{"id": "eval_00000", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTBA_GIBZE\n\n### Answer:", "output": "Tubulin is the major constituent of microtubules, a cylinder consisting of laterally associated linear protofilaments composed of alpha- and beta-tubulin heterodimers.  Microtubules grow by the addition of GTP-tubulin dimers to the microtubule end, where a stabilizing cap forms.  Below the cap, tubulin dimers are in GDP-bound state, owing to GTPase activity of alpha-tubulin.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00001", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nGL14_ARATH\n\n### Answer:", "output": "Belongs to the germin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00002", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRUVC_SHESH\n\n### Answer:", "output": "Crossover junction endodeoxyribonuclease RuvC, Holliday junction nuclease RuvC, Holliday junction resolvase RuvC.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00003", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nSFTPC SFTPB ABCA3 SFTPA1\n</cell>", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as SFTPA1, SFTPC, ABCA3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "spleen", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SFTPC", "SFTPB", "ABCA3", "SFTPA1"]}}
{"id": "eval_00004", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSAVERNIVSRLRDKGFAVVRAPAKDPIPDIIALKNGVIILIEMKSRKDIEGKIYVRREQAEGIIEFARKSGGSLFLGVKKPGVLKFIPFEKLRRTETGNYVADSEIEGLDLEDLVRLVEAKISRTLD\n\n### Answer:", "output": "<SEQ_3Di>DVVLVVLQVVLVVLVWDKDADPDPLDDARIWTDDQHATEGEHADFDADPVRWTKDAQRVLVSHVVVCVVSVYFYWYWDDDVVFIFTHGQVQFDQDPNRMTTDDPPDDGHGSVRVSVVVVVVNVVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00005", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTTSQKHRDFVAEPMGEKPVGSLAGIGEVLGKKLEERGFDKAYVVLGQFLVLKKDEDLFREWLKDTCGANAKQSRDCFGCLREWCDAFL\n\n### Answer:", "output": "<SEQ_2D>CCCCHHHHHHTTSCCTTCBGGGSTTCCHHHHHHHHHTTCCBHHHHHHHHHHTTTCHHHHHHHHHHHHCCCHHHHHHHHHHHHHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00006", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL23_TOBAC\n\n### Answer:", "output": "60S ribosomal protein L23.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00007", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRIIEGKLQLQGNEKVAILTSRFNHIITDRLKEGAMDCFKRHGGDEDLLDIVLVPGAYELPFILERLLGSGKYDGVCVLGAIIRGGTPHFDYVSAEATKGIAHAMLKYSMPVSFGVLTTDNIEQAIERAGSKAGNKGFEAMSTLIELLSLCQTLKG\n\n### Sequence 2:\nMQIIEGKLQLQGNERVAILTSRFNHIITDRLKEGAMDCFKRHGGDEKLLDIVLVPGAYELPLILDKLLESEKYDGVCVLGAIIRGGTPHFDYVSAEATKGIASAMLKYSMPVSFGVLTTDNIEQAIERAGSKAGNKGFEAMSTLIELLSLCQTLKG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00008", "category": "Mutation", "instruction": "Describe the effect of mutation O67648-A234D on protein UDP-3-O-acyl-N-acetylglucosamine deacetylase.", "input": "### Wild-type:\nMGLEKTVKEKLSFEGVGIHTGEYSKLIIHPEKEGTGIRFFKNGVYIPARHEFVVHTNHSTDLGFKGQRIKTVEHILSVLHLLEITNVTIEVIGNEIPILDGSGWEFYEAIRKNILNQNREIDYFVVEEPIIVEDEGRLIKAEPSDTLEVTYEGEFKNFLGRQKFTFVEGNEEEIVLARTFCFDWEIEHIKKVGLGKGGSLKNTLVLGKDKVYNPEGLRYENEPVRHKVFDLIGALYLLGSPVKGKFYSFRGGHSLNVKLVKELAKKQKLTRDLPHLPSVQAL\n\n### Mutant:\nMGLEKTVKEKLSFEGVGIHTGEYSKLIIHPEKEGTGIRFFKNGVYIPARHEFVVHTNHSTDLGFKGQRIKTVEHILSVLHLLEITNVTIEVIGNEIPILDGSGWEFYEAIRKNILNQNREIDYFVVEEPIIVEDEGRLIKAEPSDTLEVTYEGEFKNFLGRQKFTFVEGNEEEIVLARTFCFDWEIEHIKKVGLGKGGSLKNTLVLGKDKVYNPEGLRYENEPVRHKVFDLIGDLYLLGSPVKGKFYSFRGGHSLNVKLVKELAKKQKLTRDLPHLPSVQAL\n\n### Answer:", "output": "Almost gain of activity. 1.5-fold increase in zinc content.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00009", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMLP3A_HUMAN\n\n### Answer:", "output": "Microtubule-associated proteins 1A/1B light chain 3A, Autophagy-related protein LC3 A, Autophagy-related ubiquitin-like modifier LC3 A, MAP1 light chain 3-like protein 1, MAP1A/MAP1B light chain 3 A, MAP1A/MAP1B LC3 A, Microtubule-associated protein 1 light chain 3 alpha.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00010", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSNKGQLLQDPFLNALRKEHVPVSIYLVNGIKLQGNIESFDQYVVLLRNTVTQMVYKHAISTVVPARPVNFHPDAEASS\n\n### Sequence 2:\nMSNKGQLLQDPFLNALRKEHVPVSIYLVNGIKLQGNIESFDQYVVLLRNTVTQMVYKHAISTVVPARPVNFHPDAEASS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00011", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSTU1_ASPFU\n\n### Answer:", "output": "Belongs to the CLASP family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00012", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CC(=O)C1=C(N=[N+]=[N-])c2cccc3cccc(c23)C1=O", "output": "The SMILES represents an unnamed molecule with MW=263.26, LogP=3.65, TPSA=82.9. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "CC(=O)C1=C(N=[N+]=[N-])c2cccc3cccc(c23)C1=O", "compound_name": null, "pubchem_cid": null, "MW": 263.26, "LogP": 3.65, "TPSA": 82.9, "QED": 0.36, "functional_groups": ["aromatic ring", "ketone"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00013", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTVLTKGEIVLFALRKFAIASNASLTDVEPQSIEDGVNDLEDMMSEWMINPGDIGYAFATGDEQPLPDDESGLPRKYKHAVGYQLLLRMLSDYSLEPTPQVLSNAQRSYDALMTDTLVVPSMRRRGDFPVGQGNKYDVFTSDRYYPGDLPLIDGDI\n\n### Answer:", "output": "<SEQ_3Di>DFAFQQVLVVVACPVVPNDDPPDPDDDDVVVSLVLSVLLLVVLVVCCPDLHDQQFDDDPDDDRDDRGGTGRHDPLLSQLSSLVSNVVCVPVVPDDDDPVSVVSNVVSVVVSSVVDDCPPPDDDDAADFADCVVVCPPPNVDGGDHRDDDDDPVRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00014", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nSATB2:low SLC17A7:medium RBFOX3:high\n</cell>", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as RBFOX3, SATB2, SLC17A7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "prostate", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SATB2", "SLC17A7", "RBFOX3"]}}
{"id": "eval_00015", "category": "Mutation", "instruction": "Describe the effect of mutation P53762-A171D on protein Aryl hydrocarbon receptor nuclear translocator.", "input": "### Wild-type:\nMAATTANPEMTSDVPSLGPTIASGNPGPGIQGGGAVVQRAIKRRSGLDFDDEVEVNTKFLRCDDDQMCNDKERFARSDDEQSSADKERLARENHSEIERRRRNKMTAYITELSDMVPTCSALARKPDKLTILRMAVSHMKSLRGTGNTSTDGSYKPSFLTDQELKHLILEAADGFLFIVSCETGRVVYVSDSVTPVLNQPQSEWFGSTLYDQVHPDDVDKLREQLSTSENALTGRVLDLKTGTVKKEGQQSSMRMCMGSRRSFICRMRCGTSSVDPVSMNRLSFLRNRCRNGLGSVKEGE\n\n### Mutant:\nMAATTANPEMTSDVPSLGPTIASGNPGPGIQGGGAVVQRAIKRRSGLDFDDEVEVNTKFLRCDDDQMCNDKERFARSDDEQSSADKERLARENHSEIERRRRNKMTAYITELSDMVPTCSALARKPDKLTILRMAVSHMKSLRGTGNTSTDGSYKPSFLTDQELKHLILEDADGFLFIVSCETGRVVYVSDSVTPVLNQPQSEWFGSTLYDQVHPDDVDKLREQLSTSENALTGRVLDLKTGTVKKEGQQSSMRMCMGSRRSFICRMRCGTSSVDPVSMNRLSFLRNRCRNGLGSVKEGE\n\n### Answer:", "output": "Reduces transcription activity. Markedly reduces interaction with AHR. Impairs heterodimer formation with EPAS1. Markedly decreases heterodimer formation with HIF1A. Significantly destabilizes ARNT?s heterodimeric interactions with both NPAS1 and NPAS3. Compromises SIM1:ARNT heterodimer stability. Compromise NPAS4:ARNT heterodimer stability. Compromise AHR:ARNT heterodimer stability. The mutation in NPAS1 leads to functional changes in the protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00016", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPGGYKGECGDDVDPMPFLAPPEKERIEAMNKPYDIKKSCWVKDEKEGFIAGEIQSEQGDQVTVKTVNNQTVTVKKDDVQQMNPPKFYQASDMADMTFLNEASVLNNLRQRYTNMRIYTYSGLFCVTVNPYKWLPIYGARVANMYKGKKRTEMPPHLFSISDNAYHDMLMNRENQSMLITGESGAGKTENTKKVIQYFANVGGTGKQSSDGKGQGSLEDQIIQANPVLEAFGNAKTIRNNNSSRFGKFIRIHFGTTGKLAGADIESYLLEKSRVISQQAAERGYHIFYQILSNKKPELIESLLLVPNPKEYHWVSQGVTVVENMDDGEELQITDVAFDVLGFSAEEKIGIYKLTGGIMHFGNMKFKQKPREEQAEVDTTEVADKVSHLMGLNSGELQKGITRPRVKVGNEFVQKGQNVEQCNNSIGALGKAIYDKMFKWLVVRINKTLDTKMQRQFFIGVLDIAGFEIFEFNSFEQLCINFTNEKLQQFFNHHMFVLEQEEYKREGIEWVFIDFGLDLQACIDLLEKPMGIFSILEEQCVFPKATDATFKAALYDNHLGKSNNFLKPKGGKGKGPEAHFELVHYAGTVAYNITGWLEKNKDPLNETVVGLFQKSSLGLLALLFKEEEAPAGSKKQKRGSSFMTVSNFYREQLNKLMATLHSTAPHFVRCIVPNEFKQSGVVDAHLIMHQLACNGVLEGIRICRKGFPNRMQYPEFKQRYQVLNPNVIPQGFVDNKKASELLLGSIDLDVNEYKIGHTKVFFRAGILAKLEDMRDERLAKIMTMLQCRLRGFLMRIEFKKMLERRIGLKVIQRNTRKFLELRFWGWWKLYNKVKPLLNVARQEEEMKAKEEELRNAMSKTQELLSRVKELEEKMATLSQEKNDLTIQLQAEQENVIDAEERLTQMMKTKMDLESQISDMRERLEEEEGTAASLSATKRKLEGEMSDLKRDLEGLETTLAKTEKEKQALDHRVRTLTGDLSLREDSIAKLQKEKRALEELHQKTLDDLQAEEDKVNHLTKTNSKLSTQIHELEDNWEQEKKIRAEVEKARRKAESDLKMTIDNLNDMERSKLDLEEVVKKRDMEINSVNSKYEDEQSLNSTLQRKLKEHQARIEELEEELEAERSMRAKVEKQRSDLSRDLEDLSDRLEEAGGATSAQIEQNRKREAELLKLRRELEEAALQSEAAASTLRKKHTDSMAELTEHVENLQRVKSKLEKDKQVMKAEIDDLNASMETVQKSKMNAEAHIRKLEDSLSEANAKVAELERNQAEINAVRTRLQAENGELSREYEESQSRLNQILRIKTSLTSQVDDYKRQLDEESKSRSAAMVSLANTKHDLDLVKEQLEEEQGGKSELQRLVSKLNTEVTTWRTKYETDAIQRTEELEETKRKLAARLQEAEETAEAAQARAASLEKNKQRLQAEVEDLTIDLEKANAAAAALDKKQRVFDKMLAEWQQKCEELQVEVDSSQKECRMYMTESFKIKTAYEESLEHLESVKKENKTLQEEIKELIDQLGEGGRSVHELQKLKKKLEIEKEELQVALEEAESSLEVEESKVIRIQLELAQVKADIDRRIHEKEEEFEATRKNHQRAIESLQASLEAEAKGRAEALRLKKKMETDLNEMEIQLDHANKNNSELVKTLKRLQQQIKDLQVQMDEDARQHEELREQYNLQERRLSLLQTELEEVRSGLEGSERSRKLLEQEVVEITERHNEVNIQNQSLLVVKRKLESDVQRISSEHEELISEFRSADERAKKAMTDAARMAEELRQEQDHCMHLEKIKKNYEITIKDLQAKMEEAEQLALKGGKRTIMKLEARIKELETELDGEQKQHVETVKTLRKNERRLKELVFQTEEDHKTNQRMQELVEKLQNKLKVYKRQIEEAEEQANQTLARYRKTVHELDDAEERAGMAETALNKLRTRHRVAGKGITSV\n\n### Sequence 2:\nPFLRKSEKERLEAQTRPFDLKKDVFVPDDKEEFVKAKIVSREGGKVTAETENGKTVTVKEDQVMQQNPPKFDKIEDMAMLTFLHEPAVLYNLKERYASWMIYTYSGLFCVTVNPYKWLPVYNAEVVAAYRGKKRSEAPPHIFSISDNAYQYMLTDRENQSILITGESGAGKTVNTKRVIQYFAVIAAIGDRSKKDQTPGKGTLEDQIIQANPALEAFGNAKTVRNDNSSRFGKFIRIHFGATGKLASADIETYLLEKSRVIFQLKAERDYHIFYQILSNKKPELLDMLLITNNPYDYAFISQGETTVASIDDSEELMATDNAFDVLGFTSEEKNSMYKLTGAIMHFGNMKFKQKQREEQAEPDGTEEADKSAYLMGLNSADLLKGLCHPRVKVGNEYVTKGQNVQQVAYATGALAKAVYEKMFNWMVTRINATLETKQPRQYFIGVLDIAGFEIFDFNSFEQLCINFTNEKLQQFFNHHMFVLEQEEYKKEGIEWEFIDFGMDLQACIDLIEKPMGIMSILEEECMFPKATDMTFKAKLYDNHLGKSNNFQKPRNIKGK-QEAHFSLIHYAGTVDYNILGWLQKNKDPLNETVVALYQKSSLKLLSNLFANYAGADAPVEKGKGKAKKGSSFQTVSALHRENLNKLMTNLRSTHPHFVRCIIPNETKSPGVIDNPLVMHQLRCNGVLEGIRICRKGFPNRILYGDFRQRYRILNPAAIPEGQFIDSRKGAEKLLSSLDIDHNQYKFGHTKVFFKAGLLGLLEEMRDERLSRIITRIQAQSRGVLSRMEYKKLLERRDSLLIIQWNIRAFMGVKNWPWMKLYFKIKPLLKSAETEKEMATMKEEFARIKEALEKSE---ARRKELEEKMVSLLQEKNDLQLQVQAEQDNLADAEERCDQLIKNKIQLEAKVKEMTERLEDEEEMNAELTAKKRKLEDECSELKRDIDDLELTLAKVEKEKHATENKVKNLTEEMAGLDEIIAKLTKEKKALQEAHQQALDDLQAEEDKVNTLTKAKVKLEQQVDDLEGSLEQEKKVRMDLERAKRKLEGDLKLTQESIMDLENDKQQLDERLKKKDFELNALNARIEDEQALGSQLQKKLKELQARIEELEEELEAERTARAKVEKLRSDLSRELEEISERLEEAGGATSVQIEMNKKREAEFQKMRRDLEEATLQHEATAAALRKKHADSVAELGEQIDNLQRVKQKLEKEKSEFKLELDDVTSNMEQIIKAKANLEKMCRTLEDQMNEHRSKAEETQRSVNDLTSQRAKLQTENGELSRQLDEKEALISQLTRGKLTYTQQLEDLKRQLEEEVKAKNALAHALQSARHDCDLLREQYEEETEAKAELQRVLSKANSEVAQWRTKYETDAIQRTEELEEAKKKLAQRLQDAEEAVEAVNAKCSSLEKTKHRLQNEIEDLMVDVERSNAAAAALDKKQRNFDKILAEWKQKYEESQSELESSQKEARSLSTELFKLKNAYEESLEHLETFKRENKNLQEEISDLTEQLGSSGKTIHELEKVRKQLEAEKLELQSALEEAEASLEHEEGKILRAQLEFNQIKAEIERKLAEKDEEMEQAKRNHLRVVDSLQTSLDAETRSRNEALRVKKKMEGDLNEMEIQLSHANRMAAEAQKQVKGLQSLLKDTQIQLDDAVRANDDLKENIAIVERRNNLLQAELEELRAVVEQTERSRKLAEQELIETSERVQLLHSQNTSLINQKKKMDADLSQLQTEVEEAVQECRNAEEKAKKAITDAAMMAEELKKEQDTSAHLERMKKNMEQTIKDLQHRLDEAEQIALKGGKKQLQKLEARVRELENELEAEQKRNAESVKGMRKSERRIKELTYQTEEDRKNLLRLQDLVDKLQLKVKAYKRQAEEAEEQANTNLSKFRKVQHELDEAEERADIAESQVNKLRAKSRDIGAKGL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00017", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "LUM, COL1A2, DCN, COL3A1", "output": "The expression pattern is consistent with fibroblast in a context associated with type 2 diabetes, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "heart", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["LUM", "COL1A2", "DCN", "COL3A1"]}}
{"id": "eval_00018", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "P2RY12, CSF1R, TMEM119", "output": "The expression pattern is consistent with microglia in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "breast", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["P2RY12", "CSF1R", "TMEM119"]}}
{"id": "eval_00019", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRSXC_ECO55\n\n### Answer:", "output": "Part of a membrane-bound complex that couples electron transfer with translocation of ions across the membrane.  Required to maintain the reduced state of SoxR.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00020", "category": "Mutation", "instruction": "Describe the effect of mutation Q66915-E124D on protein Capsid protein.", "input": "### Wild-type:\nMCSTCANVLKYYNWDPHFKLVINPNKFLSIGFCDNPLMCCYPELLPEFGTVWDCDQSPLQIYLESILGDDEWSSTYEAIDPVVPPMHWNEAGKIFQPHPGVLMHHIIGEVAKAWDPNLPLFRLEADDGSITAPEQGTVVGGVIAEPSSQMSTAADMASGKSVDSEWEAFFSFHTSVNWSTSETQGKILFKQSLGPLLNPYLEHLSKLYVAWSGSVEVRFSISGSGVFGGKLAAIVVPPGVDPIQSTSMLQYPHVLFDARQVEPVIFTIPDLRSTLYHLMSDTDTTSLVIMVYNDLINPYA\n\n### Mutant:\nMCSTCANVLKYYNWDPHFKLVINPNKFLSIGFCDNPLMCCYPELLPEFGTVWDCDQSPLQIYLESILGDDEWSSTYEAIDPVVPPMHWNEAGKIFQPHPGVLMHHIIGEVAKAWDPNLPLFRLDADDGSITAPEQGTVVGGVIAEPSSQMSTAADMASGKSVDSEWEAFFSFHTSVNWSTSETQGKILFKQSLGPLLNPYLEHLSKLYVAWSGSVEVRFSISGSGVFGGKLAAIVVPPGVDPIQSTSMLQYPHVLFDARQVEPVIFTIPDLRSTLYHLMSDTDTTSLVIMVYNDLINPYA\n\n### Answer:", "output": "Partial loss of precursor cleavage by calcivirin.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00021", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGKITLYEDRGFQGRHYECSSDHPNLQPYLSRCNSARVDSGCRMLYEQPNYSGLQYFLRRGDYADHQQWMGLSDSVRSCRLIPHSGSHRIRLYEREDYRGQMIEFTEDCSCLQDRFRFNEIHSLNVLEGSWVLYELSNYRGRQYLLMPGDYRRYQDWGATNARVGSLRRVIDFS\n\n### Answer:", "output": "<SEQ_3Di>DKKKFAQAFPRDHFIDIDDFKDQFCVVRHVFTFWMAAQAFKKWFALAGRPAHFIFIDFGDTGRGVCVSPDPGRRTTIMGGFDDFPWWKKKFAPAFQRDDDIDMDGFWFQFVCVVPVDQFTFWMATATAKKWWALAGRNDHTIFIGGGGTGRGVCVRRDPGGGTTIMHGDDHDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00022", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCD3D CD4 S100A4 IL7R CD45RO\n</cell>", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD3D, CD45RO, IL7R support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "bone marrow", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD3D", "CD4", "S100A4", "IL7R", "CD45RO"]}}
{"id": "eval_00023", "category": "Mutation", "instruction": "Describe the effect of mutation Q13637-M91S on protein Ras-related protein Rab-32.", "input": "### Wild-type:\nMAGGGAGDPGLGAAAAPAPETREHLFKVLVIGELGVGKTSIIKRYVHQLFSQHYRATIGVDFALKVLNWDSRTLVRLQLWDIAGQERFGNMTRVYYKEAVGAFVVFDISRSSTFEAVLKWKSDLDSKVHLPNGSPIPAVLLANKCDQNKDSSQSPSQVDQFCKEHGFAGWFETSAKDNINIEEAARFLVEKILVNHQSFPNEENDVDKIKLDQETLRAENKSQCC\n\n### Mutant:\nMAGGGAGDPGLGAAAAPAPETREHLFKVLVIGELGVGKTSIIKRYVHQLFSQHYRATIGVDFALKVLNWDSRTLVRLQLWDIAGQERFGNSTRVYYKEAVGAFVVFDISRSSTFEAVLKWKSDLDSKVHLPNGSPIPAVLLANKCDQNKDSSQSPSQVDQFCKEHGFAGWFETSAKDNINIEEAARFLVEKILVNHQSFPNEENDVDKIKLDQETLRAENKSQCC\n\n### Answer:", "output": "Impairs interaction with ANKRD27; when associated with S-93. The mutation in VARP affects its binding to Rab32:GTP, which in turn regulates VAMP7 participation in SNARE complex formation and membrane fusion events.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00024", "category": "Mutation", "instruction": "Describe the effect of mutation P51692-F699Y on protein Signal transducer and activator of transcription 5B.", "input": "### Wild-type:\nMAVWIQAQQLQGEALHQMQALYGQHFPIEVRHYLSQWIESQAWDSVDLDNPQENIKATQLLEGLVQELQKKAEHQVGEDGFLLKIKLGHYATQLQNTYDRCPMELVRCIRHILYNEQRLVREANNGSSPAGSLADAMSQKHLQINQTFEELRLVTQDTENELKKLQQTQEYFIIQYQESLRIQAQFGPLAQLSPQERLSRETALQQKQVSLEAWLQREAQTLQQYRVELAEKHQKTLQLLRKQQTIILDDELIQWKRRQQLAGNGGPPEGSLDVLQSWCEKLAEIIWQNRQQIRRAEHLC\n\n### Mutant:\nMAVWIQAQQLQGEALHQMQALYGQHFPIEVRHYLSQWIESQAWDSVDLDNPQENIKATQLLEGLVQELQKKAEHQVGEDGFLLKIKLGHYATQLQNTYDRCPMELVRCIRHILYNEQRLVREANNGSSPAGSLADAMSQKHLQINQTFEELRLVTQDTENELKKLQQTQEYFIIQYQESLRIQAQFGPLAQLSPQERLSRETALQQKQVSLEAWLQREAQTLQQYRVELAEKHQKTLQLLRKQQTIILDDELIQWKRRQQLAGNGGPPEGSLDVLQSWCEKLAEIIWQNRQQIRRAEHLC\n\n### Answer:", "output": "Enhances phosphorylation by HCK. The mutation in STAT5B promotes BCR/ABL-dependent activation of STAT5 and elevation of expression of STAT5 downstream effectors A1 and pim-1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00025", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSDNDELQQIAHLRREYTKGGLRRRDLPADPLTLFERWLSQACEAKLADPTAMVVATVDEHDQPYQRIVLLKHYDEKGMVFYTNLGSRKAHQIENNPRVSLLFPWHTLERQVMVIGKAERLSTLEVMKYFHSRPRDSQIGAWVSKQSSRISARGILESKFLELKQKFQQGEVPLPSFWGGFRVSLEQIEFWQGGEHRLHDRFLYQRENDAWKIDRLAP\n\n### Sequence 2:\nMLTTGSKNLDALLGGGIDKGILTQVYGPFATGKTTLAMQIGLLNEGKVAYIDTEGGFSPERLAKMVESRGMDSNSTLQKFLIFEAFDFKEQKKTISNLKKIVNEKFSMIVVDSITNHYRVEEKKSSMTTDLGKQLQVLLWLARKYNLAVIVTNQVYFDSKQNALKPLAEHIMGYKCKDILRLEKLRPGLRIAVLERHRFKPEGGIVHFEITDKGIEDIEKIKSSQTTL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00026", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "CD38, JCHAIN, MZB1", "output": "This cell is most consistent with plasma cell, an antibody-secreting plasma cell. Key markers such as CD38, MZB1, JCHAIN support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasma cell", "tissue": "spleen", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD38", "JCHAIN", "MZB1"]}}
{"id": "eval_00027", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nDEDETTALVCDNGSGLVKAGFAGDDAPRAVFPSIVGRPRHQGVMVGMGQKDSYVGDEAQSKRGILTLKYPIEHGIITNWDDMEKIWHHTFYNELRVAPEEHPTLLTEAPLNPKANREKMTQIMFETFNVPAMYVAIQAVLSLYASGRTTGIVLDSGDGVTHNVPIYEGYALPHAIMRLDLAGRDLTDYLMKILTERGYSFVTTAEREIVRDIKEKLCYVALDFENEMATAASSSSLEKSYELPDGQVITIGNERFRCPETLFQPSFIGMESAGIHETTYNSIMKCDIDIRKDLYANNVMSGGTTMYPGIADRMQKEITALAPSTMKIKIIAPPERKYSVWIGGSILASLSTFQQMWITKQEYDEAGPSIVHRKCF\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCEEEEECSSEEEEEETTCSSCSEEEECCEEEESSSCCCCTTTCCSCEETHHHHHTGGGEEEECSEETTEECCHHHHHHHHHHHHHTTTCCCGGGSCEEEEECTTCCHHHHHHHHHHHHHTSCCSEEEEEEHHHHHHHHHTCSSEEEEEEETTEEEEEEEETTEECGGGCEEESCSHHHHHHHHHHHHGGGTCCCCSHHHHHHHHHHHHHHCCCCSSHHHHHHHTTSCSTTCEEEECTTSCEEEECTHHHHHHHTTTCGGGGTCCSCCHHHHHHHHHHTSCHHHHHHHHTCEEECSGGGGSTTHHHHHHHHHHHHSCTTSCCCEECCTTGGGHHHHHHHHHHHCTGGGGGCEEHHHHHHHCGGGGGCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00028", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nATPB_BUXMI\n\n### Answer:", "output": "Produces ATP from ADP in the presence of a proton gradient across the membrane.  The catalytic sites are hosted primarily by the beta subunits.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00029", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: COC(=O)C1CCC(=O)N1C(c1cc(OC)c(OC)c(OC)c1)c1cccc2ccccc12\nDescriptors: MW=449.5, LogP=4.12, TPSA=74.3, HBD=0, HBA=6", "output": "Descriptors: MW=449.5, LogP=4.12, TPSA=74.3, HBD=0, HBA=6, QED=0.51, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "COC(=O)C1CCC(=O)N1C(c1cc(OC)c(OC)c(OC)c1)c1cccc2ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 449.5, "LogP": 4.12, "TPSA": 74.3, "QED": 0.51, "functional_groups": ["aromatic ring", "ester", "amide", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00030", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTBP_TOBAC\n\n### Answer:", "output": "General transcription factor that functions at the core of the DNA-binding multiprotein factor TFIID.  Binding of TFIID to the TATA box is the initial transcriptional step of the pre-initiation complex (PIC), playing a role in the activation of eukaryotic genes transcribed by RNA polymerase II.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00031", "category": "Mutation", "instruction": "Describe the effect of mutation P01130-N564H on protein Low-density lipoprotein receptor.", "input": "### Wild-type:\nMGPWGWKLRWTVALLLAAAGTAVGDRCERNEFQCQDGKCISYKWVCDGSAECQDGSDESQETCLSVTCKSGDFSCGGRVNRCIPQFWRCDGQVDCDNGSDEQGCPPKTCSQDEFRCHDGKCISRQFVCDSDRDCLDGSDEASCPVLTCGPASFQCNSSTCIPQLWACDNDPDCEDGSDEWPQRCRGLYVFQGDSSPCSAFEFHCLSGECIHSSWRCDGGPDCKDKSDEENCAVATCRPDEFQCSDGNCIHGSRQCDREYDCKDMSDEVGCVNVTLCEGPNKFKCHSGECITLDKVCNMAR\n\n### Mutant:\nMGPWGWKLRWTVALLLAAAGTAVGDRCERNEFQCQDGKCISYKWVCDGSAECQDGSDESQETCLSVTCKSGDFSCGGRVNRCIPQFWRCDGQVDCDNGSDEQGCPPKTCSQDEFRCHDGKCISRQFVCDSDRDCLDGSDEASCPVLTCGPASFQCNSSTCIPQLWACDNDPDCEDGSDEWPQRCRGLYVFQGDSSPCSAFEFHCLSGECIHSSWRCDGGPDCKDKSDEENCAVATCRPDEFQCSDGNCIHGSRQCDREYDCKDMSDEVGCVNVTLCEGPNKFKCHSGECITLDKVCNMAR\n\n### Answer:", "output": "In FHCL1. The mutation in the LDLR gene is a known cause of familial hypercholesterolemia. This mutation disrupts the structure of the LDLR protein, impairing its ability to bind to LDL particles and remove them from the bloodstream. As a result, individuals with this mutation have elevated levels of LDL cholesterol. Both mutations cosegregated with hypercholesterolemia within the families. Each of these mutations had little or no effect on receptor function in transfected COS cells, but when both mutations were present simultaneously, receptor function, as assessed by flow cytometric measurement of fluorescent LDL uptake in cells, was reduced by 75%. Immunostainable receptors on the cell surface were decreased by 80% as measured by flow cytometry. The two mutations therefore acted in synergy to affect receptor function, possibly during intracellular receptor transport, since Northern blot analysis suggested that mRNA levels were unaffected.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00032", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAKLTLEVEQYTQNLREKLQEAGYSYYVLDQPIMEDAVYDHLLRELQELESQYPQLIVSDSPTQRVGEKPATKFVSVKHNIPLYSLENAFNIQELKSWQERWQRQTLLENQRLTLPLRGEVKIKNSIDESDNFSTVNSLAPVITGSNFDEIDLASVDYICELKIDGSALALTYENGLLVRGATRGDGIMGEDITQNVKTINSIPLKLKVDNPPLLVEVRGEAFLSINVFENINAEREKIGEIIFANPRNAAAGTLRQLDSKIVAKRRLDFFAYTLYLEENESNFLQFYTQSQCLELLKKLGFKVNPNGQVCSSLNEVEAYYQYWDKQRENLPYLTDGVVVKLNSLSLQEKLGFTQKYPRWAIALKYPAEEIPTIVQGVTVQVGRTGAITPVAELKPVQLAGTTVQRASLHNSDRLAELDLHIGDTVIVRKAGEIIPEVVRVLPELRPKKAQRFQMPTHCPECSQPLIKPRNEAVTRCINTSCPAILRGSLAHWASRAALDINGLGEKLVQQLVNSGLVKSVADLYNLTMEDLTSLERMGKKSANKLIEAMGTSKAQPWRRVLYGLGIRHVGTVNAELLTQKFPTVAQLSQAKATDIETVHGIGPEIAQYVEQWFRVPANQNLVERLKIAGVTMETKDSTSGMAKLTETVLAIQNLAARQNLVARQNLENLVERLKIAGVTMETKDSTSDVVEVQSGLLRGKTFVLTGTLPTMRRDEAKNLIQNAGGKVTSSVSSKTDFIVVGEDGGSKLEKAKKLGVKELSESELLEALAVTGI\n\n### Sequence 2:\nMAGVSAVSKVSSLVCDLSSTSGLIRRTANPHPNVWGYDLVHSLKSPYIDSSYRERAEVLVSEIKAMLNPAITGDGESMITPSAYDTAWVARVPAIDGSARPQFPQTVDWILKNQLKDGSWGIQSHFLLSDRLLATLSCVLVLLKWNVGDLQVEQGIEFIKSNLELVKDETDQDSLVTDFEIIFPSLLREAQSLRLGLPYDLPYIHLLQTKRQERLAKLSREEIYAVPSPLLYSLEGIQDIVEWERIMEVQSQDGSFLSSPASTACVFMHTGDAKCLEFLNSVMIKFGNFVPCLYPVDLLERLLIVDNIVRLGIYRHFEKEIKEALDYVYRHWNERGIGWGRLNPIADLETTALGFRLLRLHRYNVSPAIFDNFKDANGKFICSTGQFNKDVASMLNLYRASQLAFPGENILDEAKSFATKYLREALEKSETSSAWNNKQNLSQEIKYALKTSWHASVPRVEAKRYCQVYRPDYARIAKCVYKLPYVNNEKFLELGKLDFNIIQSIHQEEMKNVTSWFRDSGLPLFTFARERPLEFYFLVAAGTYEPQYAKCRFLFTKVACLQTVLDDMYDTYGTLDELKLFTEAVRRWDLSFTENLPDYMKLCYQIYYDIVHEVAWEAEKEQGRELVSFFRKGWEDYLLGYYEEAEWLAAEYVPTLDEYIKNGITSIGQRILLLSGVLIMDGQLLSQEALEKVDYPGRRVLTELNSLISRLADDTKTYKAEKARGELASSIECYMKDHPECTEEEALDHIYSILEPAVKELTREFLKPDDVPFACKKMLFEETRVTMVIFKDGDGFGVSKLEVKDHIKECLIEPLPL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00033", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMHGRLKVKTSEEQAEAKRLEREQKLKLYQSATQAVFQKRQAGELDESVLELTSQILGANPDFATLWNCRREVLQQLETQKSPEELAALVKAELGFLESCLRVNPKSYGTWHHRCWLLGRLPEPNWTRELELCARFLEVDERNFHCWDYRRFVATQAAVPPAEELAFTDSLITRNFSNYSSWHYRSCLLPQLHPQPDSGPQGRLPEDVLLKELELVQNAFFTDPNDQSAWFYHRWLLGRADPQDALRCLHVSRDEACLTVSFSRPLLVGSRMEILLLMVDDSPLIVEWRTPDGRNRPSHVWLCDLPAASLNDQLPQHTFRVIWTAGDVQKECVLLKGRQEGWCRDSTTDEQLFRCELSVEKSTVLQSELESCKELQELEPENKWCLLTIILLMRALDPLLYEKETLQYFQTLKAVDPMRATYLDDLRSKFLLENSVLKMEYAEVRVLHLAHKDLTVLCHLEQLLLVTHLDLSHNRLRTLPPALAALRCLEVLQASDNAIESLDGVTNLPRLQELLLCNNRLQQPAVLQPLASCPRLVLLNLQGNPLCQAVGILEQLAELLPSVSSVLT\n\n### Sequence 2:\nMHGRLKVKTSEEQAEAKRLEREQKLKLYQSATQAVFQKRQAGELDESVLELTSQILGANPDFATLWNCRREVLQHLETEKSPEESAALVKAELGFLESCLRVNPKSYGTWHHRCWLLSRLPEPNWARELELCARFLEADERNFHCWDYRRFVAAQAAVAPAEELAFTDSLITRNFSNYSSWHYRSCLLPQLHPQPDSGPQGRLPENVLLKELELVQNAFFTDPNDQSAWFYHRWLLGRAEPHDVLCCVHVSREEACLSVCFSRPLTVGSRMGTLLLMVDEAPLSVEWRTPDGRNRPSHVWLCDLPAASLNDQLPQHTFRVIWTGSDSQKECVLLKDRPECWCRDSATDEQLFRCELSVEKSTVLQSELESCKELQELEPENKWCLLTIILLMRALDPLLYEKETLQYFSTLKAVDPMRAAYLDDLRSKFLLENSVLKMEYADVRVLHLAHKDLTVLCHLEQLLLVTHLDLSHNRLRALPPALAALRCLEVLQASDNALENVDGVANLPRLQELLLCNNRLQQSAAIQPLVSCPRLVLLNLQGNSLCQEEGIQERLAEMLPSVSSILT\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00034", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKLPIYLDYSATCPVDPRVAEKMVQYMTMDGTFGNPASRSHRYGWQAEEAVDTAREQIADLINADPREIVFTSGATESDNLAIKGAAHFYSKKGKHVITCKTEHKAVLDPCRQLEREGFEVTYLEPESNGLVDLSKLQAAMREDTVLVSIMHVNNEIGVIQDITAIGDLCRERKIVFHVDAAQSAGKLPIDVQEMKVDLISFSAHKAYGPKGIGALYVRRKPRIRLEAQMHGGGHERGFRSGTLPTHQIVGMGEAFRVAKEDMQKDYDHALALRNRLLDGVKDLEAVTVNGDLEQRVPHNLNVSFAFVEGESLLMSLKDLAVSSGSACTSASLEPSYVLRALGLDDELAHSSVRFSFGRFTTEAEIDYAIEQIRVAVTKLRDMSPLWDMYKEGIDLSTVEWAHH\n\n### Sequence 2:\nMSCSQNKTSVSLAWRECISIASVLIGAYASYKYYKLFKTRDIPRPKEGVEELIGNTPLVKIRSLTKATGVNIYAKLELCNPAGSAKDRVALNIIKTAEELGELVRGEPGWVFEGTSGSTGISIAVVCNALGYRAHISLPDDTSLEKLALLESLGATVNKVKPASIVDPNQYVNAAKKACNELKKSGNGIRAVFADQFENEANWKVHYQTTGPEIAHQTKGNIDAFIAGCGTGGTITGVAKFLKERAKIPCHVVLADPQGSGFYNRVNYGVMYDYVEKEGTRRRHQVDTIVEGIGLNRITHNFHMGEKFIDESIRVNDNQAIRMAKYLSVNDGLFVGSSTAINAVAAIQVAKTLPHGSNIVIIACDSGSRHLSKFWKEAKEIDHDVSLEEVINI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00035", "category": "Mutation", "instruction": "Describe the effect of mutation P70271-K272A on protein PDZ and LIM domain protein 4.", "input": "### Wild-type:\nMTHSVTLRGPSPWGFRLVGGRDFSAPLTISRVHAGSKAALAALCPGDLIQAINGESTELMTHLEAQNRIKGCHDHLTLSVSRPENKNWPSAPDDKAQAHRIHIDPESQDCSPATSRRSSVSGISLEDNRSGLGSPYGQPPRLPVPHNGSSNEATLPAQMSALHVSPPTSADTARVLPRNRDCRVDLGSEVYRMLREPAEPTASEPKQSGSFRYLQGMLEAGEGGDRPGSGGPRNLKPAASKLGAPLSGLQGLPECTRCGHGIVGTIVKARDKLYHPECFMCSDCGLNLKQRGYFFLDERL\n\n### Mutant:\nMTHSVTLRGPSPWGFRLVGGRDFSAPLTISRVHAGSKAALAALCPGDLIQAINGESTELMTHLEAQNRIKGCHDHLTLSVSRPENKNWPSAPDDKAQAHRIHIDPESQDCSPATSRRSSVSGISLEDNRSGLGSPYGQPPRLPVPHNGSSNEATLPAQMSALHVSPPTSADTARVLPRNRDCRVDLGSEVYRMLREPAEPTASEPKQSGSFRYLQGMLEAGEGGDRPGSGGPRNLKPAASKLGAPLSGLQGLPECTRCGHGIVGTIVKARDALYHPECFMCSDCGLNLKQRGYFFLDERL\n\n### Answer:", "output": "No loss of interaction with PTPN13 (via PDZ domains); when associated with A-274. Deletion of the extreme RIL C-terminus did not eliminate binding, suggesting the presence of a PDZ binding site within the RIL LIM moiety.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00036", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nXGPT_ACTSZ\n\n### Answer:", "output": "Xanthine-guanine phosphoribosyltransferase, XGPRT, Xanthine phosphoribosyltransferase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00037", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKQPSALSALVEALRALPGVGPKSAQRMAYHLMQHDREGAERLGRSLLFATEHLRHCDKCNTFTEAQICEVCSDEERDPTLLCVVETPADQIMLEQTMTYRGLYFVLMGRLSPLDGIGPKEIHFDRLVRRASDGIVKEVVLATNFTNEGEATAHYLGQTLKARGLAVTRLARGVPVGGELEYVDAGTIARAMLDRRTM\n\n### Sequence 2:\nMKQPSALSALVEALRVLPGVGPKSAQRMAVHLMQHDREGAERLGRSLLFATEHLQHCEKCNTFTEAQICEVCSDEERDPTLLCVVETPADQIMLEQTMTYRGLYFVLMGRLSPLDGIGPKEIHFDRLVRRASDGVVKEVVLATNFTNEGEATAHYLGQTLKARGLAVTRLARGVPVGGELEYVDAGTIARAMLDRRTM\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00038", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CCN(C)C(=O)c1ccc(C(=C2CCN(Cc3ccc(F)cc3)CC2)c2ccccc2NC(C)=O)cc1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, amine, halide. Its descriptor profile: MW=499.63, LogP=5.97, TPSA=52.65, HBD=1, HBA=3, QED=0.44.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "CCN(C)C(=O)c1ccc(C(=C2CCN(Cc3ccc(F)cc3)CC2)c2ccccc2NC(C)=O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 499.63, "LogP": 5.97, "TPSA": 52.65, "QED": 0.44, "functional_groups": ["aromatic ring", "amide", "amine", "halide"], "moleculenet_labels": {"label": 2.79}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00039", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAROA_PROM3\n\n### Answer:", "output": "3-phosphoshikimate 1-carboxyvinyltransferase, 5-enolpyruvylshikimate-3-phosphate synthase, EPSP synthase, EPSPS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00040", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "SFTPC, SFTPA1, SFTPB", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as SFTPC, SFTPA1, SFTPB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "prostate", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SFTPC", "SFTPA1", "SFTPB"]}}
{"id": "eval_00041", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMRLNSLSPAEGAKHSAKRLGRGIGSGLGKTGGRGHKGQKSRTGGGVRRGFEGGQMPLYRRLPKFGFTSLKSFHVAEIRLNDLAKVDGNEVTLESLKAANVITKDILSVKVILAGKIEKAVVVKGLGVTKGAKAAIEAAGGSIEE\n\n### Sequence 2:\nMRLNSLSPAEGAKHSAKRLGRGIGSGLGKTGGRGHKGQKSRTGGGVRRGFEGGQMPLYRRLPKFGFTSLKSLHVAEIRLNDLAKVDGNEVTLEALKAANIITKNILSVKVILAGKIERALVIKGLRVTKGAKAAIEAVGGSIEE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00042", "category": "Mutation", "instruction": "Describe the effect of mutation Q9GZX7-R112D on protein Single-stranded DNA cytosine deaminase.", "input": "### Wild-type:\nMDSLLMNRRKFLYQFKNVRWAKGRRETYLCYVVKRRDSATSFSLDFGYLRNKNGCHVELLFLRYISDWDLDPGRCYRVTWFTSWSPCYDCARHVADFLRGNPNLSLRIFTARLYFCEDRKAEPEGLRRLHRAGVQIAIMTFKDYFYCWNTFVENHERTFKAWEGLHENSVRLSRQLRRILLPLYEVDDLRDAFRTLGL\n\n### Mutant:\nMDSLLMNRRKFLYQFKNVRWAKGRRETYLCYVVKRRDSATSFSLDFGYLRNKNGCHVELLFLRYISDWDLDPGRCYRVTWFTSWSPCYDCARHVADFLRGNPNLSLRIFTADLYFCEDRKAEPEGLRRLHRAGVQIAIMTFKDYFYCWNTFVENHERTFKAWEGLHENSVRLSRQLRRILLPLYEVDDLRDAFRTLGL\n\n### Answer:", "output": "Greatly reduced nuclear import; when associated with A-193.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00043", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: BACE1\nLigand SMILES: O=c1ccn(C2OC(CO)C(O)C2[Se]c2ccccc2)c(=O)[nH]1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "O=c1ccn(C2OC(CO)C(O)C2[Se]c2ccccc2)c(=O)[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 383.26, "LogP": -1.39, "TPSA": 104.55, "QED": 0.56, "functional_groups": ["aromatic ring", "hydroxyl", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_00044", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCC(C)C1NC(=O)C(Cc2ccc(O)cc2)NC(=O)C(N)CSSCC(C(=O)N2CCCC2C(=O)NC(CC(C)C)C(=O)NCC(N)=O)NC(=O)C(CC(N)=O)NC(=O)C(CCC(N)=O)NC1=O", "output": "This molecule is relatively large and polar, with hydrophilic character. It contains aromatic ring, amide, amine, hydroxyl, phenol. Its descriptor profile: MW=1007.21, LogP=-3.61, TPSA=399.53, HBD=12, HBA=15, QED=0.07.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "SIDER", "canonical_smiles": "CCC(C)C1NC(=O)C(Cc2ccc(O)cc2)NC(=O)C(N)CSSCC(C(=O)N2CCCC2C(=O)NC(CC(C)C)C(=O)NCC(N)=O)NC(=O)C(CC(N)=O)NC(=O)C(CCC(N)=O)NC1=O", "compound_name": null, "pubchem_cid": null, "MW": 1007.21, "LogP": -3.61, "TPSA": 399.53, "QED": 0.07, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "phenol"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 1, "Infections and infestations": 0, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 0, "Renal and urinary disorders": 0, "Pregnancy, puerperium and perinatal conditions": 1, "Ear and labyrinth disorders": 0, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00045", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: COc1cc2c(Nc3ccc(NC(=O)c4ccccc4)cc3)ncnc2cc1OCCCN1CCOCC1\nDescriptors: MW=513.6, LogP=4.74, TPSA=97.84, HBD=2, HBA=8", "output": "Descriptors: MW=513.6, LogP=4.74, TPSA=97.84, HBD=2, HBA=8, QED=0.29, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "COc1cc2c(Nc3ccc(NC(=O)c4ccccc4)cc3)ncnc2cc1OCCCN1CCOCC1", "compound_name": null, "pubchem_cid": null, "MW": 513.6, "LogP": 4.74, "TPSA": 97.84, "QED": 0.29, "functional_groups": ["aromatic ring", "amide", "amine", "ether"], "moleculenet_labels": {"label": 3.5}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00046", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "MS4A1, CD27, CD79A", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as MS4A1, CD79A, CD27 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "blood", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["MS4A1", "CD27", "CD79A"]}}
{"id": "eval_00047", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYI24_ECOLX\n\n### Answer:", "output": "Insertion element IS2 uncharacterized 16.4 kDa protein, ORF4.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00048", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nENV_EIAV5\n\n### Answer:", "output": "Envelope glycoprotein, Env polyprotein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00049", "category": "Mutation", "instruction": "Describe the effect of mutation P35398-C288Q on protein Nuclear receptor ROR-alpha.", "input": "### Wild-type:\nMESAPAAPDPAASEPGSSGADAAAGSRETPLNQESARKSEPPAPVRRQSYSSTSRGISVTKKTHTSQIEIIPCKICGDKSSGIHYGVITCEGCKGFFRRSQQSNATYSCPRQKNCLIDRTSRNRCQHCRLQKCLAVGMSRDAVKFGRMSKKQRDSLYAEVQKHRMQQQQRDHQQQPGEAEPLTPTYNISANGLTELHDDLSNYIDGHTPEGSKADSAVSSFYLDIQPSPDQSGLDINGIKPEPICDYTPASGFFPYCSFTNGETSPTVSMAELEHLAQNISKSHLETCQYLREELQQITW\n\n### Mutant:\nMESAPAAPDPAASEPGSSGADAAAGSRETPLNQESARKSEPPAPVRRQSYSSTSRGISVTKKTHTSQIEIIPCKICGDKSSGIHYGVITCEGCKGFFRRSQQSNATYSCPRQKNCLIDRTSRNRCQHCRLQKCLAVGMSRDAVKFGRMSKKQRDSLYAEVQKHRMQQQQRDHQQQPGEAEPLTPTYNISANGLTELHDDLSNYIDGHTPEGSKADSAVSSFYLDIQPSPDQSGLDINGIKPEPICDYTPASGFFPYCSFTNGETSPTVSMAELEHLAQNISKSHLETQQYLREELQQITW\n\n### Answer:", "output": "Less effect on transcriptional activity with cholesterol sulfate as substrate as compared to cholesterol as substrate. The mutation results in an increased affinity of cholesterol sulfate when compared with cholesterol.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00050", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: HIV protease\nLigand SMILES: CC(C)Cn1c(=O)n(C)c(=O)c2c(C(=O)N3C[C@H](O)CO3)c(Cc3ccccc3C(F)(F)F)sc21\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "CC(C)Cn1c(=O)n(C)c(=O)c2c(C(=O)N3C[C@H](O)CO3)c(Cc3ccccc3C(F)(F)F)sc21", "compound_name": null, "pubchem_cid": null, "MW": 511.52, "LogP": 2.78, "TPSA": 93.77, "QED": 0.57, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide"], "moleculenet_labels": {"label": 3.15}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00051", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMRSTPLGTTAVSPASIKHNSYGYGRFVSSSGVSNFSIHRRRRHSSFSISQAPSQINSGACNASQIVDLFPAVSPEIVVREARLEDCWEVAETHCSSFFPGYSFPLDVVLRVDRLMAMVMGFSIPPGCQRTCLVAVIGSSVDETICFGSDDFKIGAFDAKISLNKGYVAGILTVDTVADYLPRKGPLRQRRTGIAYVSNVAVRENFRRKGIAKRLIWKAEALAKNWGCRAIGLHCDLNNLGATKLYKDQGFRSIKIPEGATWPQPKTSPDTRFTFMMKLVNNNNTQALEQFR\n\n### Sequence 2:\nMGSKEDVGKGCPAAGGVSSFTIQSILGGGPSEAPREPAGWPARKRSLSVSSEEEEPEEGWKAPACFCPDPHGPKEPSPKHHTPIPFPCLGTPKGSGGAGPAASERTPFLSPSHPDFKEEKERLLPAGSPSPGPERPRDGGAERQTGAAKKKTRTVFSRSQVYQLESTFDMKRYLSSSERACLASSLQLTETQVKTWFQNRRNKWKRQLSAELEAANMAHASAQTLVGMPLVFRDSSLLRVPVPRSLAFPAPLYYPSSNLSALPLYNLYNKLDY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00052", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nZFP1_PONAB\n\n### Answer:", "output": "Zinc finger protein 1 homolog, Zfp-1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00053", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nDAH42_RANDH\n\n### Answer:", "output": "Secreted.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00054", "category": "Mutation", "instruction": "Describe the effect of mutation Q9VHV5-T34N on protein ADP-ribosylation factor-like protein 8.", "input": "### Wild-type:\nMLALINRILEWFKSIFWKEEMELTLVGLQFSGKTTFVNVIASGQFAEDMIPTVGFNMRKITRGNVTIKVWDIGGQPRFRSMWERYCRGVNAIVYMVDAADLDKLEASRNELHSLLDKPQLAGIPVLVLGNKRDLPGALDETGLIERMNLSSIQDREICCYSISCKEKDNIDITLQWLIQHSKSQSR\n\n### Mutant:\nMLALINRILEWFKSIFWKEEMELTLVGLQFSGKNTFVNVIASGQFAEDMIPTVGFNMRKITRGNVTIKVWDIGGQPRFRSMWERYCRGVNAIVYMVDAADLDKLEASRNELHSLLDKPQLAGIPVLVLGNKRDLPGALDETGLIERMNLSSIQDREICCYSISCKEKDNIDITLQWLIQHSKSQSR\n\n### Answer:", "output": "Probably constitutively inactive (GDP-locked form). Abolishes interaction with CG11448. The small GTPase Arl8 has emerged as a major regulatory GTPase on lysosomes. Studies in mammalian cells have shown that it regulates both fusion with late endosomes and also lysosomal motility.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00055", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "CXCR2, S100A9, S100A8, FCGR3B", "output": "This profile is most consistent with neutrophil, which is commonly found in heart under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "heart", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CXCR2", "S100A9", "S100A8", "FCGR3B"]}}
{"id": "eval_00056", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "NC(=O)NNC(N)=O", "output": "The molecule contains the following functional groups: amide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "NC(=O)NNC(N)=O", "compound_name": null, "pubchem_cid": null, "MW": 118.1, "LogP": -1.76, "TPSA": 110.24, "QED": 0.28, "functional_groups": ["amide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": null, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amide"]}}
{"id": "eval_00057", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "Nc1ccccn1", "output": "The molecule contains the following functional groups: pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "Nc1ccccn1", "compound_name": null, "pubchem_cid": null, "MW": 94.12, "LogP": 0.66, "TPSA": 38.91, "QED": 0.51, "functional_groups": ["pyridine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["pyridine"]}}
{"id": "eval_00058", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTRUB_RICFE\n\n### Answer:", "output": "tRNA pseudouridine synthase B, tRNA pseudouridine(55) synthase, Psi55 synthase, tRNA pseudouridylate synthase, tRNA-uridine isomerase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00059", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNNRE_BOVIN\n\n### Answer:", "output": "NAD(P)H-hydrate epimerase, Apolipoprotein A-I-binding protein, AI-BP, NAD(P)HX epimerase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00060", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: CC1(C)N=C(C2CCCN2C(=O)OCc2ccccc2)OC1=O\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "CC1(C)N=C(C2CCCN2C(=O)OCc2ccccc2)OC1=O", "compound_name": null, "pubchem_cid": null, "MW": 316.36, "LogP": 2.52, "TPSA": 68.2, "QED": 0.8, "functional_groups": ["aromatic ring", "ester", "amide", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00061", "category": "Mutation", "instruction": "Describe the effect of mutation O00358-R102C on protein Forkhead box protein E1.", "input": "### Wild-type:\nMTAESGPPPPQPEVLATVKEERGETAAGAGVPGEATGRGAGGRRRKRPLQRGKPPYSYIALIAMAIAHAPERRLTLGGIYKFITERFPFYRDNPKKWQNSIRHNLTLNDCFLKIPREAGRPGKGNYWALDPNAEDMFESGSFLRRRKRFKRSDLSTYPAYMHDAAAAAAAAAAAAAAAAIFPGAVPAARPPYPGAVYAGYAPPSLAAPPPVYYPAASPGPCRVFGLVPERPLSPELGPAPSGPGGSCAFASAGAPATTTGYQPAGCTGARPANPSAYAAAYAGPDGAYPQGAGSAIFAAA\n\n### Mutant:\nMTAESGPPPPQPEVLATVKEERGETAAGAGVPGEATGRGAGGRRRKRPLQRGKPPYSYIALIAMAIAHAPERRLTLGGIYKFITERFPFYRDNPKKWQNSICHNLTLNDCFLKIPREAGRPGKGNYWALDPNAEDMFESGSFLRRRKRFKRSDLSTYPAYMHDAAAAAAAAAAAAAAAAIFPGAVPAARPPYPGAVYAGYAPPSLAAPPPVYYPAASPGPCRVFGLVPERPLSPELGPAPSGPGGSCAFASAGAPATTTGYQPAGCTGARPANPSAYAAAYAGPDGAYPQGAGSAIFAAA\n\n### Answer:", "output": "In congenital hypothyroidism; with absence of thyroid agenesis; loss of sequence-specific DNA binding; loss of transcriptional activity. The novel TTF-2 mutation in a female child resulted in syndromic congenital hypothyroidism (CH) in the absence of thyroid agenesis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00062", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nYIDZ_ECOLI\n\n### Answer:", "output": "Involved in anaerobic NO protection.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00063", "category": "Mutation", "instruction": "Describe the effect of mutation P0A8Y8-M68A on protein Proofreading thioesterase EntH.", "input": "### Wild-type:\nMIWKRHLTLDELNATSDNTMVAHLGIVYTRLGDDVLEAEMPVDTRTHQPFGLLHGGASAALAETLGSMAGFMMTRDGQCVVGTELNATHHRPVSEGKVRGVCQPLHLGRQNQSWEIVVFDEQGRRCCTCRLGTAVLG\n\n### Mutant:\nMIWKRHLTLDELNATSDNTMVAHLGIVYTRLGDDVLEAEMPVDTRTHQPFGLLHGGASAALAETLGSAAGFMMTRDGQCVVGTELNATHHRPVSEGKVRGVCQPLHLGRQNQSWEIVVFDEQGRRCCTCRLGTAVLG\n\n### Answer:", "output": "130-fold decrease in activity toward salicylyl-CoA. The mutation in the EntH protein alters its interaction with substrates, potentially affecting its ability to distinguish between normal and aberrant precursors in the enterobactin assembly line.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00064", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "PDCD1, TOX, HAVCR2, CD8A, LAG3", "output": "The expression pattern is consistent with exhausted CD8 T cell in a context associated with breast carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "lung", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["PDCD1", "TOX", "HAVCR2", "CD8A", "LAG3"]}}
{"id": "eval_00065", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDERL2_CAEEL\n\n### Answer:", "output": "Derlin-2, DER1-like protein 2, cDerlin-2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00066", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLPFTLQQLRILKAVATEKNFTKAAELLYLSQPSLSKQIKTLEKNLDILLVNRENNKISLTENGKIFLQYSERILALCEESCRALIDLKNGERGSLTVGASQTIGTYLMPRVLALFAQNYPQIDLKVQVNSTRIVAKNILNREIDIAVVGGEIPLDLKKNLTVEKFVEDEFSLIIPKSHPFANKKIVTKEDLYHLNFISLNSNSTIRKFIDNILIQNQIDTKQLKIIMQLNSIEAIKTAVSLGQGAAFVSSAAIEKEIELKTIEILKIENIRITRTLSIISNSESYKSKAFEFFSNELKKLKNEIEN\n\n### Sequence 2:\nLPFTLDQLRILKAIAKEGSFKKAANSLYVSQPAISLQIQNLERQLNVALFERGNKKATLTEAGSLLLRYGGRILALCEETCRALDDLQNLQGGTLIIGASQTTGTYLMPRLIGLFRQRYPQVAVQLQVHSTRLISWSVANGQVDLAIIGGEVPTELQDVLQVTSYAEDELALILPKSHPFSKLGDIQKEDLYRLRFIALDTQSTIRKVIDKVLSQHGIDSSRFKIEMELNSIEAIKNAVQSGLGAAFVSVSAIAKELELGIVHWAQIENVTIKRMLSIIVNPNRYKSKATETFSQEI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00067", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMALSTATKAATDALAANRAPTSVNAQEVHRWLQSFNWDFKNNRTKYATKYKMANETKEQFKLIAKEYARMEAVKDERQFGSLQDALTRLNAGVRVHPKWNETMKVVSNFLEVGEYNAIAATGMLWDSAQAAEQKNGYLAQVLDEIRHTHQCAYVNYYFAKNGQDPAGHNDARRTRTIGPLWKGMKRVFSDGFISGDAVECSLNLQLVGEACFTNPLIVAVTEWAAANGDEITPTVFLSIETDELRHMANGYQTVVSIANDPASAKYLNTDLNNAFWTQQKYFTPVLGMLFEYGSKFKVEPWVKTWDRWVYEDWGGIWIGRLGKYGVESPRSLKDAKQDAYWAHHDLYLLAYALWPTGFFRLALPDQEEMEWFEANYPGWYDHYGKIYEEWRARGCEDPSSGFIPLMWFIENNHPIYIDRVSQVPFCPSLAKGASTLRVHEYNGEMHTFSDQWGERMWLAEPERYECQNIFEQYEGRELSEVIAELHGLRSDGKTLIAQPHVRGDKLWTLDDIKRLNCVFKNPVKAFN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCHHHHHTTGGGGCCCCTTCSCSSCCSCCCCTTCCCSSCCCHHHHHHHHHHHHHHHHHHHHTHHHHTTTTTSSCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHSSCCTTTTTHHHHGGGSHHHHHHHHHTTHHHHSSCHHHHHHHHTTTHHHHTHHHHHHHHHHHHHHTTCCHHHHHHHHHHTTHHHHHHHHHHHHHTTTTSTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHSCSSCCSCHHHHHHHHHIIIIIIHHHHHGGGGTCCCCTTHHHHHHHHHHHHHHHHHHHHHTGGGSSSBCCCCCHHHHHHHHHHSTTHHHHHHHHHHHHHHTTTTCTTSCCCHHHHHHHTTCCCEECTTTCCEECTTTCSSCCCCEEEEETTEEEEESSHHHHHHHHHCGGGCCCCCHHHHHTTCBHHHHHHHTTCBCTTSSBBSCBSSCSSSCCCBHHHHHTTCCBCCCGGGGCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00068", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nERG2_CANAL\n\n### Answer:", "output": "C-8 sterol isomerase ERG2, Ergosterol biosynthetic protein 2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00069", "category": "Mutation", "instruction": "Describe the effect of mutation Q9LZJ3-N268A on protein Xyloglucan 6-xylosyltransferase 1.", "input": "### Wild-type:\nMIEKCIGAHRFRRLQRFMRQGKVTILCLVLTVIVLRGTIGAGKFGTPEKDIEEIREHFFYTRKRGEPHRVLVEVSSKTTSSEDGGNGGNSYETFDINKLFVDEGDEEKSRDRTNKPYSLGPKISDWDEQRRDWLKQNPSFPNFVAPNKPRVLLVTGSAPKPCENPVGDHYLLKSIKNKIDYCRIHGIEIFYNMALLDAEMAGFWAKLPLIRKLLLSHPEIEFLWWMDSDAMFTDMVFELPWERYKDYNLVMHGWNEMVYDQKNWIGLNTGSFLLRNSQWSLDLLDAWAPMGPKGKIREEA\n\n### Mutant:\nMIEKCIGAHRFRRLQRFMRQGKVTILCLVLTVIVLRGTIGAGKFGTPEKDIEEIREHFFYTRKRGEPHRVLVEVSSKTTSSEDGGNGGNSYETFDINKLFVDEGDEEKSRDRTNKPYSLGPKISDWDEQRRDWLKQNPSFPNFVAPNKPRVLLVTGSAPKPCENPVGDHYLLKSIKNKIDYCRIHGIEIFYNMALLDAEMAGFWAKLPLIRKLLLSHPEIEFLWWMDSDAMFTDMVFELPWERYKDYNLVMHGWNEMVYDQKNWIGLATGSFLLRNSQWSLDLLDAWAPMGPKGKIREEA\n\n### Answer:", "output": "No effect on activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00070", "category": "Mutation", "instruction": "Describe the effect of mutation Q7Z4S6-D352E on protein Kinesin-like protein KIF21A.", "input": "### Wild-type:\nMLGAPDESSVRVAVRIRPQLAKEKIEGCHICTSVTPGEPQVFLGKDKAFTFDYVFDIDSQQEQIYIQCIEKLIEGCFEGYNATVFAYGQTGAGKTYTMGTGFDVNIVEEELGIISRAVKHLFKSIEEKKHIAIKNGLPAPDFKVNAQFLELYNEEVLDLFDTTRDIDAKSKKSNIRIHEDSTGGIYTVGVTTRTVNTESEMMQCLKLGALSRTTASTQMNVQSSRSHAIFTIHVCQTRVCPQIDADNATDNKIISESAQMNEFETLTAKFHFVDLAGSERLKRTGATGERAKEGISINCG\n\n### Mutant:\nMLGAPDESSVRVAVRIRPQLAKEKIEGCHICTSVTPGEPQVFLGKDKAFTFDYVFDIDSQQEQIYIQCIEKLIEGCFEGYNATVFAYGQTGAGKTYTMGTGFDVNIVEEELGIISRAVKHLFKSIEEKKHIAIKNGLPAPDFKVNAQFLELYNEEVLDLFDTTRDIDAKSKKSNIRIHEDSTGGIYTVGVTTRTVNTESEMMQCLKLGALSRTTASTQMNVQSSRSHAIFTIHVCQTRVCPQIDADNATDNKIISESAQMNEFETLTAKFHFVDLAGSERLKRTGATGERAKEGISINCG\n\n### Answer:", "output": "In CFEOM1; de novo mutation. The mutation in the KIF21A gene leads to phenotypic characteristics of congenital fibrosis of the extraocular muscles (CFEOM) and Möbius syndrome.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00071", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSYT_STRP1\n\n### Answer:", "output": "Threonine--tRNA ligase, Threonyl-tRNA synthetase, ThrRS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00072", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGCSECCVRCLGGVPYASLIATILCFVGVALFCGCGHEALTGTEKLIELYFSNDFMDYALLVNVIQVFQYIIYGTASFSFLYGVLLLAEGFYTTSAVKALFGEFRTTVCGRCVSATFIFLTYALGVTWMGVFAFSALPVYIYYTMWSTCQMVKYVTENGTGFDDICVDARQYGILPWNASPGKICGLSLAAVCNTSEFELTYHLFIATFAGAAATVIALLTYMMSSTYNYAVLKFLSRDDCCTKF\n\n### Sequence 2:\nMAFSPNPLSLSVPDPAFESWLRDSGYLELLDHRTSAAAAAASSSASVSSSAAATSAASDDVVSSITGGFFASLLSRLVTVSSLLTINPFSKLSADDFSGDTTPWTTGFIGNCDSYSFPSSSQQARMRVHENIKRFARNYATLFIVFFACALYQMPLALVGLLGSLALWELFKYCSDKWKFDRHPSMRKLSIGIGQCATAVLLTFLNVQMALFSALAISYSVMILHAGFRKLTPSKKPTRGR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00073", "category": "Mutation", "instruction": "Describe the effect of mutation O43602-R192W on protein Neuronal migration protein doublecortin.", "input": "### Wild-type:\nMELDFGHFDERDKTSRNMRGSRMNGLPSPTHSAHCSFYRTRTLQALSNEKKAKKVRFYRNGDRYFKGIVYAVSSDRFRSFDALLADLTRSLSDNINLPQGVRYIYTIDGSRKIGSMDELEEGESYVCSSDNFFKKVEYTKNVNPNWSVNVKTSANMKAPQSLASSNSAQARENKDFVRPKLVTIIRSGVKPRKAVRVLLNKKTAHSFEQVLTDITEAIKLETGVVKKLYTLDGKQVTCLHDFFGDDDVFIACGPEKFRYAQDDFSLDENECRVMKGNPSATAGPKASPTPQKTSAKSPGP\n\n### Mutant:\nMELDFGHFDERDKTSRNMRGSRMNGLPSPTHSAHCSFYRTRTLQALSNEKKAKKVRFYRNGDRYFKGIVYAVSSDRFRSFDALLADLTRSLSDNINLPQGVRYIYTIDGSRKIGSMDELEEGESYVCSSDNFFKKVEYTKNVNPNWSVNVKTSANMKAPQSLASSNSAQARENKDFVRPKLVTIIRSGVKPWKAVRVLLNKKTAHSFEQVLTDITEAIKLETGVVKKLYTLDGKQVTCLHDFFGDDDVFIACGPEKFRYAQDDFSLDENECRVMKGNPSATAGPKASPTPQKTSAKSPGP\n\n### Answer:", "output": "In LISX1 and SBHX.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00074", "category": "Mutation", "instruction": "Describe the effect of mutation P60824-K101R on protein Cold-inducible RNA-binding protein.", "input": "### Wild-type:\nMASDEGKLFVGGLSFDTNEQALEQVFSKYGQISEVVVVKDRETQRSRGFGFVTFENIDDAKDAMMAMNGKSVDGRQIRVDQAGKSSDNRSRGYRGGSAGGKGFFRGGRSRGRGFSRGGGDRGYGGGRFESRSGGYGGSRDYYASRSQGGSYGYRSSGGSYRDSYDSYATHNE\n\n### Mutant:\nMASDEGKLFVGGLSFDTNEQALEQVFSKYGQISEVVVVKDRETQRSRGFGFVTFENIDDAKDAMMAMNGKSVDGRQIRVDQAGKSSDNRSRGYRGGSAGGRGFFRGGRSRGRGFSRGGGDRGYGGGRFESRSGGYGGSRDYYASRSQGGSYGYRSSGGSYRDSYDSYATHNE\n\n### Answer:", "output": "Enhances translational repression. The mutation in the CIRP protein leads to a functional change in its ability to migrate into stress granules (SGs). This mutation affects the demethylation of arginine residues in the RGG domain, which is necessary for CIRP to exit the nucleus and be recruited into SGs. Additionally, this mutation may also impact the up-regulation of mRNA translation by CIRP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00075", "category": "Mutation", "instruction": "Describe the effect of mutation P23890-E461R on protein Transcriptional activator CadC.", "input": "### Wild-type:\nMQQPVVRVGEWLVTPSINQISRNGRQLTLEPRLIDLLVFFAQHSGEVLSRDELIDNVWKRSIVTNHVVTQSISELRKSLKDNDEDSPVYIATVPKRGYKLMVPVIWYSEEEGEEIMLSSPPPIPEAVPATDSPSHSLNIQNTATPPEQSPVKSKRFTTFWVWFFFLLSLGICVALVAFSSLDTRLPMSKSRILLNPRDIDINMVNKSCNSWSSPYQLSYAIGVGDLVATSLNTFSTFMVHDKINYNIDEPSSSGKTLSIAFVNQRQYRAQQCFMSIKLVDNADGSTMLDKRYVITNGNQL\n\n### Mutant:\nMQQPVVRVGEWLVTPSINQISRNGRQLTLEPRLIDLLVFFAQHSGEVLSRDELIDNVWKRSIVTNHVVTQSISELRKSLKDNDEDSPVYIATVPKRGYKLMVPVIWYSEEEGEEIMLSSPPPIPEAVPATDSPSHSLNIQNTATPPEQSPVKSKRFTTFWVWFFFLLSLGICVALVAFSSLDTRLPMSKSRILLNPRDIDINMVNKSCNSWSSPYQLSYAIGVGDLVATSLNTFSTFMVHDKINYNIDEPSSSGKTLSIAFVNQRQYRAQQCFMSIKLVDNADGSTMLDKRYVITNGNQL\n\n### Answer:", "output": "Abolishes response to the external low pH signal. The mutation in the CadC protein is found to be crucial for pH detection. It is part of a cluster of negatively charged amino acids (D-198, D-200, E-461, E-468, and D-471) that form a negatively charged patch on the surface of the periplasmic domain of CadC. Alterations in this patch can affect the activation of the receptor upon lowering of the environmental pH.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00076", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nYGK3_SCHPO\n\n### Answer:", "output": "To yeast YGR017w.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00077", "category": "Mutation", "instruction": "Describe the effect of mutation P14871-G218D on protein Type II methyltransferase M.FokI.", "input": "### Wild-type:\nMRFIGSKVNLLDNIQEVIEENVKDDAHVFMDLFSGTGIVGENFKKDYQVLSNDSLYFSYILLKAKIENNSIPNFSELKKIGIKEPLHYLENEEFEISHEFFLTHNYSPYMGCERMYFTVENASRIDFIRLTLNRWKNESLINELEFAYLLAILIEAVPFISNISGTYGAYLKHWDKRALGKLKLRTLDIGNNHYANKTYNEDANSLIEKVYGDILYIGPPYNGRQYISNYHLLETIALYDYPEIYGKTGLRPYVESKSLYCQKKEVGNAFNHLIEKANFRHILVSYSSEGLLLEEEIESI\n\n### Mutant:\nMRFIGSKVNLLDNIQEVIEENVKDDAHVFMDLFSGTGIVGENFKKDYQVLSNDSLYFSYILLKAKIENNSIPNFSELKKIGIKEPLHYLENEEFEISHEFFLTHNYSPYMGCERMYFTVENASRIDFIRLTLNRWKNESLINELEFAYLLAILIEAVPFISNISGTYGAYLKHWDKRALGKLKLRTLDIGNNHYANKTYNEDANSLIEKVYGDILYIDPPYNGRQYISNYHLLETIALYDYPEIYGKTGLRPYVESKSLYCQKKEVGNAFNHLIEKANFRHILVSYSSEGLLLEEEIESI\n\n### Answer:", "output": "Loss of demethylation; when associated with A-548. The mutation in the MFokI gene resulted in preservation of the methylation of the strand carrying 5'-GGATG-3' in the target DNA.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00078", "category": "Mutation", "instruction": "Describe the effect of mutation Q84HF5-E372Q on protein Kynurenine 3-monooxygenase.", "input": "### Wild-type:\nMTATDNARQVTIIGAGLAGTLVARLLARNGWQVNLFERRPDPRIETGARGRSINLALAERGAHALRLAGLEREVLAEAVMMRGRMVHVPGTPPNLQPYGRDDSEVIWSINRDRLNRILLDGAEAAGASIHFNLGLDSVDFARQRLTLSNVSGERLEKRFHLLIGADGCNSAVRQAMASVVDLGEHLETQPHGYKELQITPEASAQFNLEPNALHIWPHGDYMCIALPNLDRSFTVTLFLHHQSPAAQPASPCFAQLVDGHAARRFFQRQFPDLSPMLDSLEQDFEHHPTGKLATLRLTTW\n\n### Mutant:\nMTATDNARQVTIIGAGLAGTLVARLLARNGWQVNLFERRPDPRIETGARGRSINLALAERGAHALRLAGLEREVLAEAVMMRGRMVHVPGTPPNLQPYGRDDSEVIWSINRDRLNRILLDGAEAAGASIHFNLGLDSVDFARQRLTLSNVSGERLEKRFHLLIGADGCNSAVRQAMASVVDLGEHLETQPHGYKELQITPEASAQFNLEPNALHIWPHGDYMCIALPNLDRSFTVTLFLHHQSPAAQPASPCFAQLVDGHAARRFFQRQFPDLSPMLDSLEQDFEHHPTGKLATLRLTTW\n\n### Answer:", "output": "Strongly decreases kynurenine 3-monooxygenase activity. The mutation in hKMO causes allosteric inhibition by Ro 61-8048.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00079", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPLSY_COXBN\n\n### Answer:", "output": "Glycerol-3-phosphate acyltransferase, Acyl-PO4 G3P acyltransferase, Acyl-phosphate--glycerol-3-phosphate acyltransferase, G3P acyltransferase, GPAT, Lysophosphatidic acid synthase, LPA synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00080", "category": "Cell", "instruction": "Based on the marker genes, describe the enriched biological pathway.", "input": "MYH6, MYH7, TNNI3, TNNT2", "output": "The marker profile is enriched for cardiac contraction and sarcomere organization, consistent with cardiomyocyte function.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_pathway", "top_genes": ["MYH6", "MYH7", "TNNI3", "TNNT2"]}}
{"id": "eval_00081", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRUVB_LACLA\n\n### Answer:", "output": "The RuvA-RuvB-RuvC complex processes Holliday junction (HJ) DNA during genetic recombination and DNA repair, while the RuvA-RuvB complex plays an important role in the rescue of blocked DNA replication forks via replication fork reversal (RFR).  RuvA specifically binds to HJ cruciform DNA, conferring on it an open structure.  The RuvB hexamer acts as an ATP-dependent pump, pulling dsDNA into and through the RuvAB complex.  RuvB forms 2 homohexamers on either side of HJ DNA bound by 1 or 2 RuvA tetramers; 4 subunits per hexamer contact DNA at a time.  Coordinated motions by a converter formed by DNA-disengaged RuvB subunits stimulates ATP hydrolysis and nucleotide exchange.  Immobilization of the converter enables RuvB to convert the ATP-contained energy into a lever motion, pulling 2 nucleotides of DNA out of the RuvA tetramer per ATP hydrolyzed, thus driving DNA branch migration.  The RuvB motors rotate together with the DNA substrate, which together with the progressing nucleotide cycle form the mechanistic basis for DNA recombination by continuous HJ branch migration.  Branch migration allows RuvC to scan DNA until it finds its consensus sequence, where it cleaves and resolves cruciform DNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00082", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMANAIIEKAKERMTQSHQSLAREFGGIRAGRANASLLDRVHVEYYGVETPLNQIASITIPEARVLLVTPFDKSSLKDIERALNASDLGITPANDGSVIRLVIPALTEETRRDLAKEVKKVGENAKVAVRNIRRDAMDEAKKQEKAKEITEDELKTLEKDIQKVTDDAVKHIDDMTANKEKELLEV\n\n### Sequence 2:\nMNNFSHWNEAGRPKMVDISEKEITTRTAIARSTITLSNEVYQAIQQGGIKKGDPTQVAQIAGIMGAKKTADIIPMCHPIMLQGTDLQFDYEKIDNGYALHIQATVKCNGKTGVEMEALTAVSIAALTFYDMCKAVDKTMVIKETYLVEKTGGKSGTFTHK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00083", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nAVLMVGPNFRVGKKIGCFGELRLGKNLYNNEHVAIKMEPMKSKAPQLHLEYRFYKLLGSHADNAPDGIPRIYHLGTCGGRYNAMVLELLGLSLEDLFNICARKFSLKTVLMIAKQLLHRIEYVHSRHLIYRDVKPENFLIGRTSTKREKIIHIIDFGLAKEYIDLDTNRHIPYREHKSLTGTARYMSINTHMGREQSRRDDLEALGHMFMYFLRGSLPWQGLKADTLKERYQKIGDTKRATPIEVLCDGHPEEFATYLRYVRRLDFFETPDYDFLRRLFQDLFDRKGYTDEGEFDWTGKT\n\n### Answer:", "output": "<SEQ_3Di>DFKDWDPFKTWDDFPDDQFTKTWIAGPVPRDIKIKTKDFLPDDDDQPVVQVVLCVLLDLDQPDADAQEWHWDDWGAIDDRMTMTIIHDAAAFQVVVCVVLVLAFDLLQLLLVVLQVLVVLVSSVVSQKDQQADDRRQWTFGDVVPVGNLGTHGDDRRPMDGQQDPPPRHGDDKDFPADADYDLQQFDLVSLRRTDDASLRVQSSSLSVSLSRHPRGFPLPPPDDDDPNVNSVSSSVVVVVDQLCRSQVPHDCLSVVSNVVSNPDDRRDDDDSVVNSVSSVVVNCVVPDDPPSDTPCVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00084", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTDRFKGDDDIPXXXXXXXXXXXXXDYYQQVQRGKQDKKISRKEAHKNAVIAAREGKLAELAENVSGDGKRAINYQILKNKGLTPKRNKDNRNSRVKKRKKYQKAQKKLKSVRAVYSGGQSGVYEGEKTGIKKGLTRSVKFKN\n\n### Answer:", "output": "<SEQ_3Di>DDDPDPDDDDDDPVVVVVVVVVVDPVVVVVVVVVVVVVVVVVVVVVVVVVVVVPVVVVVVVVPDDPDPDDDDDPCVPVVDPDDDDDDPPVVDVVVVVVVVVVVVVVVVCVVDPDDDDDDDDDDDDDDVDDDDPDPPDDDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00085", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nGLNA_SUIBO\n\n### Answer:", "output": "Cytoplasm.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00086", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nFCGR3A:low CDKN1C:low LST1:high\n</cell>", "output": "This cell is most consistent with CD16+ monocyte, a non-classical CD16+ monocyte. Key markers such as FCGR3A, LST1, CDKN1C support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "bone marrow", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["FCGR3A", "CDKN1C", "LST1"]}}
{"id": "eval_00087", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMASKNSINKPKIKINAHSHAKTLGRKRAARARKNPSTKSSTSRYAEVSGTATKPSDSTAIALYTGEAPKPTGVVTNTTLSNKRAKKLARNQKYIDQRKNEGLGKEHEMDVEVETEEQRQSRLEQIKKALWSVVQNQSEGLYKVAVDAEGTTLGVQAF\n\n### Sequence 2:\nMSLLKEFKAFASRGNVIDMAVGIIIGAAFGKIVSSFVADIIMPPIGIILGGVNFSDLSFVLLAAQGDAPAVVIAYGKFIQTVVDFTIIAFAIFMGLKAINSLKRKEEEAPKAPPAPTKDQELLSEIRDLLKAQQDK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00088", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGLVPRGSHMWKIVFARIDDRLIHGQVMTRWMKGFPEASIVIIDDELAVDEFMKNIYTMAAPPGVKVKVFGVDAALKEWSQKTSVEEKVFLLFKNIDTCKRVMDGGLPITTLNIGGVAKTPQRKGISQSVSLSEDEVKTLLELKTKYNVDVYLQMIPDSEKIHLTTVVEKYFPELK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCBSEEEEEEEEETTCSCTTHHHHHHTTCTTCEEEEECHHHHHCHHHHHHHHHTSCTTCCEEEECHHHHHHHHHSBCSSCEEEEEEESSHHHHHHHHHTTCCCSEEEEEEBCCCTTSEECSSSBEECHHHHHHHHHHHHHHCCEEEECSSTTSCCEEHHHHHHHHCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00089", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSFEKISSLDDIWVGCMETFETSDGTEVLIVNSEEHGVKAYQAMCPHQEILLSEGSYEGGVITCRAHLWTFNDGTGHGINPDDAALAEYPVEVKGDDIYVSTKGILPNKAHS\n\n### Answer:", "output": "<SEQ_3Di>DKDWWFFCVVAPAQAWDWTADPVRAIWIWHQHPVAGIWIWHQADLADGDGQNPAHGDPQWGFRPPPTFIAGNNFQQTPVVRVDGTWTWDWDHDPGTIITDCPPTDTHHDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00090", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "Cl[C@H]1[C@H](Cl)[C@@H](Cl)[C@H](Cl)[C@@H](Cl)[C@@H]1Cl", "output": "The molecule contains the following functional groups: halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "Cl[C@H]1[C@H](Cl)[C@@H](Cl)[C@H](Cl)[C@@H](Cl)[C@@H]1Cl", "compound_name": null, "pubchem_cid": null, "MW": 290.83, "LogP": 3.64, "TPSA": 0.0, "QED": 0.6, "functional_groups": ["halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": null, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["halide"]}}
{"id": "eval_00091", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMGRMHAPGKGISQSALPYRRSVPTWLKLGPDDVKEQIYKLAKKGLTPSQIGVILRDSHGVAQTRHVAGNKILRILKAKGLAPTIPEDLYFLIKKAVAIRKHLERNRKDKDSKFRLILVESRIHRLARYYKTRKVLPPVWKYESATASALVA\n\n### Sequence 2:\nMSQQIKLLVGLANPGPEYAKTRHNAGAWVVEELARVHNVTLKNEPKFFGLTGRIMVHGEDLRLLIPTTFMNLSGKAVAALAKFYQIKPEEIMVAHDELDLPPGIGKFKKGGGHGGHNGLKDIISKQGNNKEFYRLRLGIGHPGHKDKVAGYVLGKAPAKEQECIEAVVDESVRSLDILLKDGLPKAQNRLHTFKAE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00092", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nAYTNSFTRGVYYPDKVFRSSVLHSTQDLFLPFFSNVTWFHNPVLPFNDGVYFASTEKSNIIRGWIFGTTLDSKTQSLLIVNNATNVVIKVCNNCTFEYVSKNLREFVFKNYFKIYSKHTPINDLPQGFSALEPLVDLPIGINITRFQTLLAAYYVGYLQPRTFLLKYNENGTITDAVDCALDPLSETKCTLKSFTVEKGIYQTSNFRVQPTESIVRFPNKSTNLVKNKCVNFNFNGLTGTGVLTESNKKFLPFQQFGRDIADTTDAVRDPQTLEILDITPCSFGGVSVITPGTNTSNEVAVLYQDVNCTEVPVAIHADQLTPTWRVYSTGSNVFQTRAGCLIGAEHVNNSYECDIPIGAGICASYQTSQSIIAYTMSLGAENSVAYSNNSIAIPTNFTISVTTEILPVSMTKTSVDCTMYICGDSTECSNLLLQYGSFCTQLNRALTGIAVEQDKNTQEVFAQVKQIYKTPPIKDFGGFNFSQILPDPSKSKRSFIEDLLFNKVTKFNGLTVLPPLLTDEMIAQYTSALLAGTITSGWTFGAGAALQIPFAMQMAYRFNGIGVTQNVLYENQKLIANQFNSAIGKIQDSLSSTASALGKLQDVVNQNAQALNTLVKQLSSNFGAISSVLNDILSRLDPPEAEVQIDRLITGRLQSLQTYVTQQLIRAAEIRASANLAATKMSECVLGQSKRVDFCGKGYHLMSFPQSAPHGVVFLHVTYVPAQEKNFTTAPAICHDGKAHFPREGVFVSNGTHWFVTQRNFYEPQIITTDNTFVSGNCDVVIGIVNNTVYDP\n\n### Answer:", "output": "<SEQ_3Di>DKDFQAAEWAADPDDDFDAQAKDKDWDWTFDHRDIWDKALPDKFWQPFKKKKKDDDDVQQFFWKKWDDPPPDPFWIWIWGDDDQKIWIWRDVVTDDIDMDVKMKIWMWGPCIWIWIDIDNDPVNDHAYDYTYTDDRDPDPGDTTMMDTDGGIMIMDHIDGWMKIFGHHRVRGTHDMAGCQQAQLSVVCVVVSHPWDAFEKDFRDKAFADADEEFEADAVVGDDDDAPGFHFYHAHHDTDTFGWDADPDDDDPPDAFDADPVRHTFWGQHRHDRGIIGGHGGDIFIKMKGDHTVVVDPQIKIFGFPDALVCQCPRVPLVVGDPNHVQRNDDPQWDQDNRGIIGFAHEDADEEAADYDRDPQKHWDFADVTGIYIYGHDPDDDDDDPDDDWKDKAFAFWFKAKDKDWFFFAAWLKAFQQCLQQPNPHPVLVVLVVVPPCLNVVLRVLSVVLNVVRLVLVCLQQVQADDQDDADPDPPQLLADCQQAAFCPPPVRGGNVLVLQAVLDDPPSRIDTDDHPDDPVNSRRSQNRLVRLCSAANLVSVVPPGDGDRNLVLLQVLCVVLFHHSVVCVVVVNVRNVSNVVNVVVVSVVSNPDRCPVVVVVVSVVVVVVLVVVLVVLLQDQLPAPGSDLVVLVVPDDPVVNSVSSVSSSVSNSVSSVVSSVVSVVVSVVSSVSSVLSSVCCSDQRSHADPDACSQHHARWRHWGWHDHDSIIMITTTHMGRDDIDMWTWFQWEAAPQWTKHFPAFTWTDPPPDIFTAHRHHDDGDAAACVGIPTDGGDPPDDRYDYDHHDYD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00093", "category": "Mutation", "instruction": "Describe the effect of mutation Q9C9Z9-S371L on protein Glycosyltransferase-like KOBITO 1.", "input": "### Wild-type:\nMKSTHHHRAPLISASSSSSSSSQNHSFVSRLLLLLTLLPVSLACLAFILQWRGGGLADPASASVRSSTSVPGGSDLNHEVFPGMETVSSVSPKSHQSSSDCSNLARSSSPSFPYYADWKFGVDTSLKPKICITTSTSAGLDQILPWMFYHKVLGVSTFFLFVEGKAATPSISKVLESIPGVKVIYRTKELEEKQAKSRIWNETWLSSFFYKPCNYELFVKQSLNMEMAIVMARDAGMDWILHLDTDELIYPAGAREYSLRRLLLDVPPNVDMVIFPNYESSVERDDIKDPFTEVSMFKKN\n\n### Mutant:\nMKSTHHHRAPLISASSSSSSSSQNHSFVSRLLLLLTLLPVSLACLAFILQWRGGGLADPASASVRSSTSVPGGSDLNHEVFPGMETVSSVSPKSHQSSSDCSNLARSSSPSFPYYADWKFGVDTSLKPKICITTSTSAGLDQILPWMFYHKVLGVSTFFLFVEGKAATPSISKVLESIPGVKVIYRTKELEEKQAKSRIWNETWLSSFFYKPCNYELFVKQSLNMEMAIVMARDAGMDWILHLDTDELIYPAGAREYSLRRLLLDVPPNVDMVIFPNYESSVERDDIKDPFTEVSMFKKN\n\n### Answer:", "output": "In kob1-3; increased plasmodesmatal permeability, dwarf plants and stomata clusters in the erl1-2 erl2-1 background. The kob1-3 mutation leads to the formation of stomata clusters in the erl1 erl2 background but not in the wild type. Cell-to-cell mobility assays demonstrated an increase in intercellular protein trafficking in kob1-3, including increased diffusion of SPEECHLESS, suggesting that the formation of stomata clusters is due to an escape of cell fate-specifying factors from stomatal lineage cells. While plasmodesmatal permeability is increased in kob1-3, we did not detect drastic changes in callose accumulation at the neck regions of the plasmodesmata. Previously, KOBITO1 has been proposed to function in cellulose biosynthesis. Our data demonstrate that disruption of cellulose biosynthesis in the erl1 erl2 background does not lead to the formation of stomata clusters, indicating that cellulose biosynthesis is not a major determining factor for regulating plasmodesmatal permeability. Analysis of KOBITO1 structure suggests that it is a glycosyltransferase-like protein. KOBITO1 might be involved in a carbohydrate metabolic pathway that is essential for both cellulose biosynthesis and the regulation of plasmodesmatal permeability.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00094", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "O=C([O-])CCCO", "output": "The SMILES represents an unnamed molecule with MW=103.1, LogP=-1.49, TPSA=60.36. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "ClinTox", "canonical_smiles": "O=C([O-])CCCO", "compound_name": null, "pubchem_cid": null, "MW": 103.1, "LogP": -1.49, "TPSA": 60.36, "QED": 0.47, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"FDA_APPROVED": 1, "CT_TOX": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00095", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nAPOB:medium VIL1:low ALPI:medium\n</cell>", "output": "This cell is most consistent with enterocyte, an intestinal enterocyte. Key markers such as ALPI, APOB, VIL1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "enterocyte", "tissue": "pancreas", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["APOB", "VIL1", "ALPI"]}}
{"id": "eval_00096", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAAPAVSGVSQQVRYFGTSVVRPFAKLVRPPVQVYGVEGRYATALYSAASKQKKLEQVEKELLRVAQILKEPKVAASVLNPYVKHSIKVKSLSDIIAKERFSPLTTNLINLLAENGRLSNTQGVVSAFSTMMSVHRGEVPCTVTTASPLEEATLSELKTVLKSFLSQGQILKLEVKTDPSIMGGMIVRIGEKYVDMSAKTKIQKLSKAMREVI\n\n### Sequence 2:\nMAAPAVSGLSRQVRYFSTSVVRPFAKLVRPPVQVYGIEGRYATALYSAASKQKKLEQVEKELLRVAQILKEPKVAASVLNPYVKHSVKVKSLSDIIAKERFSPLTTNLINLLAENGRLSNTQGVVSAFSTMMSVHRGEIPCTVTTASPLEETTLSELKTVLKSFLSQGQILKLEVKTDPSIMGGMIVRIGEKYVDMSAKTKIQKLSKAMREVI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00097", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSDIPEHWEEDASWGPHRLAVLVPFRERFEELLVFVPHMRRFLSRKKIRHHIYVLNQVDHFRFNRAALINVGFLESSNSTDYIAMHDVDLLPLNEELDYGFPEAGPFHVASPELHPLYHYKTYVGGILLLSKQHYRLCNGMSNRFWGWGREDDEFYRRIKGAGLQLFRPSGITTGYKTFRHLHDPAWRKRDQKRIAAQKQEQFKVDREGGLNTVKYHVASRTALSVGGAPCTVLNIMLDCDKTATPWCTFS\n\n### Answer:", "output": "<SEQ_2D>CCCCCGGGSCCGGGCSCEEEEEEEESSCHHHHHHHHHHHHHHHHHTTCEEEEEEEEECSSSCCCHHHHHHHHHHHSCTTCCEECCCCTTEEECCTTSCCSCCTTSCEESSCTTTSSSCCCTTCCCSCEEEEHHHHHHTTCCCSTTTTSSCHHHHHHHHHHHTTCCCBCCCSCCCCTTSEEECCCCCCCCCCCCCCCCCCCCCCCCCCCCSTTTCCEEEEEEEEEEETTEEEEEEEEEECCCTTTCGGGSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00098", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMZT1_CHAGB\n\n### Answer:", "output": "Required for gamma-tubulin complex recruitment to the microtubule organizing center (MTOC).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00099", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CNC(=O)c1ccc(C)c(-n2c(C)cc(OCc3ccc(F)cc3F)c(Br)c2=O)c1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, ether, halide, pyridine. Its descriptor profile: MW=477.31, LogP=4.43, TPSA=60.33, HBD=1, HBA=3, QED=0.59.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "CNC(=O)c1ccc(C)c(-n2c(C)cc(OCc3ccc(F)cc3F)c(Br)c2=O)c1", "compound_name": null, "pubchem_cid": null, "MW": 477.31, "LogP": 4.43, "TPSA": 60.33, "QED": 0.59, "functional_groups": ["aromatic ring", "amide", "ether", "halide", "pyridine"], "moleculenet_labels": {"label": 2.8}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00100", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSKLGVLLTICLLLFPLTALPLDGDQPADQRAERTQAEKHSLPDPRMGCCPFPCKTSCTTLCCG\n\n### Sequence 2:\nMPPHGHHHHGHHGHHEHVTYTTTPIVQPMVQPMPVYGQPPMMVQPQVVMGAPVVYPPTNVVIETSHHHGHHGHHGHHGHHGHFF\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00101", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSREGFQIPTNLDAAAAGTSQARTATLKYICAECSSKLSLSRTDAVRCKDCGHRILLKARTKRLVQFEAR\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCEESSSCCBCCCCSSSCCCCSSSCCSCEECCCCSSCEEEECC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00102", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPSB1_SULIN\n\n### Answer:", "output": "Belongs to the peptidase T1B family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00103", "category": "Mutation", "instruction": "Describe the effect of mutation P40967-F160W on protein Melanocyte protein PMEL.", "input": "### Wild-type:\nMDLVLKRCLLHLAVIGALLAVGATKVPRNQDWLGVSRQLRTKAWNRQLYPEWTEAQRLDCWRGGQVSLKVSNDGPTLIGANASFSIALNFPGSQKVLPDGQVIWVNNTIINGSQVWGGQPVYPQETDDACIFPDGGPCPSGSWSQKRSFVYVWKTWGQYFQVLGGPVSGLSIGTGRAMLGTHTMEVTVYHRRGSRSYVPLAHSSSAFTITDQVPFSVSVSQLRALDGGNKHFLRNQPLTFALQLHDPSGYLAEADLSYTWDFGDSSGTLISRALVVTHTYLEPGPVTAQVVLQAAIPLTS\n\n### Mutant:\nMDLVLKRCLLHLAVIGALLAVGATKVPRNQDWLGVSRQLRTKAWNRQLYPEWTEAQRLDCWRGGQVSLKVSNDGPTLIGANASFSIALNFPGSQKVLPDGQVIWVNNTIINGSQVWGGQPVYPQETDDACIFPDGGPCPSGSWSQKRSFVYVWKTWGQYWQVLGGPVSGLSIGTGRAMLGTHTMEVTVYHRRGSRSYVPLAHSSSAFTITDQVPFSVSVSQLRALDGGNKHFLRNQPLTFALQLHDPSGYLAEADLSYTWDFGDSSGTLISRALVVTHTYLEPGPVTAQVVLQAAIPLTS\n\n### Answer:", "output": "Loss-of-function. Enables fibril formation. Exerts dominant positive effect; when associated with A-211.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00104", "category": "Mutation", "instruction": "Describe the effect of mutation P11884-R86Q on protein Aldehyde dehydrogenase, mitochondrial.", "input": "### Wild-type:\nMLRAALSTARRGPRLSRLLSAAATSAVPAPNQQPEVFCNQIFINNEWHDAVSKKTFPTVNPSTGEVICQVAEGNKEDVDKAVKAARAAFQLGSPWRRMDASDRGRLLYRLADLIERDRTYLAALETLDNGKPYVISYLVDLDMVLKCLRYYAGWADKYHGKTIPIDGDFFSYTRHEPVGVCGQIIPWNFPLLMQAWKLGPALATGNVVVMKVAEQTPLTALYVANLIKEAGFPPGVVNIVPGFGPTAGAAIASHEDVDKVAFTGSTEVGHLIQVAAGSSNLKRVTLELGGKSPNIIMSDA\n\n### Mutant:\nMLRAALSTARRGPRLSRLLSAAATSAVPAPNQQPEVFCNQIFINNEWHDAVSKKTFPTVNPSTGEVICQVAEGNKEDVDKAVKAAQAAFQLGSPWRRMDASDRGRLLYRLADLIERDRTYLAALETLDNGKPYVISYLVDLDMVLKCLRYYAGWADKYHGKTIPIDGDFFSYTRHEPVGVCGQIIPWNFPLLMQAWKLGPALATGNVVVMKVAEQTPLTALYVANLIKEAGFPPGVVNIVPGFGPTAGAAIASHEDVDKVAFTGSTEVGHLIQVAAGSSNLKRVTLELGGKSPNIIMSDA\n\n### Answer:", "output": "In allele Aldh2*2 and allele Aldh2*3; in strains UChA and UChB. This mutation is found in 94% of the UChA (low drinker) rats and is associated with an increase in ethanol consumption and an increase in the relative K(m) for NAD+ of ALDH2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00105", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "PRF1, GZMB, CD8B, CD8A, CD3D, NKG7, GZMA", "output": "The expression pattern is consistent with cytotoxic T cell in a context associated with COVID-19, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "spleen", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["PRF1", "GZMB", "CD8B", "CD8A", "CD3D", "NKG7", "GZMA"]}}
{"id": "eval_00106", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMIQPQSYLNVADNSGARKLMCIRVLGSSNRKYAHIGDMVIAVVKETVPNMPLKKSEIVRAVIVRTRKGLKRDNGMVLRFDDNAAVVINQEGNPRGTRVFGPVARELRDLNFTKIVSLAPEVL\n\n### Sequence 2:\nMEHVNEILATVGRILHETTTANTNVANKSTERLGAYIGAGITMVGGATVGLGQGYIFGKAVEAVARNPEVEKQVFKLIFIGSAISESSSIYSLLIAFILIFVSGA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00107", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nLCL3_KLULA\n\n### Answer:", "output": "Belongs to the LCL3 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00108", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRSMG_SALPA\n\n### Answer:", "output": "Specifically methylates the N7 position of guanine in position 527 of 16S rRNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00109", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "O=C(CCCN1CCC2(CC1)CC(=O)N(N1CCOCC1)C2=O)c1ccc(F)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, ether, ketone, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "O=C(CCCN1CCC2(CC1)CC(=O)N(N1CCOCC1)C2=O)c1ccc(F)cc1", "compound_name": null, "pubchem_cid": null, "MW": 417.48, "LogP": 1.88, "TPSA": 70.16, "QED": 0.52, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "ketone", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "ether", "ketone", "halide"]}}
{"id": "eval_00110", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: N#CC1(C2SCCS2=O)C2=C(C(=O)c3ccccc3C2=O)N2CCCC21\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "N#CC1(C2SCCS2=O)C2=C(C(=O)c3ccccc3C2=O)N2CCCC21", "compound_name": null, "pubchem_cid": null, "MW": 384.48, "LogP": 2.13, "TPSA": 78.24, "QED": 0.74, "functional_groups": ["aromatic ring", "amine", "ketone", "nitrile"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00111", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y6K1-W330R on protein DNA.", "input": "### Wild-type:\nMPAMPSSGPGDTSSSAAEREEDRKDGEEQEEPRGKEERQEPSTTARKVGRPGRKRKHPPVESGDTPKDPAVISKSPSMAQDSGASELLPNGDLEKRSEPQPEEGSPAGGQKGGAPAEGEGAAETLPEASRAVENGCCTPKEGRGAPAEAGKEQKETNIESMKMEGSRGRLRGGLGWESSLRQRPMPRLTFQAGDPYYISKRKRDEWLARWKREAEKKAKVIAGMNAVEENQGPGESQKVEEASPPAVQQPTDPASPTVATTPEPVGSDAGDKNATKAGDDEPEYEDGRGFGIGELVWGKL\n\n### Mutant:\nMPAMPSSGPGDTSSSAAEREEDRKDGEEQEEPRGKEERQEPSTTARKVGRPGRKRKHPPVESGDTPKDPAVISKSPSMAQDSGASELLPNGDLEKRSEPQPEEGSPAGGQKGGAPAEGEGAAETLPEASRAVENGCCTPKEGRGAPAEAGKEQKETNIESMKMEGSRGRLRGGLGWESSLRQRPMPRLTFQAGDPYYISKRKRDEWLARWKREAEKKAKVIAGMNAVEENQGPGESQKVEEASPPAVQQPTDPASPTVATTPEPVGSDAGDKNATKAGDDEPEYEDGRGFGIGELVWGKL\n\n### Answer:", "output": "In HESJAS; no effect on protein expression; changed DNA methylation; results in aberrant expression of genes involved in developmental processes. Substitutions in the PWWP domain abrogate binding to the histone modifications H3K36me2 and H3K36me3, and alter DNA methylation in patient cells. Polycomb-associated DNA methylation valleys, hypomethylated domains encompassing developmental genes, become methylated with concomitant depletion of H3K27me3 and H3K4me3 bivalent marks.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00112", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMPS2_YEASZ\n\n### Answer:", "output": "Component of the spindle pole body (SPB) required for insertion of the nascent SPB into the nuclear envelope and for the proper execution of spindle pole body (SPB) duplication.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00113", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSHMEAVQNRIVEAAERVPGVRGVIHLRARYVGQDIWADMIIGVDPENTVEQAHEICEAVQAAVCGKIRRIESLHVSADAREIGDTTKPSFSDQPLSFDEVMLSKVDN\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHHHHHTSTTCCCEEEEEEEEETTEEEEEEEEEECTTSCHHHHHHHHHHHHHHHHHHCTTEEEEEEEEEECCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00114", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nENOPH_DROGR\n\n### Answer:", "output": "Belongs to the HAD-like hydrolase superfamily. MasA/MtnC family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00115", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTRHO_CAUVC\n\n### Answer:", "output": "Catalyzes oxygen-dependent 5-hydroxyuridine (ho5U) modification at position 34 in tRNAs.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00116", "category": "Mutation", "instruction": "Describe the effect of mutation Q3ED78-C365S on protein Zinc finger CCCH domain-containing protein 7.", "input": "### Wild-type:\nMRPFKQSKFSLVWTQNDPQPRMPIAHMRNQNIVPQLVPWKRVTYWRRLMNSVSAFRNGSSLNISRKLSMMRKRHTIYTRSTNGYSLRKSKVLSVGGSHLKWSKSIERDSRKANEEATLAVAAYSKKESEKQSGQNNTSTASRNHLARERVFRFGSLRYKMDSSRRTLQRISDVDSPCSGPSENGKGVKRPFIPKRLVIGNEEYVRFGNGNQLVRDPKKRTRVLANEKVRWSLHNARLRLAKKKKYCQFFTRFGKCNKDDGKCPYVHDPSKIAVCTKFLNGLCANANCKLTHKVIPERMPD\n\n### Mutant:\nMRPFKQSKFSLVWTQNDPQPRMPIAHMRNQNIVPQLVPWKRVTYWRRLMNSVSAFRNGSSLNISRKLSMMRKRHTIYTRSTNGYSLRKSKVLSVGGSHLKWSKSIERDSRKANEEATLAVAAYSKKESEKQSGQNNTSTASRNHLARERVFRFGSLRYKMDSSRRTLQRISDVDSPCSGPSENGKGVKRPFIPKRLVIGNEEYVRFGNGNQLVRDPKKRTRVLANEKVRWSLHNARLRLAKKKKYCQFFTRFGKCNKDDGKCPYVHDPSKIAVCTKFLNGLCANANCKLTHKVIPERMPD\n\n### Answer:", "output": "Loss of RNA binding and reduction of ribonuclease activity; when associated with K-366 and H-369.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00117", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nZMY10_MOUSE\n\n### Answer:", "output": "Zinc finger MYND domain-containing protein 10, Protein BLu.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00118", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nTNNT2:medium MYH7:high MYH6:low\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as TNNT2, MYH7, MYH6 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "ovary", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["TNNT2", "MYH7", "MYH6"]}}
{"id": "eval_00119", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD79A, CD79B, MS4A1, IGHM, TCL1A, IGHD", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as CD79A, IGHD, CD79B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "breast", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD79A", "CD79B", "MS4A1", "IGHM", "TCL1A", "IGHD"]}}
{"id": "eval_00120", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRS20_STRS2\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bS20 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00121", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMVFEGHLVGTGLKVGVVVGRFNEFITSKLLGGALDGLKRHGVEENDIDVAWVPGAFEIPLIAKKMANSGKYDAVITLGTVIRGATTHYDYVCNEVAKGVASLSLQTDIPVIFGVLTTETIEQAIERAGTKAGNKGYESAVAAIEMAHLSKHWA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCTTCCEEEEEESTTHHHHHHHGGGIIIIIGGGTCCGGGEEEEEESSGGGHHHHHHHHHTTTCCSEEEEEEEEECCSSTHHHHHHHHHHHHHHHTSSSSCSCEEEEEEEESCHHHHHHTEEETTEEHHHHHHHHHHHHHTTTTSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00122", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nTTR:high IRX2:medium GCG:high\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as IRX2, TTR, GCG support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "tumor", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["TTR", "IRX2", "GCG"]}}
{"id": "eval_00123", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "KRT8, KRT18, CDH1, EPCAM", "output": "This profile is most consistent with epithelial cell, which is commonly found in brain under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "epithelial cell", "tissue": "brain", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["KRT8", "KRT18", "CDH1", "EPCAM"]}}
{"id": "eval_00124", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nKLF1 HBB GATA1 HBA1\n</cell>", "output": "This cell is most consistent with erythroid progenitor, an erythroid progenitor cell. Key markers such as KLF1, GATA1, HBA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "brain", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["KLF1", "HBB", "GATA1", "HBA1"]}}
{"id": "eval_00125", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMIQISDTAKSHFLKLIQREGVPGMGVRLSAVDPGTPRADARLEFADPSELVGDEWLIDCGDFTLYVASASVAWLDGAEIDYVTQATGSQQLIIKAPKIKGQELSQVASLVERVCWVVENEINPQLASHGGRVEVQEVSAEGVVLLRFGGGCHGCGMADVTLKQGVEKTLMERVHGVIAVRDATDHSTGAAPYISRDFSP\n\n### Sequence 2:\nMIQISDRAKAHFLRLIQREGVPGMGVRLSAVDPGTARADARLEFAEPSELVGDEWLIDCGDFTLYVASASVAWLDSAEIDYVTQATGSQQLIIKAPKIKGQELSQVASLVERVCWVVENEINPQLASHGGRVEVQEVSAEGVVLLRFGGGCHGCGMVDVTLKQGVEKTLMERVHGVIAVRDATDHSTGAAPYISRNFA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00126", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nIRX2:low GCG:medium TTR:high\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as GCG, IRX2, TTR support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "heart", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["IRX2", "GCG", "TTR"]}}
{"id": "eval_00127", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTRMFO_STRSV\n\n### Answer:", "output": "Methylenetetrahydrofolate--tRNA-(uracil-5-)-methyltransferase TrmFO, Folate-dependent tRNA (uracil-5-)-methyltransferase, Folate-dependent tRNA(M-5-U54)-methyltransferase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00128", "category": "Mutation", "instruction": "Describe the effect of mutation P09981-H56N on protein Alpha-like toxin BeM9.", "input": "### Wild-type:\nARDAYIAKPHNCVYECYNPKGSYCNDLCTENGAESGYCQILGKYGNACWCIQLPDHVPIRIPGKCH\n\n### Mutant:\nARDAYIAKPHNCVYECYNPKGSYCNDLCTENGAESGYCQILGKYGNACWCIQLPDNVPIRIPGKCH\n\n### Answer:", "output": "Important increase in activity on rNav1.2/SCN2A (enhances peak current and promotes inactivation), significant increase or gain in activity on rNav1.4/SCN4A, hNav1.5/SCN5A, and BgNav1, and no change in activity on mNav1.6/SCN8A.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00129", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKFLVNVALVFMVVYISYIYAAPEPEPAPEPEAEADAEADPEAGIGAVLKVLTTGLPALISWIKRKRQQG\n\n### Sequence 2:\nMSILSALTSISNPMKSSKSSVANGGGRLSMGSNSVACGSCGGGSSSSGTINNADGSKTTYYTYTSPIYTYNYSYSYSSSGSSSSCGCH\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00130", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nUBXN8_HUMAN\n\n### Answer:", "output": "Endoplasmic reticulum membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00131", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "COc1cc(CCc2cc(Nc3ccnc(NCc4ccccn4)n3)n[nH]2)cc(OC)c1", "output": "This molecule is medium-sized and polar, with lipophilic character. It contains aromatic ring, ether, pyridine. Its descriptor profile: MW=431.5, LogP=3.75, TPSA=109.87, HBD=3, HBA=8, QED=0.35.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "COc1cc(CCc2cc(Nc3ccnc(NCc4ccccn4)n3)n[nH]2)cc(OC)c1", "compound_name": null, "pubchem_cid": null, "MW": 431.5, "LogP": 3.75, "TPSA": 109.87, "QED": 0.35, "functional_groups": ["aromatic ring", "ether", "pyridine"], "moleculenet_labels": {"label": 3.5}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00132", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTVP18_SCHPO\n\n### Answer:", "output": "Golgi apparatus membrane protein tvp18.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00133", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "PRF1, GZMB, CD8A", "output": "This profile is most consistent with cytotoxic T cell, which is commonly found in lung under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "lung", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["PRF1", "GZMB", "CD8A"]}}
{"id": "eval_00134", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "Cc1nc(C2CCCCC2)c(-c2ccc(S(N)(=O)=O)c(F)c2)o1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amine, sulfonamide, halide. Its descriptor profile: MW=338.4, LogP=3.48, TPSA=86.19, HBD=1, HBA=4, QED=0.93.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "Cc1nc(C2CCCCC2)c(-c2ccc(S(N)(=O)=O)c(F)c2)o1", "compound_name": null, "pubchem_cid": null, "MW": 338.4, "LogP": 3.48, "TPSA": 86.19, "QED": 0.93, "functional_groups": ["aromatic ring", "amine", "sulfonamide", "halide"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": null, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00135", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGRPE_BACC2\n\n### Answer:", "output": "Protein GrpE, HSP-70 cofactor.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00136", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMEMVNKIACFVLLCMVVVAPHAEALTCGQVTSTLAPCLPYLMNRGPLGGCCGGVKGLLGQAQTTVDRQTACTCLKSAASSFTGLDLGKAASLPSTCSVNIPYKISPSTDCSKVQ\n\n### Sequence 2:\nMAMIKMSPEEIRAKSQSYGQGSDQIRQILSDLTRAQGEIAANWEGQAFSRFEEQFQQLSPKVEKFAQLLEEIKQQLNSTADAVQEQDQQLSNNFGLQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00137", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "IL7R, S100A4, CD4, CD3D, CD45RO", "output": "This profile is most consistent with memory CD4+ T cell, which is commonly found in tumor under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "tumor", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["IL7R", "S100A4", "CD4", "CD3D", "CD45RO"]}}
{"id": "eval_00138", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPHM3_PYRSX\n\n### Answer:", "output": "MFS-type transporter; part of the gene cluster that mediates the biosynthesis of the trans-fused decalin-containing tetramic acid phomasetin.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00139", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDHSO_SHEEP\n\n### Answer:", "output": "Belongs to the zinc-containing alcohol dehydrogenase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00140", "category": "Mutation", "instruction": "Describe the effect of mutation P43803-K889A on protein Protein translocase subunit SecA.", "input": "### Wild-type:\nMSILTRIFGSRNERVLRKLKKQVVKINKMEPAFEALSDDELKAKTQEFRDRLSGGETLQQILPEAFATVREASKRVLGMRHFDVQLIGGMVLTNRCIAEMRTGEGKTLTATLPCYLIALEGKGVHVVTVNDYLARRDAETNRPLFEFLGMSVGVNIPGLSPEEKRAAYAADITYATNSELGFDYLRDNLAHSKEERFQRTLGYALVDEVDSILIDEARTPLIISGQAENSSELYIAVNKLIPSLIKQEKEDTEEYQGEGDFTLDLKSKQAHLTERGQEKVEDWLIAQGLMPEGDSLYSPS\n\n### Mutant:\nMSILTRIFGSRNERVLRKLKKQVVKINKMEPAFEALSDDELKAKTQEFRDRLSGGETLQQILPEAFATVREASKRVLGMRHFDVQLIGGMVLTNRCIAEMRTGEGKTLTATLPCYLIALEGKGVHVVTVNDYLARRDAETNRPLFEFLGMSVGVNIPGLSPEEKRAAYAADITYATNSELGFDYLRDNLAHSKEERFQRTLGYALVDEVDSILIDEARTPLIISGQAENSSELYIAVNKLIPSLIKQEKEDTEEYQGEGDFTLDLKSKQAHLTERGQEKVEDWLIAQGLMPEGDSLYSPS\n\n### Answer:", "output": "Binds zinc, unable to bind secB. The mutation in SecA may lead to functional changes in the high-affinity binding with SecB.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00141", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nARLY_VIBVU\n\n### Answer:", "output": "In the N-terminal section; belongs to the lyase 1 family. Argininosuccinate lyase subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00142", "category": "Mutation", "instruction": "Describe the effect of mutation Q9H0E2-V12A on protein Toll-interacting protein.", "input": "### Wild-type:\nMATTVSTQRGPVYIGELPQDFLRITPTQQQRQVQLDAQAAQQLQYGGAVGTVGRLNITVVQAKLAKNYGMTRMDPYCRLRLGYAVYETPTAHNGAKNPRWNKVIHCTVPPGVDSFYLEIFDERAFSMDDRIAWTHITIPESLRQGKVEDKWYSLSGRQGDDKEGMINLVMSYALLPAAMVMPPQPVVLMPTVYQQGVGYVPITGMPAVCSPGMVPVALPPAAVNAQPRCSEEDLKAIQDMFPNMDQEVIRSVLEAQRGNKDAAINSLLQMGEEP\n\n### Mutant:\nMATTVSTQRGPAYIGELPQDFLRITPTQQQRQVQLDAQAAQQLQYGGAVGTVGRLNITVVQAKLAKNYGMTRMDPYCRLRLGYAVYETPTAHNGAKNPRWNKVIHCTVPPGVDSFYLEIFDERAFSMDDRIAWTHITIPESLRQGKVEDKWYSLSGRQGDDKEGMINLVMSYALLPAAMVMPPQPVVLMPTVYQQGVGYVPITGMPAVCSPGMVPVALPPAAVNAQPRCSEEDLKAIQDMFPNMDQEVIRSVLEAQRGNKDAAINSLLQMGEEP\n\n### Answer:", "output": "Reduced interaction with TOM1; when associated with Ala-9.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00143", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nENO_HYPNA\n\n### Answer:", "output": "Belongs to the enolase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00144", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nYPKTGVATSIVEKIERAEFNTAGRKPTVLLRIADFIAAMNGMDAKQDMQALWDAEIAIMNGRAQTTIISYITKYRNAIREAFGDDHPMLKIATGDAAMYDEAARVKMEKIANKHGALITFENYRQVLKICEDCLKSSDPLMIGIGLIGMTGRRPYEVFTQAEFSPAPYGKGVSKWSILFNGQAKTKQGEGTKFGITYEIPVLTRSETVLAAYKRLRESGQGKLWHGMSIDDFSSETRLLLRDTVFNLFEDVWPKEELPKPYGLRHLYAEVAYHNFAPPHVTKNSYFAAILGHNNNDLETSLSYMTYTLPEDRDNALARL\n\n### Answer:", "output": "<SEQ_3Di>DAQWWALVVLVVVQVVVVDDPDDDHPVLSVLLRVLNVVSRPDDDVVVLLVSLVVVVVVCPPDDLVVVQVSLVVNLVSCCRIPNDPDPSNVRSDDDPVSVVVVVVVVVVVLVVLVVPFFALELVVVVLVCLLVLLVDLDLLSVLLSCCQFQLAASLQLQAPKDWAFPADPPHTDFFKIKIADAPPFDDDPLACHRNIDMAGTPGTPVSNVVSSVSSCPDPSSVVSYPDDPVVCCVVPPVVSQVVQQVSCVPSGRVVDGGGSSVSSLVNLVVCCVPPPDDVADSQRSSCVSNRHGVVPSPVSSVSRHYDYPVVVVVVVVDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00145", "category": "Mutation", "instruction": "Describe the effect of mutation O95831-W196A on protein Apoptosis-inducing factor 1, mitochondrial.", "input": "### Wild-type:\nMFRCGGLAAGALKQKLVPLVRTVCVRSPRQRNRLPGNLFQRWHVPLELQMTRQMASSGASGGKIDNSVLVLIVGLSTVGAGAYAYKTMKEDEKRYNERISGLGLTPEQKQKKAALSASEGEEVPQDKAPSHVPFLLIGGGTAAFAAARSIRARDPGARVLIVSEDPELPYMRPPLSKELWFSDDPNVTKTLRFKQWNGKERSIYFQPPSFYVSAQDLPHIENGGVAVLTGKKVVQLDVRDNMVKLNDGSQITYEKCLIATGGTPRSLSAIDRAGAEVKSRTTLFRKIGDFRSLEKISREV\n\n### Mutant:\nMFRCGGLAAGALKQKLVPLVRTVCVRSPRQRNRLPGNLFQRWHVPLELQMTRQMASSGASGGKIDNSVLVLIVGLSTVGAGAYAYKTMKEDEKRYNERISGLGLTPEQKQKKAALSASEGEEVPQDKAPSHVPFLLIGGGTAAFAAARSIRARDPGARVLIVSEDPELPYMRPPLSKELWFSDDPNVTKTLRFKQANGKERSIYFQPPSFYVSAQDLPHIENGGVAVLTGKKVVQLDVRDNMVKLNDGSQITYEKCLIATGGTPRSLSAIDRAGAEVKSRTTLFRKIGDFRSLEKISREV\n\n### Answer:", "output": "Increases protein dimerization at lower NADH levels.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00146", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: CC(C)C(=O)NCCNCC(O)COc1ccc(O)cc1\nDescriptors: MW=296.37, LogP=0.49, TPSA=90.82, HBD=4, HBA=5", "output": "Descriptors: MW=296.37, LogP=0.49, TPSA=90.82, HBD=4, HBA=5, QED=0.5, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "CC(C)C(=O)NCCNCC(O)COc1ccc(O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 296.37, "LogP": 0.49, "TPSA": 90.82, "QED": 0.5, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"label": -0.15}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00147", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nIRCIGVSNRDFVEGMSGGTWVDVVLEHGGCVTVMAQDKPTVDIELVTTTVSNMAEVRSYCYEASISDMASDSRCPTQGEAYLDKQSDTQYVCKRTLVDRGWGNGCGLFGKGSLVTCAKFACSKKMTGKSIQPENLEYRIMLSVHGSQHSGMIVNDTGHETDENRAKVEITPNSPRAEATLGGFGSLGLDCEPRTGLDFSDLYYLTMNNKHWLVHKEWFHDIPLPWHAGADTGTPHWNNKEALVEFKDAHAKRQTVVVLGSQEGAVHTALAGALEAEMDGAKGRLSSGHLKCRLKMDKLRLKGVSYSLCTAAFTFTKIPAETLHGTVTVEVQYAGTDGPCKVPAQMAVDMQTLTPVGRLITANPVITESTENSKMMLELDPPFGDSYIVIGVGEKKITHHWHRSGSTIGKAFEATVRGAKRMAVLGDTAWDFGSVGGALNSLGKGIHQIFGAAFKSLFGGMSWFSQILIGTLLMWLGLNTKNGSISLMCLALGGVLIFLSTAVSA\n\n### Answer:", "output": "<SEQ_2D>CCCSSCSSEEEEECCTTTTBCCEEEETTCEEEEECSSSCCEEEECCCEEECCCEECCCEECSEECCCCEEEECCTTSCCCCCGGGGGSSEEEEEEEEEESGGGSSSSSEEEEEEEEEECEESCEECCEECCSSSEEEEEEEEESSSCCSCTTSCHHHHHHSTTEEEEEECSSSCSEEEEETTTEEEEEEEESTTSSCTTSEEEEESSSCEEEEEHHHHSSCCSCBCCSCCCSSCCCSCCSSSBCCEECSSSCEECCBCCCCHHHHHHHTTTCEEEEEETTEEEECCCEEEEEEEEEEECCSCTTCCBCCSCCCBCSCCCBCSSSCEEEEEEETTCCSSCBCCCEEESSSSSCCCCCCBSSCSCBCCCCSSCEEEEEEECCCSSEEEEECCSSTTCCCEEEECSSCHHHHHHHHHHHHHHHHHTTTTSSTTTSCCCSSSHHHHHHHHHHHHHHHHHHSCSCCHHHHHHHHHHHHHHTTSSSTTTHHHHHHHHHHHHHHHHHTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00148", "category": "Mutation", "instruction": "Describe the effect of mutation Q22799-T67A on protein Dynein light chain 1, cytoplasmic.", "input": "### Wild-type:\nMVDRKAVIKNADMSDDMQQDAIDCATQALEKYNIEKDIAAYIKKEFDKKYNPTWHCIVGRNFGSYVTHETKHFIYFYLGQVAILLFKSG\n\n### Mutant:\nMVDRKAVIKNADMSDDMQQDAIDCATQALEKYNIEKDIAAYIKKEFDKKYNPTWHCIVGRNFGSYVAHETKHFIYFYLGQVAILLFKSG\n\n### Answer:", "output": "Reduces mett-10 binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00149", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQLQARRLDGIDYNPWVEFVKLASEHDVVNLGQGFPDFPPPDFAVEAFQHAVSGDFMLNQYTKTFGYPPLTKILASFFGELLGQEIDPLRNVLVTVGGYGALFTAFQALVDEGDEVIIIEPFFDCYEPMTMMAGGRPVFVSLKPGLGSSSNWQLDPMELAGKFTSRTKALVLNTPNNPLGKVFSREELELVASLCQQHDVVCITDEVYQWMVYDGHQHISIASLPGMWERTLTIGSAGKTFSATGWKVGWVLGPDHIMKHLRTVHQNSVFHCPTQSQAAVAESFEREQLLFRQPSSYFVQFPQAMQRCRDHMIRSLQSVGLKPIIPQGSYFLITDISDFKRKMPDLPGAVDEPYDRRFVKWMIKNKGLVAIPVSIFYSVPHQKHFDHYIRFCFVKDEATLQAMDEKLRKWKVELENL\n\n### Answer:", "output": "<SEQ_3Di>DPDDPLCPPLPDDPVVVVVVLLVPFQFQEFADQAAPDDFDPLLVVLLVCLVPDDPCLQDFDAQQAHQLLLVLVQVVVCVVLVHRFDSPQFKGKAQFLVRVLLLLLLLFAAAAAEEEEEPLAPLVVQSSNSSRNYHYDYFYFCAPLAALAPRDGDLVVVLVSDDLRHAEYEAEPCGPPAQAADDPVRLVSVQVSCVVSVYAYEYEDQWQLQFDDPDHHDDSCSPPPNLARYKYKYGLCSVRSNVVLRMMMIGGGNVRVVSSSVVSCVPPSGDNSSSSSSSSSLSVVCSVCPPPPPHCSNPVSVVLVVLQVLVCVLCVLLPWSWHHHRIGQKIKTFCVVVCVVPVPQDDDPPDQVLVSVQSVCCVPQRYHWAQSLSSDDPVCSNPRSGMIMGGSSYDVVSSVVVSVSSVVSNVVVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00150", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSTAIFSDRYKGQRVLGKGSFGEVILCKDKITGQECAVKVISKRQVKQKTDKESLLREVQLLKQLDHPNIMKLYEFFEDKGYFYLVGEVYTGGELFDEIISRKRFSEVDAARIIRQVLSGITYMHKNKIVHRDLKPENLLLESKSKDANIRIIDFGLSTHFEASKKMKDKIGTAYYIAPEVLHGTYDEKCDVWSTGVILYILLSGCPPFNGANEYDILKKVEKGKYTFELPQWKKVSESAKDLIRKMLTYVPSMRISARDALDHEWIQTYTKEVPSLDNAILNIRQFQGTQKLAQAALLYMGSKLTSQDETKELTAIFHKMDKNGDGQLDRAELIEGYKELMRMKGQDASMLDASAVEHEVDQVLDAVDFDKNGYIEYSEFVTVAMDRKTLLSRERLERAFRMFDSDNSGKISSTELATIFGVSDVDSETWKSVLSEVDKNNDGEVDFDEFQQMLLKLCGN\n\n### Answer:", "output": "<SEQ_3Di>DDDDDVVFKAFDAWPAAAWQATWTFIAGPPPRDTWTKGKHFCVQWPWPDDPVVVVVVVVVQQPDDDQAAWHWPDWDDDDTIIITITDDAAQAFPLVVLLPDPFDFLLNLLVQLLLVLVVLVSVVVVFWAQQQDDRNQWGFRYVDPPTGTHGPRGHVVNTIDTDPQLLHVQPRLLQGALVVVVPDDGRLRVLSSSLQVSVLQQQVDGQQDDLARVSSNVSNNVLDGDCPDPSSVQDDPQSVVLSCLSSPRPSVSRDHSVVSCVGPSSVPSNDPGDTRPNSSVSNLQVVLLRLLLSLLVVLLVLVVVLVVLLVQLVVLVCVVPPPNPQWDALVSNLVSVVVSCVVPVHDPQPVDSVSSSVLLVVLCVVPPDPRPPIDRVVSSSVSVSCVSSVVSVVSNVVSVVSNVVVPPQKDFLVSVSSNCNSNVDDPVLLVVLLCPQDVPDDGMDGVVSVVSSSVVVSPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00151", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMFLSLPMLTVLIPLVSLAGLFYSASVEDDFPQGCTSTTSLCFYSLLLPITIPVYVFFHLWTWMGIKLFRHN\n\n### Sequence 2:\nMIRSLPTMTVLIPLVSLAGLLYSASVEEGFPEGCTSASSLCFYSLLLPVTVPVYVFFHLWTWMGLKLFRHN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00152", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMENLNMDLLYMAAAVMMGLAAIGAAIGIGILGGKFLEGAARQPDLIPLLRTQFFIVMGLVDAIPMIAVGLGLYVMFAVA\n\n### Answer:", "output": "<SEQ_2D>CCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCGGGHHHHHHHHHHHHHHHTHHHHHHHHHHHHHHHTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00153", "category": "Mutation", "instruction": "Describe the effect of mutation P68688-A11C on protein Glutaredoxin 1.", "input": "### Wild-type:\nMQTVIFGRSGAPYCVRAKDLAEKLSNERDDFQYQYVDIRAEGITKEDLQQKAGKPVETVPQIFVDQQHIGGYTDFAAWVKENLDA\n\n### Mutant:\nMQTVIFGRSGCPYCVRAKDLAEKLSNERDDFQYQYVDIRAEGITKEDLQQKAGKPVETVPQIFVDQQHIGGYTDFAAWVKENLDA\n\n### Answer:", "output": "Enables AMPylation by YdiU.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00154", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "SLC1A3, AQP4, GFAP, S100B", "output": "This cell is most consistent with astrocyte, an astrocyte. Key markers such as GFAP, AQP4, S100B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "astrocyte", "tissue": "lymph node", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SLC1A3", "AQP4", "GFAP", "S100B"]}}
{"id": "eval_00155", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "RBFOX3, SATB2, SLC17A7", "output": "This profile is most consistent with excitatory neuron, which is commonly found in breast under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "breast", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["RBFOX3", "SATB2", "SLC17A7"]}}
{"id": "eval_00156", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSYE_SHELP\n\n### Answer:", "output": "Glutamate--tRNA ligase, Glutamyl-tRNA synthetase, GluRS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00157", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNQNTLYVNNLNDKINKNDLRTALYMLFSTYGTVVDIVALKTPKMRGQAHVVFFDPSAAAIAMKALKNFIFFGKEMKIQYAHSKSKIIERIVAENDSRGPLKRLRDEADLE\n\n### Sequence 2:\nQSIYIKHINEKIKKEELKRSLYCLFSQFGRLLDVVALKTPKLRGQAWVVFTEVTAASNAVRQMQNFPFYDKPMRIQYAKSKSDYVTK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00158", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDAPA_BIFA0\n\n### Answer:", "output": "Catalyzes the condensation of (S)-aspartate-beta-semialdehyde [(S)-ASA] and pyruvate to 4-hydroxy-tetrahydrodipicolinate (HTPA).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00159", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMVRPKPTSVLSFDYGHKRIGLAGCDPLGITVTALSPIHRISFQKDLTLIKSLCIEREIKGLIVGLPLDESGQNTVQSIHCQNYGIKIAKELDLPLAWVNEHSSSWDAGQKYNLQNDRTGKLDSAVAALLLEQWLREGPELQPVSKLDSPRIKF\n\n### Sequence 2:\nMDDSEEDQRLPHHREPKEFIPLDKLSELGILSWRLNADDWENDEKLKKIREARGYSYMDICDVCPEKLPNYEAKIKNFFEEHLRTDEEIRYCLEGSGYFDVRDENDQWIRVAVKKGGMIVLPAGMYHRFTLDTDNYIKAMRLFVGEPVWTPYNRPHDHLPARKEYVEKIINRGGNQAVEAR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00160", "category": "Mutation", "instruction": "Describe the effect of mutation P0AC23-L89A on protein Formate channel FocA.", "input": "### Wild-type:\nMKADNPFDLLLPAAMAKVAEEAGVYKATKHPLKTFYLAITAGVFISIAFVFYITATTGTGTMPFGMAKLVGGICFSLGLILCVVCGADLFTSTVLIVVAKASGRITWGQLAKNWLNVYFGNLVGALLFVLLMWLSGEYMTANGQWGLNVLQTADHKVHHTFIEAVCLGILANLMVCLAVWMSYSGRSLMDKAFIMVLPVAMFVASGFEHSIANMFMIPMGIVIRDFASPEFWTAVGSAPENFSHLTVMNFITDNLIPVTIGNIIGGGLLVGLTYWVIYLRENDHH\n\n### Mutant:\nMKADNPFDLLLPAAMAKVAEEAGVYKATKHPLKTFYLAITAGVFISIAFVFYITATTGTGTMPFGMAKLVGGICFSLGLILCVVCGADAFTSTVLIVVAKASGRITWGQLAKNWLNVYFGNLVGALLFVLLMWLSGEYMTANGQWGLNVLQTADHKVHHTFIEAVCLGILANLMVCLAVWMSYSGRSLMDKAFIMVLPVAMFVASGFEHSIANMFMIPMGIVIRDFASPEFWTAVGSAPENFSHLTVMNFITDNLIPVTIGNIIGGGLLVGLTYWVIYLRENDHH\n\n### Answer:", "output": "Does not affect intracellular formate levels.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00161", "category": "Mutation", "instruction": "Describe the effect of mutation P0C1V0-A77E on protein Ribosomal RNA large subunit methyltransferase H.", "input": "### Wild-type:\nMKITILAVGKLKEKYWKQAIAEYEKRLGPYTKIDIIEVPDEKAPENMSDKEIEQVKEKEGQRILAKIKPQSTVITLAIQGKMLSSEGLAQELNQRMTQGQSDFVFVIGGSNGLHKDVLQRSNYALSFSKMTFPHQMMRVVLIEQVYRAFKIMRGEAYHK\n\n### Mutant:\nMKITILAVGKLKEKYWKQAIAEYEKRLGPYTKIDIIEVPDEKAPENMSDKEIEQVKEKEGQRILAKIKPQSTVITLEIQGKMLSSEGLAQELNQRMTQGQSDFVFVIGGSNGLHKDVLQRSNYALSFSKMTFPHQMMRVVLIEQVYRAFKIMRGEAYHK\n\n### Answer:", "output": "30% increase in methyltransferase activity. The function of OrfX has not been determined.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00162", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nVCAN LYZ S100A8 FCN1\n</cell>", "output": "This cell is most consistent with CD14+ monocyte, a classical CD14+ monocyte. Key markers such as S100A8, FCN1, LYZ support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "CD14+ monocyte", "tissue": "brain", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["VCAN", "LYZ", "S100A8", "FCN1"]}}
{"id": "eval_00163", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nETEEKYAMTWSSNREHVFEMPTGGAATMNAGDNLLYLARKEQALALATQLRTQFKIQDYKIYRIFPSGEVQYLHPKDGVLPYLVNEGREQVGRIKATIGKNVNPAQVKFTSKATYDR\n\n### Answer:", "output": "<SEQ_3Di>DQAKKKKKKFAACDQDWDAQLVGDTDTADNTIDIDIGRDPVSNVVVVVVCCVPVVTLPMWMWIQGPVGDIHTDPPNVSDNPVDDDPPDDDDDDQPDDPVQFADCVVQPPHPDDSRPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00164", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "C=CC(C)(O)CC/C=C(/C)CCC=C(C)C", "output": "The SMILES represents an unnamed molecule with MW=222.37, LogP=4.4, TPSA=20.23. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "C=CC(C)(O)CC/C=C(/C)CCC=C(C)C", "compound_name": null, "pubchem_cid": null, "MW": 222.37, "LogP": 4.4, "TPSA": 20.23, "QED": 0.63, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00165", "category": "Mutation", "instruction": "Describe the effect of mutation Q4VYA5-Y133F on protein 5-hydroxyisourate hydrolase.", "input": "### Wild-type:\nMKRHILATVIASLVAAPAMALAAGNNILSVHILDQQTGKPAPGVEVVLEQKKDNGWTQLNTGHTDQDGRIKALWPEKAAAPGDYRVIFKTGQYFESKKLDTFFPEIPVEFHISKTNEHYHVPLLLSQYGYSTYRGS\n\n### Mutant:\nMKRHILATVIASLVAAPAMALAAGNNILSVHILDQQTGKPAPGVEVVLEQKKDNGWTQLNTGHTDQDGRIKALWPEKAAAPGDYRVIFKTGQYFESKKLDTFFPEIPVEFHISKTNEHYHVPLLLSQYGYSTFRGS\n\n### Answer:", "output": "Reduces activity by 90%. The deep, negatively charged thyroxine-binding pocket that characterises vertebrate TTR contrasts with a shallow and elongated, positively charged cleft in S. dublin TLP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00166", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSSADSTPPWLTLVGIGEDGYPGLGKQARRALLQASRIVGAARQLELLPPCIGAARETWPTPFSLEPLLARRGQPTCVLASGDPMLFGVGASLARQLPATELRVLPAPSSLSLAAARLGWALQEVACLSLVARPLAALQAQVHDGRRLLVLSNDGDSPAAIARLLNARGFGASRLSVLEHLGGPLERRLDGLAGDWSLPRAADLNLVAVECRAGAGAVRLPLTPGLADEAYRHDGQLTKRDVRAVTLARLAPCPGELLWDVGAGCGSIGIEWMRAHPSCRAIAIEANDERQEHIRHNRDALGVPTLHLVAGSAPEALLELPEPDAIFIGGGVTAEGVLETCWERLRPGGRLLANAVTLQSERVLLDWQQRLGGDLTRLGVAHARPLGGFDTWRQALPITLLELRKPAARKPPRDA\n\n### Sequence 2:\nIIVVGIGADGMTGLSEHSRSELRRATVIYGSKRQLALLDDTVTAERWEWPTPMLPAVQGLSPDGADLHVVASGDPLLHGIGSTLIRLFGHDNVTVLPHVSAVTLACARMGWNVYDTEVISLVTAQPHTA----VRRGGRAIVLSGDRSTPQALAVLLTEHGRGDSKFSVLEQLGGPAERRRDGTARAWACDPPLDVDELNVI------AVRYLLDERTSWAPDEAFAHDGQITKHPIRVLTLAALAPRPGQRLWDVGAGSGAIAVQWCRSWPGCTAVAFERDERRRRNIGFNAAAFGVSV--DVRGDAPDAFDDAARPSVIFLGGGVTQPGLLEACLDSLPAGGNLVANAVTVESEAALAHAYSRLGGELRRFQHYLGEPLGGFTGWRPQLPVTQWSVTK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00167", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMKLMIVGNTGSGKTTLLQQLMKTKKSDLGMQSATVGIDVKDWPIQIRDKRKRDLVLNVWDFAGREEFYSTHPHFMTQRALYLAVYDLSKGQAEVDAMKPWLFNIKARASSSPVILVGTHLDVSDEKQRKACMSKITKELLNKRGFPAIRDYHFVNATEESDALAKLRKTII\n\n### Answer:", "output": "<SEQ_2D>CCCEECCCTTSSHHHHCCCCCCCCCCCCCCCCCCCCEEECCCCCCCCCCCCCCCCCCCEEECSHHHHHTTSHHHHHHSCCEEEEEETTSCSSTTTTTHHHHHHHHHHCTTCEEEEEEECCTTCCCCCCCCCCCCCCCCCCCCCCCCEEEEEEECCTTSCCCCCCCCTTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00168", "category": "Mutation", "instruction": "Describe the effect of mutation O46567-S552T on protein Glucocorticoid receptor.", "input": "### Wild-type:\nMDSKESLTPGKEENPSSVLTQERGNVMDFCKILRGGATLKVSVSSTSLAAASQSDSKQQRLLVDFPKGSVSNAQQPDLSKAVSLSMGLYMGETETKVMGNDLGFPQQGQISLSSGETDLQLLEESIANLNRSTSVPENPKSSASSSVSAAPKEKEFPKTHSDVSSEQQNLKGQTGTNGGNVKLYTADQSTFDILQDLEFSSGSPGKETNQSPWKSDLLIDENCLLSPLAGEEDSFLLEGNSNEDCKPLILPDTKPKIKDNGDLVLSSSSNVTLPQVKTEKEDFIELCTPGVIKQEKLSTV\n\n### Mutant:\nMDSKESLTPGKEENPSSVLTQERGNVMDFCKILRGGATLKVSVSSTSLAAASQSDSKQQRLLVDFPKGSVSNAQQPDLSKAVSLSMGLYMGETETKVMGNDLGFPQQGQISLSSGETDLQLLEESIANLNRSTSVPENPKSSASSSVSAAPKEKEFPKTHSDVSSEQQNLKGQTGTNGGNVKLYTADQSTFDILQDLEFSSGSPGKETNQSPWKSDLLIDENCLLSPLAGEEDSFLLEGNSNEDCKPLILPDTKPKIKDNGDLVLSSSSNVTLPQVKTEKEDFIELCTPGVIKQEKLSTV\n\n### Answer:", "output": "Decreased nuclear translocation; when associated with S-617 and A-619. Truncated gsmGR constructs lacking the LDB after amino acid 552 also showed decreased translocation even in the absence of cortisol.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00169", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "HNF4A, TTR, APOB", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as TTR, APOB, HNF4A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "heart", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["HNF4A", "TTR", "APOB"]}}
{"id": "eval_00170", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNIKR_SHIB3\n\n### Answer:", "output": "Nickel-responsive regulator.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00171", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nENLYFQGAMSRSYNDELQFLEKINKNCWRIKKGFVPNMQVEGVFYVNDALEKLMFEELRNACRGGFLPAMKQIGNVAALPGIVHRSIGLPDVHSGYGFAIGNMAAFDMNDPEAVVSPGGVGFDINCGVRLLRTNLDESDVQPVKEQLAQAMFDHIPVVIPMNAKDLEEALEMGVDWSLREGYAWAEDKEHCEEYGRMLQADPNKVSARAKKRGLPQLGTLGAGNHYAEIQVVDEIFNEYAAKKMGIDHKGQVCVMIHSGSRGLGHQVATDALVAMEKAMKRDKIIVNDRQLACARIASPEGQDYLKGMAAAGNYAWVNRSSMTFLTRQAFAKVFNTTPDDLDLHVIYDVSHNIAKVEQHVVDGKERTLLVHRKGSTRAFPPHHPLIAVDYQLTGQPVLIGGTMGTCSYVLTGTEQGMTETFGTTCHGAGRFQDVLDKLADMGIAIRVASPKLVYKNVTDVVNTCHDAGISKKAIKLRPIAVIKG\n\n### Answer:", "output": "<SEQ_3Di>DADDPVLLQAAALVVLCVQKAADDQFKIKRHPSSHPQQQEIEMEGAHPVLVVVLSVQNNVCNVVVLGRQVSLLRLVSRQPQWPDGWYAAHQWGCDQSHTQRIKTKGFLVPQSTWDFLSLQDLQQFWWKWKWFWPDFLVVCQVCQQLLQVLLQVLQVCDDDDDPVLLQCLLQAQCVSVVVVVLADQLQLLQEVVSRADNLADSVLQDPVLSVQQHSPVRHAQPFQWKKFKWWFDDFDACLVCVLLVNNDGRTIMMMITHFLGSSSNRLSVVLLVVQVVVCVVVVTDGNDSSRRIGGCPDPSNSSSQNSSSSSNSSGSSSVSNSLSSSQVSVCVVVVHHSVNTRRHISHMWGARHWDFDWAQDPNRIGTITMRGRQKTAWAAACDPSHRPSCNHPTRWAKADDPLWAKIFIKHAHPVLCVNVVRMHGRYHYDPVVSVVVCVVVQAAEDGDDPVVRHDHRCSSVVSCCVVGRIHTTTMIGGRYIYTD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00172", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYGCW_ECOLI\n\n### Answer:", "output": "Uncharacterized oxidoreductase YgcW.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00173", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nK1549_HUMAN\n\n### Answer:", "output": "May play a role in photoreceptor function.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00174", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nLITH_PIG\n\n### Answer:", "output": "Might act as an inhibitor of spontaneous calcium carbonate precipitation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00175", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGSA2_BACC4\n\n### Answer:", "output": "Glutamate-1-semialdehyde 2,1-aminomutase 2, GSA 2, Glutamate-1-semialdehyde aminotransferase 2, GSA-AT 2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00176", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: Cc1ccc(NC(=O)C(=O)C2NC(=S)NC2=O)c([N+](=O)[O-])c1\nDescriptors: MW=322.3, LogP=-0.22, TPSA=130.44, HBD=3, HBA=6", "output": "Descriptors: MW=322.3, LogP=-0.22, TPSA=130.44, HBD=3, HBA=6, QED=0.23, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "Cc1ccc(NC(=O)C(=O)C2NC(=S)NC2=O)c([N+](=O)[O-])c1", "compound_name": null, "pubchem_cid": null, "MW": 322.3, "LogP": -0.22, "TPSA": 130.44, "QED": 0.23, "functional_groups": ["aromatic ring", "amide", "amine", "ketone", "nitro"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00177", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKILKPYMRNTSISCYLCFLLNSHFLTEAGIHVFILGCVSVGLPKTEANWIDVRYDLEKIESLIQSIHIDTTLYTDSDFHPSCKVTAMNCFLLELQVILHEYSNMTLNETVRNVLYLANSTLSSNKNVAESGCKECEELEEKTFTEFLQSFIRIVQMFINTS\n\n### Sequence 2:\nMRISKPHLRITSIQCYVCLLLNTHFLTEAGIRVFILGCISAGIPKTEANWEDVRKDLQKIENLIQSLHMDATLYTESDVHPRCKVTAMNCFLLELEVISHESRDGDIEETVKNLILLANSSLSSNGNITESGCKVCEELEEKNITEFLESFKHIVQMFIN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00178", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: CC(C)c1cccc(C[NH2+]C[C@@H](O)[C@@H]2C[C@H](C)CCCCCCCCC(=O)N(C)[C@@H](C)C(=O)N2)c1\nEvidence: MoleculeNet BACE reports activity against BACE1.", "output": "The ligand is relevant to beta-secretase (BACE1). MoleculeNet BACE reports activity against BACE1. The relationship is best described as target-associated activity evidence rather than an unsupported binding prediction.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BACE", "canonical_smiles": "CC(C)c1cccc(C[NH2+]C[C@@H](O)[C@@H]2C[C@H](C)CCCCCCCCC(=O)N(C)[C@@H](C)C(=O)N2)c1", "compound_name": null, "pubchem_cid": null, "MW": 488.74, "LogP": 3.73, "TPSA": 86.25, "QED": 0.57, "functional_groups": ["aromatic ring", "amide", "hydroxyl"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": true}}
{"id": "eval_00179", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nLEUC_BRUA1\n\n### Answer:", "output": "3-isopropylmalate dehydratase large subunit, Alpha-IPM isomerase, IPMI, Isopropylmalate isomerase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00180", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CC(=O)[O-].CC(=O)[O-].[Mg+2]", "output": "This molecule is small and moderately polar, with hydrophilic character. It contains no standard functional groups detected. Its descriptor profile: MW=142.39, LogP=-2.87, TPSA=80.26, HBD=0, HBA=4, QED=0.34.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CC(=O)[O-].CC(=O)[O-].[Mg+2]", "compound_name": null, "pubchem_cid": null, "MW": 142.39, "LogP": -2.87, "TPSA": 80.26, "QED": 0.34, "functional_groups": [], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": null, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00181", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSECE_METKA\n\n### Answer:", "output": "Essential subunit of the Sec protein translocation channel SecYEG.  Clamps together the 2 halves of SecY.  May contact the channel plug during translocation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00182", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nALLSFERKYRVPGGTLVGGNLFDFWVGPFYVGFFGVATFFFAALGIILIAWSAVLQGTWNPQLISVYPPALEYGLGGAPLAKGGLWQIITICATGAFVSWALREVEICRKLGIGYHIPFAFAFAILAYLTLVLFRPVMMGAWGYAFPYGIWTHLDWVSNTGYTYGNFHYNPAHMIAISFFFTNALALALHGALVLSAANPEKGKEMRTPDHEDTFFRDLVGYSIGTLGIHRLGLLLSLSAVFFSALCMIITGTIWFDQWVDWWQWWVKLPWWANIPGGING\n\n### Answer:", "output": "<SEQ_2D>CCCTTTGGGCCSCCCSSSSSTTCCEETTEECCTHHHHHHHHHHHHHHHHHHHHHTTSCCCTTTCEECCSCSSSCSSCCCTTTTHHHHHHHHHHHHHHHHHHHHHHHHHHSSCCCCCSTTTHHHHHHHHHHTTTHHHHHHSCGGGSCCEESSSSHHHHHHHHHTTSCSTTCHHHHHHHHHHHHHHHHHHHHHHHHTTTTSCSSSSCCCCHHHHHHHHHHHTSCCCTTTTHHHHHHHHHHHHHHHHTTTTTSBTTTBCSCHHHHTTTTTTCGGGSSCSSSSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00183", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVHLTPEEKSAVTALWGKVNVDEVGGEALGRLLVVYPWTQRFFESFGDLSTPDAVMGNPKVKAHGKKVLGAFSDGLAHLDNLKGTFATLSELHCDKLHVDPENFRLLGNVLVGVLAHHFGKEFTPPVQAAYQKVVAGVANALAHKYH\n\n### Answer:", "output": "<SEQ_3Di>DPDDPVLLVLVVPLVVPDPLLVLLLQLLLVLCVVPVVVCVLVVVQDDNPDSCRSSVGPSSSVRRSVVSVLLVVLSVPPVPLLVSCVVVLVCCAVPVVPDLVVLVSSLVSSLVVSCVVVPVVCDPSNSVSSVVSSVSNSVSSNPVHD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00184", "category": "Mutation", "instruction": "Describe the effect of mutation P78325-W35R on protein Disintegrin and metalloproteinase domain-containing protein 8.", "input": "### Wild-type:\nMRGLGLWLLGAMMLPAIAPSRPWALMEQYEVVLPWRLPGPRVRRALPSHLGLHPERVSYVLGATGHNFTLHLRKNRDLLGSGYTETYTAANGSEVTEQPRGQDHCFYQGHVEGYPDSAASLSTCAGLRGFFQVGSDLHLIEPLDEGGEGGRHAVYQAEHLLQTAGTCGVSDDSLGSLLGPRTAAVFRPRPGDSLPSRETRYVELYVVVDNAEFQMLGSEAAVRHRVLEVVNHVDKLYQKLNFRVVLVGLEIWNSQDRFHVSPDPSVTLENLLTWQARQRTRRHLHDNVQLITGVDFTGTT\n\n### Mutant:\nMRGLGLWLLGAMMLPAIAPSRPWALMEQYEVVLPRRLPGPRVRRALPSHLGLHPERVSYVLGATGHNFTLHLRKNRDLLGSGYTETYTAANGSEVTEQPRGQDHCFYQGHVEGYPDSAASLSTCAGLRGFFQVGSDLHLIEPLDEGGEGGRHAVYQAEHLLQTAGTCGVSDDSLGSLLGPRTAAVFRPRPGDSLPSRETRYVELYVVVDNAEFQMLGSEAAVRHRVLEVVNHVDKLYQKLNFRVVLVGLEIWNSQDRFHVSPDPSVTLENLLTWQARQRTRRHLHDNVQLITGVDFTGTT\n\n### Answer:", "output": "The functional changes caused by the mutation are not mentioned in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00185", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRPB4_YEAST\n\n### Answer:", "output": "DNA-dependent RNA polymerase catalyzes the transcription of DNA into RNA using the four ribonucleoside triphosphates as substrates.  Component of RNA polymerase II which synthesizes mRNA precursors and many functional non-coding RNAs.  Pol II is the central component of the basal RNA polymerase II transcription machinery.  It is composed of mobile elements that move relative to each other.  RPB4 is part of a subcomplex with RPB7 that binds to a pocket formed by RPB1, RPB2 and RPB6 at the base of the clamp element.  The RBP4-RPB7 subcomplex seems to lock the clamp via RPB7 in the closed conformation thus preventing double-stranded DNA to enter the active site cleft.  The RPB4-RPB7 subcomplex binds single-stranded DNA and RNA.  The RPB4-RPB7 subcomplex is necessary for promoter-directed transcription initiation but is not required for recruitment of Pol II to active preinitiation complexes and seems to be dispensable for transcription elongation and termination.  The RPB4-RPB7 subcomplex recruits FCP1 to Pol II.  Involved in DNA repair of damage in the transcribed strand.  RPB4 is dispensable under optimal growth conditions, but becomes essential during heat or cold shock and under nutrient depletion.  Suppresses the RBP9-mediated transcription-coupled repair (TCR) subpathway of nucleotide excision repair (NER) but facilitates the RAD26-mediated TCR subpathway.  Under stress conditions only, involved in mRNA export to the cytoplasm.  Involved in mRNA decay.  Promotes or enhances the deadenylation process of specific mRNAs and may recruit PAT1 and the LSM1-7 complex to these mRNAs, thus stimulating their decapping and further decay.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00186", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nT151B_HUMAN\n\n### Answer:", "output": "Belongs to the TMEM151 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00187", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CC1COC(=O)O1", "output": "The SMILES represents an unnamed molecule with MW=102.09, LogP=0.54, TPSA=35.53. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "CC1COC(=O)O1", "compound_name": null, "pubchem_cid": null, "MW": 102.09, "LogP": 0.54, "TPSA": 35.53, "QED": 0.42, "functional_groups": ["carboxylic acid", "ester", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00188", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMATSLSVSRFMSSSATVISVAKPLLSPTVSFTAPLSFTRSLAPNLSLKFRNRRTNSASATTRSFATTPVTASISVGDKLPDSTLSYLDPSTGDVKTVTVSSLTAGKKTILFAVPGAFTPTCSQKHVPGFVSKAGELRSKGIDVIACISVNDAFVMEAWRKDLGINDEVMLLSDGNGEFTGKLGVELDLRDKPVGLGVRSRRYAILADDGVVKVLNLEEGGAFTNSSAEDMLKAL\n\n### Sequence 2:\nMEALKKIAGVTAAQYVTDGMTIGLGTGSTAYYFVEEIGRRIKEEGLQVVGVTTSSVTTKQAEGLGIPLTSIDDIDCIDLTVDGADEVDKAFNGIKGGGAALLMEKIVATPTKEYIWVVDESKLVDHLGAFKLPVEVVQYGADRLFRVFERAGYKPSFRMKGDKRLITDMQNFIIDLNLGCIENPCEFGRLLDQTVGVVEHGLFNGMVDKVIVAGQAGVTVLEANQST\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00189", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMADTLERVTKIIVDRLGVDEADVKLEASFKEDLGADSLDVVELVMELEDEFDMEISDEDAEKIATVGDAVNYIQNQQ\n\n### Sequence 2:\nMEVTDVRLRRVNTEGRMRAIASITLDHEFVVHDIRVIDGNNGLFVAMPSKRTPDGEFRDIAHPINSGTRSKIQDAVLTEYHRLGELEEVEFEEAGAS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00190", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGMKTRDKIIQASLELFNEHGERTITTNHIAAHLDISPGNLYYHFRNKEDIIRCIFDQYEQHLLLGFKPYADQKVDLELLMSYFDAMFYTMWQFRFMYANLADILARDDTLKARYLKVQQAVLEQSIAVLNQLKKDGILQIEDERIADLADTIKMIIGFWISYKLTQSSIATISKASLYEGLLRVLMIFKAYSTPDSLANFDRLEQHFRSQSN\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHHHHHHHHCGGGCCHHHHHHHHTCCHHHHHHHCSSHHHHHHHHHHHHHHHHHHHTCCCSSCCCSHHHHHHHHHHHHHHHHHTHHHHHSHHHHHHTCHHHHHHHHHHHHHHHHHHHHHHHHHHHTTSBCCCGGGHHHHHHHHHHHHHHHHHHHHHHSSCCCCCHHHHHHHHHHHHHHHHTTBCHHHHHHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00191", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nCCR7:high SELL:medium CD3E:high TCF7:low IL7R:low\n</cell>", "output": "This cell is most consistent with naive CD4+ T cell, a T-cell subtype with naive/central-memory-like profile. Key markers such as TCF7, CD3E, SELL support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "intestine", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CCR7", "SELL", "CD3E", "TCF7", "IL7R"]}}
{"id": "eval_00192", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKAKEIRDLTTSEIEEQIKSSKEELFNLRFQLATGQLEETARIRTVRKTIARLKTVAREREIEQSKANQ\n\n### Sequence 2:\nMKANEIRDLTTAEIEQKVKSLKEELFNLRFQLATGQLENTARIREVRKAIARMKTVIREREIAANK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00193", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nETS5_CAEEL\n\n### Answer:", "output": "ETS domain-containing transcription factor ets-5.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00194", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CC1OC(OCC2OC(Oc3c(-c4ccc(OCCO)c(OCCO)c4)oc4cc(OCCO)cc(O)c4c3=O)C(O)C(O)C2O)C(O)C(O)C1O", "output": "This molecule is relatively large and polar, with hydrophilic character. It contains aromatic ring, hydroxyl, ether, phenol. Its descriptor profile: MW=742.68, LogP=-2.69, TPSA=297.12, HBD=10, HBA=19, QED=0.08.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "SIDER", "canonical_smiles": "CC1OC(OCC2OC(Oc3c(-c4ccc(OCCO)c(OCCO)c4)oc4cc(OCCO)cc(O)c4c3=O)C(O)C(O)C2O)C(O)C(O)C1O", "compound_name": null, "pubchem_cid": null, "MW": 742.68, "LogP": -2.69, "TPSA": 297.12, "QED": 0.08, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 0, "Product issues": 0, "Eye disorders": 0, "Investigations": 0, "Musculoskeletal and connective tissue disorders": 0, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 0, "Reproductive system and breast disorders": 0, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 0, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 0, "Respiratory, thoracic and mediastinal disorders": 0, "Psychiatric disorders": 0, "Renal and urinary disorders": 0, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 0, "Cardiac disorders": 0, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00195", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSLTRLLIKDFRNIENADLALSPGFNFLVGANGSGKTSVLEAIYTLGHGRAFRSLQVGRVIRHEQESFILHGRLQGSERETSIGLTKDKQGDSKVRIDGTDGHKVAELALLMPMQLITPEGFTLLNGGPKYRRAFLDWGCFHNEAGFFNAWSNLKRLLKQRNAALRQVTRYAQVRPWDMELVPLAEQISRWRAEYSAGIAEDMADTCKQFLPEFSLTFSFQRGWEKETDYAEVLERSFERDRMLTYTAHGPHKADFRIRADGAPVEDTLSRGQLKLLMCALRLAQGEFLTRESGRRCLYLIDDFASELDDARRGLLASRLKATQSQVFVSAISAEHVLDMSDKNSKMFTVEKGKITD\n\n### Sequence 2:\nMSLTRLLIKDFRNIESADLALSPGFNFLVGANGSGKTSVLEAIYTLGHGRAFRSLQIGRVIRHEQEAFVLHGRLQGEERETAIGLTKDKQGDSKVRIDGTDGHKVAELAHLMPMQLITPEGFTLLNGGPKYRRAFLDWGCFHNEAGFFTAWSNLKRLVKQRNAALRQVSRYAQLRPWDLELIPLAEQISRWRAEYSAAIVEDMADTCQQFLPEFTLTFSFQRGWEKETDYAEVLERNFERDRMLTYTAHGPHKADFRIRADGAPVEDTLSRGQLKLLMCALRLAQGEFLTRVSGRRCLYLIDDFASELDDARRGLLSSRLKATQSQVFVSAISAEHVMDMSDKNSKMFRVEKGKITD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00196", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nFNKELDPVQKLFVDKIREYRTKRQTSGGPVDAGPEYQQDLDRELFKLKQMYGKADMNTFPNFTFEDPKFEVVEKPQS\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCHHHHHHHHHHHHHHHHTTCCTTTTTCHHHHHHHHHHHHHHHHHTTTCCSSSCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00197", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL14_LACLA\n\n### Answer:", "output": "Binds to 23S rRNA.  Forms part of two intersubunit bridges in the 70S ribosome.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00198", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMEMTNAQRLILSNQYKMMTMLDPANAERYRRLQTIIERGYGLQMRELDREFGELKEETCRTIIDIMEMYHALHVSWSNLQDQQSIDERRVTFLGFDAATEARYLGYVRFMVNVEGRYTHFDAGTHGFNAQTPMWEKYQRMLNVWHACPRQYHLSANEINQIINA\n\n### Sequence 2:\nMATKKNEQIKGRTDGLVAENRRSRREYSVEDTLEAGIMLVGSEVKSLREGRANIAESYASVEQGELWLINAEIQTYGGANRFNHEPRRKRKLLVSRRELSKLSQEVERAGRTIVPLKLYFNDKGRAKLLIGVATGRKAHDKREHEAKRDWARDKARIMKAG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00199", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nSLC1A3 AQP4 S100B GFAP\n</cell>", "output": "This cell is most consistent with astrocyte, an astrocyte. Key markers such as SLC1A3, AQP4, S100B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "astrocyte", "tissue": "lymph node", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SLC1A3", "AQP4", "S100B", "GFAP"]}}
{"id": "eval_00200", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMFVNTVIAFIIIFGTLVFFHELGHLLLAQRAGILCREFAIGFGPKIFSFKKNETVYTIRLLPVGGFVRMAGEDPEMIEVKPGYTVGLLFNKEDQVEKVIINQKEKYPDALVIEVETADLEHDMKITGYEQGKEDELSSFTVSETSFFIVDGEEVQIAPYNRQFGSKPVWQRIKAIAAGPIMNFILAYVILVMLGLIQGVPSNEPMLGQLTDNGRAAEAGLKEGDYIQSINGEKMRSWTDIVSAVKENPEKEMDVAVKRDNKTLHISVTPEAVKDENKKTIGRFGSYAPTEKGVLSAVAYGATSTVDVTKAILTNLSKLVTGQFKLDMLSGPVGIYDMTDQVAKTGIVNLFQFAAFLSINLGIVNLLPIPALDGGRLLFLFIEAIRGKPINREKEAFVVFIGVAFLMLLMLVVTWNDIQRLFL\n\n### Sequence 2:\nMNTLLAFIAIFSVLVFVHEWGHLYFAKKAGILCYEFAIGMGPKLFAFERNDTIYTIRLLPIGGYVRMAGEEPEQPTIRPGYEIGLVLDEKDTVKELIVNNKSKHPEAQVVQVERIDLVHDLFVETIDE-DTGELVRYPIDEKAFIVQDEVAQIIAPWKRQFGSKPLPKRAMAIFAGPLMNFILGFVILLGLSLYQGVTLSSEIVIN--GENSPAEAAGLQDGDVITAVNGVEVDSWKEMTTEVKKYPGEEVSIDYERNGEALQTNATLSQVEVMPDEYEGFLGVSGVPEFSLLGSLQYAGNEFINMATSIFDTLGLIFTGQFSLDYISGPVGIYDITDQAVSLGIQTVIFFAALLSINLGVINLMPIPALDGGRLMFLAYEGIRGKPVSPEKEGAIQFIGFALVMLLMIVVTWNDISKLF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00201", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "SLC17A7, SATB2, RBFOX3", "output": "The expression pattern is consistent with excitatory neuron in a context associated with lung adenocarcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "liver", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["SLC17A7", "SATB2", "RBFOX3"]}}
{"id": "eval_00202", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nHSLU_METSB\n\n### Answer:", "output": "ATP-dependent protease ATPase subunit HslU, Unfoldase HslU.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00203", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMEDEMPKTLYVGNLSRDVTEALILQLFSQIGPCKNCKMIMDTAGNDPYCFVEFHEHRHAAAALAAMNGRKIMGKEVKVNWATTPSSQKKDTS\n\n### Answer:", "output": "<SEQ_2D>CGGGCCCEEEEESCCTTSCHHHHHHHHHTTSCEEEEEEECCSSSCCCEEEEEESCHHHHHHHHHHHTTCEETTEECEEEEECCCSSCCCSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00204", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRS19_METHJ\n\n### Answer:", "output": "30S ribosomal protein S19.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00205", "category": "Mutation", "instruction": "Describe the effect of mutation O75506-K16V on protein Heat shock factor-binding protein 1.", "input": "### Wild-type:\nMAETDPKTVQDLTSVKQTLLQQMQDKFQTMSDQIIGRIDDMSSRIDDLEKNIADLMTQAGVEELESENKIPATQKS\n\n### Mutant:\nMAETDPKTVQDLTSVVQTLLQQMQDKFQTMSDQIIGRIDDMSSRIDDLEKNIADLMTQAGVEELESENKIPATQKS\n\n### Answer:", "output": "Loss of interaction with HSF1; in association with K-19. The mutation in HSF1 leads to slow DNA release and sustained transcriptional activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00206", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRSMA_FRATF\n\n### Answer:", "output": "Belongs to the class I-like SAM-binding methyltransferase superfamily. rRNA adenine N(6)-methyltransferase family. RsmA subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00207", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nISPE_OCEIH\n\n### Answer:", "output": "4-diphosphocytidyl-2-C-methyl-D-erythritol kinase, CMK, 4-(cytidine-5'-diphospho)-2-C-methyl-D-erythritol kinase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00208", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSFPEGKDILFMGNEAAKLAEAFQKSLRKPGHKRSQSIIGKHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCCGGGTSCCSHHHHHHHHHHHHHTCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00209", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "Nc1ccc(Oc2ccc(N)c(N)c2)cc1", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains aromatic ring, ether. Its descriptor profile: MW=215.26, LogP=2.23, TPSA=87.29, HBD=3, HBA=4, QED=0.67.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "Nc1ccc(Oc2ccc(N)c(N)c2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 215.26, "LogP": 2.23, "TPSA": 87.29, "QED": 0.67, "functional_groups": ["aromatic ring", "ether"], "moleculenet_labels": {"NR-AR": 1.0, "NR-AR-LBD": 0.0, "NR-AhR": 1.0, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 1.0, "SR-ATAD5": 1.0, "SR-HSE": null, "SR-MMP": 1.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00210", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMLATRRLLGWSLPARVSVRFSGDTTAPKKTSFGSLKDEDRIFTNLYGRHDWRLKGSLSRGDWYKTKEILLKGPDWILGEIKTSGLRGRGGAGFPTGLKWSFMNKPSDGRPKYLVVNADEGEPGTCKDREILRHDPHKLLEGCLVGGRAMGARAAYIYIRGEFYNEASNLQVAIREAYEAGLIGKNACGSGYDFDVFVVRGAGAYICGEETALIESIEGKQGKPRLKPPFPADVGVFGCPTTVANVETVAVSPTICRRGGTWFAGFGRERNSGTKLFNISGHVNHPCTVEEEMSVPLKELIEKHAGGVTGGWDNLLAVIPGGSSTPLIPKSVCETVLMDFDALVQAQTGLGTAAVIVMDRSTDIVKAIARLIEFYKHESCGQCTPCREGVDWMNKVMARFVRGDARPAEIDSLWEISKQIEGHTICALGDGAAWPVQGLIRHFRPELEERMQRFAQQHQARQAAS\n\n### Sequence 2:\nMLAARRLLGWSLPARVSVRFSGDTTAPKKTSFGSLKDEDRIFTNLYGRHDWRLKGALSRGDWYKTKEILLKGPDWILGEIKTSGLRGRGGAGFPTGLKWSFMNKPSDGRPKYLVVNADEGEPGTCKDREIIRHDPHKLVEGCLVGGRAMGARAAYIYIRGEFYNEASNLQVAIREAYEAGLIGKNACGSGYDFDVFVVRGAGAYICGEETALIESIEGKQGKPRLKPPFPADVGVFGCPTTVANVETVAVSPTICRRGGTWFAGFGRERNSGTKLFNISGHVNYPCTVEEEMSVPLKELIEKHAGGVTGGWDNLLAVIPGGSSTPLIPKSVCETVLMDFDALVQAQTGLGTAAVIVMDRSTDIVKAIARLIEFYKHESCGQCTPCREGVDWMNKVMARFVRGDAQPAEIDSLWEISKQIEGHTICALGDGAAWPVQGLIRHFRPELEERMQRFAQQHQAQQAAS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00211", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NP55-I76C on protein BPI fold-containing family A member 1.", "input": "### Wild-type:\nMFQTGGLIVFYGLLAQTMAQFGGLPVPLDQTLPLNVNPALPLSPTGLAGSLTNALSNGLLSGGLLGILENLPLLDILKPGGGTSGGLLGGLLGKVTSVIPGLNNIIDIKVTDPQLLELGLVQSPDGHRLYVTIPLGIKLQVNTPLVGASLLRLAVKLDITAEILAVRDKQERIHLVLGDCTHSPGSLQISLLDGLGPLPIQGLLDSLTGILNKVLPELVQGNVCPLVNEVLRGLDITLVHDIVNMLIHGLQFVIKV\n\n### Mutant:\nMFQTGGLIVFYGLLAQTMAQFGGLPVPLDQTLPLNVNPALPLSPTGLAGSLTNALSNGLLSGGLLGILENLPLLDCLKPGGGTSGGLLGGLLGKVTSVIPGLNNIIDIKVTDPQLLELGLVQSPDGHRLYVTIPLGIKLQVNTPLVGASLLRLAVKLDITAEILAVRDKQERIHLVLGDCTHSPGSLQISLLDGLGPLPIQGLLDSLTGILNKVLPELVQGNVCPLVNEVLRGLDITLVHDIVNMLIHGLQFVIKV\n\n### Answer:", "output": "Forms an artificial disulfide bond. Reduced surfactant activity; when associated with C-214.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00212", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y294-R108E on protein Histone chaperone ASF1A.", "input": "### Wild-type:\nMAKVQVNNVVVLDNPSPFYNPFQFEITFECIEDLSEDLEWKIIYVGSAESEEYDQVLDSVLVGPVPAGRHMFVFQADAPNPGLIPDADAVGVTVVLITCTYRGQEFIRVGYYVNNEYTETELRENPPVKPDFSKLQRNILASNPRVTRFHINWEDNTEKLEDAESSNPNLQSLLSTDALPSASKGWSTSENSLNVMLESHMDCM\n\n### Mutant:\nMAKVQVNNVVVLDNPSPFYNPFQFEITFECIEDLSEDLEWKIIYVGSAESEEYDQVLDSVLVGPVPAGRHMFVFQADAPNPGLIPDADAVGVTVVLITCTYRGQEFIEVGYYVNNEYTETELRENPPVKPDFSKLQRNILASNPRVTRFHINWEDNTEKLEDAESSNPNLQSLLSTDALPSASKGWSTSENSLNVMLESHMDCM\n\n### Answer:", "output": "Reduces interaction with histone H3.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00213", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "SPP1, MRC1, CD206, CD163, APOE", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as MRC1, SPP1, APOE support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "bone marrow", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SPP1", "MRC1", "CD206", "CD163", "APOE"]}}
{"id": "eval_00214", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPRIB_SALDC\n\n### Answer:", "output": "Binds single-stranded DNA at the primosome assembly site (PAS).  During primosome assembly it facilitates the complex formation between PriA and DnaT.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00215", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nHBB1_XENBO\n\n### Answer:", "output": "Hemoglobin subunit beta-1, Beta-1-globin, Hemoglobin beta-1 chain, Hemoglobin beta-major chain.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00216", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "FCER1A, CD1C, HLA-DQA1", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as CD1C, FCER1A, HLA-DQA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "breast", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["FCER1A", "CD1C", "HLA-DQA1"]}}
{"id": "eval_00217", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKTLSIEPLTRAAFAPFGDVIETQGAKQIPINLGTTMRFHDLAKIDVADEGGRPLVNLFRGQPRTLPFEVTMLERHPLGSQAFVPLTDRPYIVVVAPAGDLDASKIRAFVTSGWQGVNYAKGVWHHPLIALGEVSDFIVVDRGGDGRNLNEQNLPESLWLTDDALLAVGA\n\n### Sequence 2:\nMHYPEPISKLIDSFMKLPGIGPKTAARLAFHVLAMKEDTVLEFAKALVDVKRHIHYCTICGHITDTDPCYICKDERRDRTMICVVQDPKDVIAMEKMKEYNGLYHVLHGAISPMEGIGPEDIKIAELLARLQDETIQEVILATDPNIEGEATAMYLSRLLKPTGIKITRIAHGLPVGGDLEYADEVTLSKALEGRREL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00218", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nURER_ECOLX\n\n### Answer:", "output": "Positive regulator of the expression of the urease operon.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00219", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCZCR_BACHK\n\n### Answer:", "output": "HTH-type transcriptional regulator CzcR.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00220", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLSSAILSEKPNVKWEDVAGLEGAKEALKEAVILPVKFPHLFKGNRKPTSGILLYGPPGTGKSYLAKAVATEANSTFFSVSSSDLVSKWMGESEKLVKQLFAMARENKPSIIFIDQVDALTGTRGEGESEASRRIKTELLVQMNGVGNDSQGVLVLGATNIPWQLDSAIRRRFERRIYIPLPDLAARTTMFEINVGDTPCVLTKEDYRTLGAMTEGYSGSDIAVVVKDALMQPIRKIQSATHFKDVSTEDDETRKLTPCSPGDDGAIEMSWTDIEADELKEPDLTIKDFLKAIKSTRPTVNEDDLLKQEQFTRDFGQEGN\n\n### Answer:", "output": "<SEQ_3Di>DDDDADDDADPDEPVLQFDDPVVVVVVCVQAQVVVVDPDPNPDPDDHFLEAAEEDAPQQQLVVVVNNSVNPRPADEHEDDQPPQADPPQPDRCVSVCVVLVVQVVGLSHEYEHAAPCPLFDPPDDPCPGVVVVVVVSVVVVSVVCVPDDNSYYYYHHYYDDVRHDPVVLLSGAHYDYRYADALVRVLSNLVSLCPPAPDPPDVVVSNVVSPVDDDDHNVLVNVLSVQQVCPLVVVLQPAQWWDFPDDVVDPARAIERDDPPDPPTDRHHSVPDDPRNYDTDHGDVVSVVVSPVPGDHDDDCVSVVVVVVCVVPDRPDRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00221", "category": "Mutation", "instruction": "Describe the effect of mutation P05186-V423A on protein Alkaline phosphatase, tissue-nonspecific isozyme.", "input": "### Wild-type:\nMISPFLVLAIGTCLTNSLVPEKEKDPKYWRDQAQETLKYALELQKLNTNVAKNVIMFLGDGMGVSTVTAARILKGQLHHNPGEETRLEMDKFPFVALSKTYNTNAQVPDSAGTATAYLCGVKANEGTVGVSAATERSRCNTTQGNEVTSILRWAKDAGKSVGIVTTTRVNHATPSAAYAHSADRDWYSDNEMPPEALSQGCKDIAYQLMHNIRDIDVIMGGGRKYMYPKNKTDVEYESDEKARGTRLDGLDLVDTWKSFKPRYKHSHFIWNRTELLTLDPHNVDYLLGLFEPGDMQYELN\n\n### Mutant:\nMISPFLVLAIGTCLTNSLVPEKEKDPKYWRDQAQETLKYALELQKLNTNVAKNVIMFLGDGMGVSTVTAARILKGQLHHNPGEETRLEMDKFPFVALSKTYNTNAQVPDSAGTATAYLCGVKANEGTVGVSAATERSRCNTTQGNEVTSILRWAKDAGKSVGIVTTTRVNHATPSAAYAHSADRDWYSDNEMPPEALSQGCKDIAYQLMHNIRDIDVIMGGGRKYMYPKNKTDVEYESDEKARGTRLDGLDLVDTWKSFKPRYKHSHFIWNRTELLTLDPHNVDYLLGLFEPGDMQYELN\n\n### Answer:", "output": "In HOPS; 16% alkaline of phosphatase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00222", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nARTDNFKLSSLANGLKVATSNTPGHFSALGLYIDAGSRFEGRNLKGCTHILDRLAFKSTEHVEGRAMAETLELLGGNYQCTSSRENLMYQASVFNQDVGKMLQLMSETVRFPKITEQELQEQKLSAEYEIDEVWMKPELVLPELLHTAAYSGETLGSPLICPRGLIPSISKYYLLDYRNKFYTPENTVAAFVGVPHEKALELTGKYLGDWQSTHPPITKKVAQYTGGESCIPPAPVFGNLPELFHIQIGFEGLPIDHPDIYALATLQTLLGGGGGGPGKGMYSRLYTHVLNQYYFVENCVAFNHSYSDSGIFGISLSCIPQAAPQAVEVIAQQMYNTFANKDLRLTEDEVSRAKNQLKSSLLMNLESKLVELEDMGRQVLMHGRKIPVNEMISKIEDLKPDDISRVAEMIFTGNVNNAGNGKGRATVVMQGDRGSFGDVENVLKAYGLGN\n\n### Answer:", "output": "<SEQ_3Di>DPQFQWDWDAFVQQAIEIETDAADQWKKKKKKFFDALQQCPVPLFPLLVLLQVLAQADFPPDHRVRSVVVCVVQPVAWGWDGHLGITMTMGMGGVVCPLVVVQSRLRSLPTRPCDVVSLVVVLVVLLVVLVVQVQDPVRVQVLLLLCQQQVCHRSNPGNRGDNVCSVVNDVVSNVVVSLVISASSRMYMYMYNDDRVVRVVSCCVRPVPTHHSPDDDRDDDGATAAGEEEAEFRDDDDPDAGKKKKKKKWKAEACVDLLNLLVQLLQLLQAWFLDDPPNNCPHNCNVLPVVPDVLWSTKGWDWAHHHWIIITTMMTMGHLQCLQCVLVSRLVVLLLDAPDPVSDDDPVSSVSSLVSVLVVVVVVCVDNRVVGVLQRVVCSRPNDRDDSVVVNVSSVPDDSVSSNVSSCCTFLLNDQTPSGYAQYMHMYIYHPVPSNPPNVVSSVVSNGTD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00223", "category": "Mutation", "instruction": "Describe the effect of mutation O77932-E253A on protein Decapping and exoribonuclease protein.", "input": "### Wild-type:\nMDPRGTKRGAEKTEVAEPRNKLPRPAPSLPTDPALYSGPFPFYRRPSELGCFSLDAQRQYHGDARALRYYSPPPTNGPGPNFDLRDGYPDRYQPRDEEVQERLDHLLCWLLEHRGRLEGGPGWLAEAIVTWRGHLTKLLTTPYERQEGWQLAASRFQGTLYLSEVETPNARAQRLARPPLLRELMYMGYKFEQYMCADKPGSSPDPSGEVNTNVAFCSVLRSRLGSHPLLFSGEVDCTDPQAPSTQPPTCYVELKTSKEMHSPGQWRSFYRHKLLKWWAQSFLPGVPNVVAGFRNPDGFV\n\n### Mutant:\nMDPRGTKRGAEKTEVAEPRNKLPRPAPSLPTDPALYSGPFPFYRRPSELGCFSLDAQRQYHGDARALRYYSPPPTNGPGPNFDLRDGYPDRYQPRDEEVQERLDHLLCWLLEHRGRLEGGPGWLAEAIVTWRGHLTKLLTTPYERQEGWQLAASRFQGTLYLSEVETPNARAQRLARPPLLRELMYMGYKFEQYMCADKPGSSPDPSGEVNTNVAFCSVLRSRLGSHPLLFSGEVDCTDPQAPSTQPPTCYVALKTSKEMHSPGQWRSFYRHKLLKWWAQSFLPGVPNVVAGFRNPDGFV\n\n### Answer:", "output": "Abolishes the decapping activity on incomplete m7G cap mRNAs; when associated with A-236. The present study demonstrates that the mutation in the DXO protein affects its decapping and exoribonuclease activities toward incompletely capped RNAs, only sparing RNAs with a 2'-O-methylated cap structure.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00224", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CC(C)C(C=Cc1ccc(Cl)c(Cl)c1)=NNC(=O)NN=C(C=Cc1ccc(Cl)c(Cl)c1)C(C)C", "output": "The molecule contains the following functional groups: aromatic ring, amide, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "CC(C)C(C=Cc1ccc(Cl)c(Cl)c1)=NNC(=O)NN=C(C=Cc1ccc(Cl)c(Cl)c1)C(C)C", "compound_name": null, "pubchem_cid": null, "MW": 540.32, "LogP": 8.35, "TPSA": 65.85, "QED": 0.26, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "halide"]}}
{"id": "eval_00225", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: Clc1ccc(C(Cl)(Cl)Cl)cn1\nDescriptors: MW=230.91, LogP=3.56, TPSA=12.89, HBD=0, HBA=1", "output": "Descriptors: MW=230.91, LogP=3.56, TPSA=12.89, HBD=0, HBA=1, QED=0.49, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "Clc1ccc(C(Cl)(Cl)Cl)cn1", "compound_name": null, "pubchem_cid": null, "MW": 230.91, "LogP": 3.56, "TPSA": 12.89, "QED": 0.49, "functional_groups": ["halide", "pyridine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 1.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00226", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFENR_POLSJ\n\n### Answer:", "output": "Ferredoxin--NADP reductase, FNR, Fd-NADP(+) reductase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00227", "category": "Mutation", "instruction": "Describe the effect of mutation P38631-S982N on protein 1,3-beta-glucan synthase component FKS1.", "input": "### Wild-type:\nMNTDQQPYQGQTDYTQGPGNGQSQEQDYDQYGQPLYPSQADGYYDPNVAAGTEADMYGQQPPNESYDQDYTNGEYYGQPPNMAAQDGENFSDFSSYGPPGTPGYDSYGGQYTASQMSYGEPNSSGTSTPIYGNYDPNAIAMALPNEPYPAWTADSQSPVSIEQIEDIFIDLTNRLGFQRDSMRNMFDHFMVLLDSRSSRMSPDQALLSLHADYIGGDTANYKKWYFAAQLDMDDEIGFRNMSLGKLSRKARKAKKKNKKAMEEANPEDTEETLNKIEGDNSLEAADFRWKAKMNQLSPLE\n\n### Mutant:\nMNTDQQPYQGQTDYTQGPGNGQSQEQDYDQYGQPLYPSQADGYYDPNVAAGTEADMYGQQPPNESYDQDYTNGEYYGQPPNMAAQDGENFSDFSSYGPPGTPGYDSYGGQYTASQMSYGEPNSSGTSTPIYGNYDPNAIAMALPNEPYPAWTADSQSPVSIEQIEDIFIDLTNRLGFQRDSMRNMFDHFMVLLDSRSSRMSPDQALLSLHADYIGGDTANYKKWYFAAQLDMDDEIGFRNMSLGKLSRKARKAKKKNKKAMEEANPEDTEETLNKIEGDNSLEAADFRWKAKMNQLSPLE\n\n### Answer:", "output": "Decrease in 1144; temperature-sensitive mutant; higher beta-glucan content of cells; when associated with F-872; K-907 and DEL-GSC2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00228", "category": "Mutation", "instruction": "Describe the effect of mutation Q9VS60-H401A on protein Ceramide phosphoethanolamine synthase.", "input": "### Wild-type:\nMCDGEIGDPVTQPRSEGGGLVTMDQETRTHYLDAATDKHLTNGSPDPEPVDPLLVAQWSIENVTSWATCMEHFSRTLLDCLRQEAIDGEVLLSLTEEDVRDMRYKLGYKLTFGELKKFWIAVLKLQLLVKNSSAESVILGIECHGNGNSVYMPLASTGCGPPSSSTCPCPQAECPSYVSDCDTYLRMGGRYVPPEYFKTAMSLGYSFVVTWITSLTMVIVHERVPDMKRYPPLPDIFLDNVPHIPWAFNMCEITGSLLFTIWVVVLTFHKYRLVLLRRFFALAGTVFLLRCVTMLITSLS\n\n### Mutant:\nMCDGEIGDPVTQPRSEGGGLVTMDQETRTHYLDAATDKHLTNGSPDPEPVDPLLVAQWSIENVTSWATCMEHFSRTLLDCLRQEAIDGEVLLSLTEEDVRDMRYKLGYKLTFGELKKFWIAVLKLQLLVKNSSAESVILGIECHGNGNSVYMPLASTGCGPPSSSTCPCPQAECPSYVSDCDTYLRMGGRYVPPEYFKTAMSLGYSFVVTWITSLTMVIVHERVPDMKRYPPLPDIFLDNVPHIPWAFNMCEITGSLLFTIWVVVLTFHKYRLVLLRRFFALAGTVFLLRCVTMLITSLS\n\n### Answer:", "output": "Abolishes CPE synthase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00229", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: HIV protease\nLigand SMILES: O=C(N[C@@H]1N=C(c2ccccc2F)c2cccc3c2N(CC3)C1=O)c1cc2ccccc2[nH]1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "O=C(N[C@@H]1N=C(c2ccccc2F)c2cccc3c2N(CC3)C1=O)c1cc2ccccc2[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 438.46, "LogP": 3.8, "TPSA": 77.56, "QED": 0.51, "functional_groups": ["aromatic ring", "amide", "halide", "indole"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 1.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 1.0, "SR-HSE": 0.0, "SR-MMP": 1.0, "SR-p53": null}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00230", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCc1cccc2c(CCO)c[nH]c12", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring, hydroxyl, indole. Its descriptor profile: MW=189.26, LogP=2.27, TPSA=36.02, HBD=2, HBA=1, QED=0.76.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CCc1cccc2c(CCO)c[nH]c12", "compound_name": null, "pubchem_cid": null, "MW": 189.26, "LogP": 2.27, "TPSA": 36.02, "QED": 0.76, "functional_groups": ["aromatic ring", "hydroxyl", "indole"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": 0.0, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00231", "category": "Mutation", "instruction": "Describe the effect of mutation P14753-F235W on protein Erythropoietin receptor.", "input": "### Wild-type:\nMDKLRVPLWPRVGPLCLLLAGAAWAPSPSLPDPKFESKAALLASRGSEELLCFTQRLEDLVCFWEEAASSGMDFNYSFSYQLEGESRKSCSLHQAPTVRGSVRFWCSLPTADTSSFVPLELQVTEASGSPRYHRIIHINEVVLLDAPAGLLARRAEEGSHVVLRWLPPPGAPMTTHIRYEVDVSAGNRAGGTQRVEVLEGRTECVLSNLRGGTRYTFAVRARMAEPSFSGFWSAFSEPASLLTASDLDPLILTLSLILVLISLLLTVLALLSHRRTLQQKIWPGIPSPESEFEGLFTTHK\n\n### Mutant:\nMDKLRVPLWPRVGPLCLLLAGAAWAPSPSLPDPKFESKAALLASRGSEELLCFTQRLEDLVCFWEEAASSGMDFNYSFSYQLEGESRKSCSLHQAPTVRGSVRFWCSLPTADTSSFVPLELQVTEASGSPRYHRIIHINEVVLLDAPAGLLARRAEEGSHVVLRWLPPPGAPMTTHIRYEVDVSAGNRAGGTQRVEVLEGRTECVLSNLRGGTRYTFAVRARMAEPSFSGFWSAWSEPASLLTASDLDPLILTLSLILVLISLLLTVLALLSHRRTLQQKIWPGIPSPESEFEGLFTTHK\n\n### Answer:", "output": "14-fold more EPO binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00232", "category": "Mutation", "instruction": "Describe the effect of mutation Q968Y9-S547G on protein Insulin-like receptor.", "input": "### Wild-type:\nMTRMNIVRCRRRHKILENLEEENLGPSCSSTTSTTAATEALGTTTEDMRLKQQRSSSRATEHDIVDGNHHDDEHITMRRLRLVKNSRTRRRTTPDSSMDCYEENPPSQKTSINYSWISKKSSMTSLMLLLLFAFVQPCASIVEKRCGPIDIRNRPWDIKPQWSKLGDPNEKDLAGQRMVNCTVVEGSLTISFVLKHKTKAQEEMHRSLQPRYSQDEFITFPHLREITGTLLVFETEGLVDLRKIFPNLRVIGGRSLIQHYALIIYRNPDLEIGLDKLSVIRNGGVRIIDNRKLCYTKTID\n\n### Mutant:\nMTRMNIVRCRRRHKILENLEEENLGPSCSSTTSTTAATEALGTTTEDMRLKQQRSSSRATEHDIVDGNHHDDEHITMRRLRLVKNSRTRRRTTPDSSMDCYEENPPSQKTSINYSWISKKSSMTSLMLLLLFAFVQPCASIVEKRCGPIDIRNRPWDIKPQWSKLGDPNEKDLAGQRMVNCTVVEGSLTISFVLKHKTKAQEEMHRSLQPRYSQDEFITFPHLREITGTLLVFETEGLVDLRKIFPNLRVIGGRSLIQHYALIIYRNPDLEIGLDKLSVIRNGGVRIIDNRKLCYTKTID\n\n### Answer:", "output": "In m596; decreases the number of muscle membrane extensions during larval development.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00233", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKLLVVGASYRTAPVAALERLAVAPADLPHALARLVAQPYVSEAVLVSTCNRVEVYAAVSGFHGGLGDICAVLAESTGVPPAALADHLYVHFDAAAVKHAFRVATGLDSMVVGEAQILGQLRDAYHWASGADTTGRLLHELMQQALRVGKRAHSETGIDRAGQSVVTAALGLATELLNSDLAARPALVVGAGAMGSLGVATLSRLGAGPVAVTNRGADRAVRLAESYGATAVPIADLTATLSTVDIVVAATAAPEAVLTREVVTRALADRKPSRGPLVLLDLAVPRDVEPGVADLPGVQVIDIDRMAALVANGPAAADAAAVERIVATEVDAFLNWLRGADVAPTVAALRGRADNVVTVELGRLAQRRPDLTDDQRDEVARTVHRVVQRLLHQPTVRVRQLAAEPGGDQYAALLRELFDLEVPQTSPVGTVPEVVVPEAVPPLGGAAEDPSTGGQ\n\n### Sequence 2:\nMSVLLVGVSHRSAPVSVLEKVAITDADRPKLIDKMLASSHISEAMIVSTCNRVEVYAVVDAFHGGLAEVGELLTRHSGLAMPDLTKHAYVRYSEAAAEHLFAVASGLDSMVVGEQQVLSQIRGAYATADAQQAVGRTLHELAQHALRVGKRVHSETGIDRAGASVVSVALDRAAQVLG-DLAGRTAVVVGAGAMGGLSVAHLARAGIGRIIVVNRTIERARRLAETAASYGVESSALELDRLHEAMSAADVVLSCTGAVGAVVTLADTHRALADRDRAEFPAGDRPLVFCDLGLPRDVEPAVAGLPGVAVIDIESLQRDPAAGAAADDTAAARSIVAEELAKYLAGQRMAEVTPTVAALRQRAAEVVEAELLRLDSRLPGLAEPERDEVARTVRRVVDKLLHAPTVRVKQLASTPGGDSYAEALRELFELK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00234", "category": "Mutation", "instruction": "Describe the effect of mutation P20591-A640R on protein Interferon-induced GTP-binding protein Mx1.", "input": "### Wild-type:\nMVVSEVDIAKADPAAASHPLLLNGDATVAQKNPGSVAENNLCSQYEEKVRPCIDLIDSLRALGVEQDLALPAIAVIGDQSSGKSSVLEALSGVALPRGSGIVTRCPLVLKLKKLVNEDKWRGKVSYQDYEIEISDASEVEKEINKAQNAIAGEGMGISHELITLEISSRDVPDLTLIDLPGITRVAVGNQPADIGYKIKTLIKKYIQRQETISLVVVPSNVDIATTEALSMAQEVDPEGDRTIGILTKPDLVDKGTEDKVVDVVRNLVFHLKKGYMIVKCRGQQEIQDQLSLSEALQREK\n\n### Mutant:\nMVVSEVDIAKADPAAASHPLLLNGDATVAQKNPGSVAENNLCSQYEEKVRPCIDLIDSLRALGVEQDLALPAIAVIGDQSSGKSSVLEALSGVALPRGSGIVTRCPLVLKLKKLVNEDKWRGKVSYQDYEIEISDASEVEKEINKAQNAIAGEGMGISHELITLEISSRDVPDLTLIDLPGITRVAVGNQPADIGYKIKTLIKKYIQRQETISLVVVPSNVDIATTEALSMAQEVDPEGDRTIGILTKPDLVDKGTEDKVVDVVRNLVFHLKKGYMIVKCRGQQEIQDQLSLSEALQREK\n\n### Answer:", "output": "Sequesters viral nucleoproteins, has antiviral activity. The mutation in the MxA protein has been mentioned in the document. However, the functional changes caused by this mutation are not described.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00235", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nMZB1:high SDC1:high JCHAIN:medium CD38:high XBP1:low\n</cell>", "output": "This cell is most consistent with plasma cell, an antibody-secreting plasma cell. Key markers such as SDC1, MZB1, XBP1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasma cell", "tissue": "lymph node", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MZB1", "SDC1", "JCHAIN", "CD38", "XBP1"]}}
{"id": "eval_00236", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nFXYD6_MOUSE\n\n### Answer:", "output": "Membrane; Single-pass type I membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00237", "category": "Mutation", "instruction": "Describe the effect of mutation Q9EPH0-R571Q on protein Prestin.", "input": "### Wild-type:\nMDHAEENEIPAETQKYLVERPIFSHPVLQERLHVKDKVTDSIGDKLKQAFTCTPKKVRNIIYMFLPITKWLPAYKFKEYVLGDLVSGISTGVLQLPQGLAFAMLAAVPPVFGLYSSFYPVIMYCFFGTSRHISIGPFAVISLMIGGVAVRLVPDDIVIPGGVNATNGTEARDALRVKVAMSVTLLSGIIQFCLGVCRFGFVAIYLTEPLVRGFTTAAAVHVFTSMLKYLFGVKTKRYSGIFSVVYSTVAVLQNVKNLNVCSLGVGLMVFGLLLGGKEFNERFKEKLPAPIPLEFFAVVMG\n\n### Mutant:\nMDHAEENEIPAETQKYLVERPIFSHPVLQERLHVKDKVTDSIGDKLKQAFTCTPKKVRNIIYMFLPITKWLPAYKFKEYVLGDLVSGISTGVLQLPQGLAFAMLAAVPPVFGLYSSFYPVIMYCFFGTSRHISIGPFAVISLMIGGVAVRLVPDDIVIPGGVNATNGTEARDALRVKVAMSVTLLSGIIQFCLGVCRFGFVAIYLTEPLVRGFTTAAAVHVFTSMLKYLFGVKTKRYSGIFSVVYSTVAVLQNVKNLNVCSLGVGLMVFGLLLGGKEFNERFKEKLPAPIPLEFFAVVMG\n\n### Answer:", "output": "Shifts the voltage-sensitivity to slightly more positive values; when associated with Q-572 and Q-577.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00238", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nCSF1R:high CX3CR1:low P2RY12:low\n</cell>", "output": "This cell is most consistent with microglia, a brain microglia cell. Key markers such as CX3CR1, CSF1R, P2RY12 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "blood", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CSF1R", "CX3CR1", "P2RY12"]}}
{"id": "eval_00239", "category": "Mutation", "instruction": "Describe the effect of mutation Q95XA8-S179A on protein CREB-regulated transcription coactivator 1 homolog.", "input": "### Wild-type:\nMSNSNTPRKFSEKIAILERKQNEENTTFEDIMRQVQSITHHPTDSSGSSTATAPMPIPQQGLLPPQQPWGHNLGGSLPNVHQMPSYSPPQWPPNWQHEIQHRPIQGHRSRSPEDHMIGSASGSPSHHYHPYGMRSNGLSRSPDRTPPQHPQYTPYGPPYNQPGQLVPPESWNQINRARSDPAIHNMGGMVPMHPHHHQQQQMAFHQMPHYLQNSMPGPSGMMAPNSQSQQHSPQLTPQGSQQGSPVQMHHQIPPPLQMGNNQQMGGGNNGMSPLQSPNHMMTPMYGYHNGSPLHSPMDSP\n\n### Mutant:\nMSNSNTPRKFSEKIAILERKQNEENTTFEDIMRQVQSITHHPTDSSGSSTATAPMPIPQQGLLPPQQPWGHNLGGSLPNVHQMPSYSPPQWPPNWQHEIQHRPIQGHRSRSPEDHMIGSASGSPSHHYHPYGMRSNGLSRSPDRTPPQHPQYTPYGPPYNQPGQLVPPESWNQINRARADPAIHNMGGMVPMHPHHHQQQQMAFHQMPHYLQNSMPGPSGMMAPNSQSQQHSPQLTPQGSQQGSPVQMHHQIPPPLQMGNNQQMGGGNNGMSPLQSPNHMMTPMYGYHNGSPLHSPMDSP\n\n### Answer:", "output": "Loss of phosphorylation. Leads to nuclear sequestration; when associated with A-76. CRTC-1 is a direct AMPK target and interacts with the CREB homologue-1 (CRH-1) transcription factor in vivo. The pro-longevity effects of activating AMPK decrease CRTC-1 and CRH-1 activity and induce transcriptional responses similar to those of CRH-1 null worms. Downregulation of crtc-1 increases lifespan in a crh-1-dependent manner, substantiating a role for CRTCs and CREB in ageing.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00240", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCYB_LEPDO\n\n### Answer:", "output": "Cytochrome b, Complex III subunit 3, Complex III subunit III, Cytochrome b-c1 complex subunit 3, Ubiquinol-cytochrome-c reductase complex cytochrome b subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00241", "category": "Mutation", "instruction": "Describe the effect of mutation P97275-K419R on protein Alkyldihydroxyacetonephosphate synthase, peroxisomal.", "input": "### Wild-type:\nMAEAAAAAAAAAAAGETSASSGSAAERDPDQDRAGRRLRVLSGHLLGRPQEALSTNECKARRAASAATAAPTATPAAPESGIIPKKRQELMKWNGWGYNDSKFFLNKKGQLELTGKRYPLSGVALPTFKDWIQNTFGINLDHKTTSKASLNPSDTPPSIVNEDFLHELKKTNISYSQEADDRVFRAHGHCLHEIFLLREGMFERIPDIVLWPTCHDDVVKIVNLACKYNLCIIPIGGGTSVSYGLMCPADETRTIISLDTSQMNRILWVDENNLTAHVEAGITGQELERQLKESGYCTGH\n\n### Mutant:\nMAEAAAAAAAAAAAGETSASSGSAAERDPDQDRAGRRLRVLSGHLLGRPQEALSTNECKARRAASAATAAPTATPAAPESGIIPKKRQELMKWNGWGYNDSKFFLNKKGQLELTGKRYPLSGVALPTFKDWIQNTFGINLDHKTTSKASLNPSDTPPSIVNEDFLHELKKTNISYSQEADDRVFRAHGHCLHEIFLLREGMFERIPDIVLWPTCHDDVVKIVNLACKYNLCIIPIGGGTSVSYGLMCPADETRTIISLDTSQMNRILWVDENNLTAHVEAGITGQELERQLKESGYCTGH\n\n### Answer:", "output": "Strongly increased activity. The mutation in alkyl-dihydroxyacetonephosphate synthase results in a change in enzyme activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00242", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMURB_STRP2\n\n### Answer:", "output": "UDP-N-acetylenolpyruvoylglucosamine reductase, UDP-N-acetylmuramate dehydrogenase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00243", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMPEETLLAFDFGEKKIGIAIGNTLTRQARPLEIIFSEIREARFARIGQLLQQWQPQRVVVGLALASDGGEQPATARCRRFANQLRGRYGLAVELVDERGSSMEAQRLLGTHAPDDAVAAAVILQRYLDALP\n\n### Sequence 2:\nMNQLQNLAEALSASSPVSGTVQTVVDGNTTTTTTTNTGLGVVAVGAGLAMIGAIGSGLGQGYAAGKTVEAVGRNPEMISKIRATFIIGAGIAETASIYSFIVALLLIFVGK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00244", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nKAD_ACTPJ\n\n### Answer:", "output": "Catalyzes the reversible transfer of the terminal phosphate group between ATP and AMP.  Plays an important role in cellular energy homeostasis and in adenine nucleotide metabolism.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00245", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "Nc1nc(F)nc2c1ncn2C1OC(CO)C(O)C1O", "output": "The SMILES represents an unnamed molecule with MW=285.24, LogP=-1.84, TPSA=139.54. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "SIDER", "canonical_smiles": "Nc1nc(F)nc2c1ncn2C1OC(CO)C(O)C1O", "compound_name": null, "pubchem_cid": null, "MW": 285.24, "LogP": -1.84, "TPSA": 139.54, "QED": 0.47, "functional_groups": ["hydroxyl", "ether", "halide"], "moleculenet_labels": {"Hepatobiliary disorders": 1, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 0, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 1, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 1, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 1, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00246", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSVNDLRPGTTFLYDGNIYLVLEQAFSKTGRQQGKVTVKAKNMRTGARVELTFTGGEKVDKAMIERKEMQYLYNDGNDAYLMNTETYEQVSIPMTRLEWEKNFLVDGLMINMTEFENEVLGIDLPVKVELTVVEAEAAVKGDTTSGAQKKAILETGLEIMVPLFVNQGTKIIVSSADGKYVGRA\n\n### Sequence 2:\nINVNEFKPGITFEDEGNIYVVLTAQHSKQGRGQANVKAKVKNLRTGSTTLKSYTGGDKVQKAHIEKKPMDYLYNDGSNIILMDQESFEQIEIDVKKVEWELNFLTEGMKILVRQYQNEILDIEIPINIELKVINAPDAVKGNTTTNPQKKVIVETGYELEVPMFIKEGETIIVSSETGKYGGKS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00247", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSYL_ALIF1\n\n### Answer:", "output": "Belongs to the class-I aminoacyl-tRNA synthetase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00248", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMVKIEVGSVGDSFSVSSLKAYLSEFIATLLFVFAGVGSAVAFAKLTSDGALDPAGLVAIAIAHAFALFVGVSIAANISGGHLNPAVTLGLAIGGNITLITGFFYWIAQCLGSIVACLLLVFVTNGKSVPTHGVSAGLGAVEGVVMEIVVTFALVYTVYATAADPKKGSLGTIAPIAIGFIVGANILAAGPFSGGSMNPARSFGPAVVSGDLSQIWIYWVGPLVGGALAGLIYGDVFIGSYEAVETREIRV\n\n### Sequence 2:\nMVKLAFGSFRDSLSAASLKAYVAEFIATLLFVFAGVGSAIAYSQLTKGGALDPAGLVAIAIAHAFALFVGVSMAANISGGHLNPAVTFGLAVGGHITILTGILYWVAQLLGASVACFLLQYVTHGQAIPTHGVS-GISEIEGVVMEIVITFALVYTVYATAADPKKGSLGTIAPMAIGFIVGANILAAGPFSGGSMNPARSFGPAVAAGNFAGNWVYWVGPLVGGGLAGLVYGDVFIASYQPVGQQE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00249", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTLFSSISSMSTSMSGSKSSISSFGSGTSMGSNSIACGGCGGSGGILGSGLGLGLGLGLDLTGGSRSRGACGGNRGNGNGNGGMGGGNGSCCGGPCCGI\n\n### Sequence 2:\nMDRLSFLTFIMLILASYRLTHLIVFDKITEFIRKPFMKKKRIVDQNGHVDEKSVPASNFGYMLNCYWCAGVWCAILIGLGYLFLPRIAIPLIFILAIAGAQAILETAVGVGVKLIDVLKSLQTMMNDKKS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00250", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSMTDLLNAEDIKKAVGAFSAIDSFDHKKFFQMVGLKKKSADDVKKVFHILDKDKSGFIEEDELGFILKGFSPDARDLSAKETKTLMAAGDKDGDGKIGVDEFSTLVAES\n\n### Sequence 2:\nMNSLSILNNIEDKVVEAINTAALSLESLSASLDIENTNENFSKFQTQSDKFYNLVKKDIHKGLIDFIDSMTDIAPFDHSSYLKKSELEVSHNFTEIILSHLEDLNNIVENNQEKQEKEKQEKEKLEKEKLEKEKQQSNEMNID\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00251", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "PECAM1, CDH5, KDR, CLDN5, VWF", "output": "This profile is most consistent with endothelial cell, which is commonly found in colon under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "endothelial cell", "tissue": "colon", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["PECAM1", "CDH5", "KDR", "CLDN5", "VWF"]}}
{"id": "eval_00252", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPIVQNLQGQMVHQAISPRTLNAWVKVVEEKAFSPEVIPMFSALSEGATPQDLNTMLNTVGGHQAAMQMLKETINEEAAEWDRLHPVHAGPIEPGQMREPRGSDIAGTTSTLQEQIGWMTHNPPIPVGEIYKRWIILGLNKIVRMYSPTSILDIRQGPKEPFRDYVDRFYKTLRAEQASQEVKNWMTETLLVQNANPDCKTILKALGPAATLEEMMTACQGVGGPGHKARVL\n\n### Answer:", "output": "<SEQ_2D>CEESCTTTCCEECCCCHHHHHHHHHHHHHHTTSTTHHHHHHHHTTTCCHHHHHHHHHHCCSCHHHHHHHHHHHHHHHHHHHHHCCCCSCCCCTTCCCCCCHHHHTTSSCCHHHHHHHHHSSSCCCHHHHHHHHHHHHHHHHHHHHSCCCSTTCCCCSSCCHHHHHHHHHHHHHHSSCCHHHHHHHHHHTHHHHSCHHHHHHHHHHCTTCCHHHHHSCSCCCCCSSTTSCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00253", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSMFRTLQRQPRTISLFTHDLENSRPCLSILEYLKSHTTNRFDLELSTKFPTLDQVHYMNAINPMILRAQIPHLTKIMKLKSYDPLFGSQLSDCVTKGFWNKEAPLWVDWEKKALGTDLQSIKELLEKD\n\n### Sequence 2:\nMCESSNKTENDIVRLSQAMDVLAKLIISKQKDGSQLQVEYEHKLKELEKFINLLLGLHESTVGSMMNTSVLDMVLRNGIEIMEKDDQKYALIPIKAKEEADKTTSTIQGVTSKKSSKKKKNKIKCSFCHEAGHTRAHCGARLTVIPKK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00254", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "S100A4, ACTA2, COL1A1, FAP", "output": "The expression pattern is consistent with cancer-associated fibroblast in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "lung", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["S100A4", "ACTA2", "COL1A1", "FAP"]}}
{"id": "eval_00255", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: COc1ccc(OC)c(-c2c(F)ccc3c(N)c(C(=O)NC(C)C)nnc23)c1\nDescriptors: MW=384.41, LogP=3.17, TPSA=99.36, HBD=2, HBA=6", "output": "Descriptors: MW=384.41, LogP=3.17, TPSA=99.36, HBD=2, HBA=6, QED=0.7, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "COc1ccc(OC)c(-c2c(F)ccc3c(N)c(C(=O)NC(C)C)nnc23)c1", "compound_name": null, "pubchem_cid": null, "MW": 384.41, "LogP": 3.17, "TPSA": 99.36, "QED": 0.7, "functional_groups": ["aromatic ring", "amide", "ether", "halide"], "moleculenet_labels": {"label": 3.77}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00256", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDPCKG_METAR\n\n### Answer:", "output": "Catalyzes the GTP-dependent phosphorylation of the 3'-hydroxyl group of dephosphocoenzyme A to form coenzyme A (CoA).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00257", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSILSLVGLGISKKFITENAIDTLNNSDIIIFDKYTSRSCDINVDVLRRLVKGGKTLIEADRSLLENNSKIIMDYLDKNYNVSIASIGDVLIATTHVSLLIEAKQRGHNVKVIPGISVHCYLISKSLLSSYKFGKSVTVTFPYNDFIDPTPYNVIKDNKERGLHTILYLDLKSEKAMTANEALQILLRLEDKHRKNVLSKSDIVIVGARLGCDDEKIVALTVEEATLYDFGNTPHIIIIPGNLHYMEADAIKWMLMS\n\n### Sequence 2:\nMSILSLVGLGISKKFITENAIDTLNNSDIIIFDKYTSRSCDINVDVLRRLVKGGKTLIEADRSLLENNSKIIMDYLDKNYNVSIASIGDVLIATTHVSLLIEAKQRGHNVKVIPGISVHCYLISKSLLSSYKFGKSVTVTFPYNDFIDPTPYNVIKDNKERGLHTILYLDLKSEKAMTANEALQILLRLEDKHRKNVLSKSDIVIVGARLGCDDEKIVALTVEEATLYDFGNTPHIIIIPGNLHYMEADAIKWMLMS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00258", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL15_STAA2\n\n### Answer:", "output": "50S ribosomal protein L15.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00259", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTIQPGTGYNNGYFYSYWNDGHGGVTYTNGPGGQFSVNWSNSGDFVGGKGWQPGTKNKVINFSGSYNPNGNSYLSVYGWSRNPLIEYYIVENFGTYNPSTGATKLGEVTSDGSVYDIYRTQRVNQPSIIGTATFYQYWSVRRNHRSSGSVNTANHFNAWAQQGLTLGTMDYQIVAVQGYFSSGSASITVS\n\n### Answer:", "output": "<SEQ_3Di>DDDAAWDDDQHKTKGKDFPPPDFKDKDQDHAQKIKIFGDQGAWIKIGIHDFFDDQFDKKWKDWDWWWFAWKFWWFWFFFVPPTETETETQDTYPDDVCVVFDWQAWFDDFHFIWTKGKDWDQQDQGPVGGHIHMYIYTHTPHHDRTGIGRNNRVQVSCVVSVNDGHHTGHIGGMMIGHRTGIMIMMGMD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00260", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRUVB_BACAH\n\n### Answer:", "output": "Holliday junction branch migration complex subunit RuvB.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00261", "category": "Mutation", "instruction": "Describe the effect of mutation A0A0A7HIF0-A100D on protein CRISPR system Cms endoribonuclease Csm3.", "input": "### Wild-type:\nMTFAKIKFSAQIRLETGLHIGGSDAFAAIGAIDSPVIKDPITNIPIIPGSSLKGKMRTLLAKVYNEKVAEKPSDDSDILSRLFGNSKDKRFKMGRLIFRAAFLSNADELDSLGVRSYTEVKFENTIDRITAEANPRQIERAIRNSTFDFELIYEITDENENQVEEDFKVIRDGLKLLELDYLGGSGSRGYGKVAFEKLKATTVFGNYDVKTLNELLTAEV\n\n### Mutant:\nMTFAKIKFSAQIRLETGLHIGGSDAFAAIGAIDSPVIKDPITNIPIIPGSSLKGKMRTLLAKVYNEKVAEKPSDDSDILSRLFGNSKDKRFKMGRLIFRDAFLSNADELDSLGVRSYTEVKFENTIDRITAEANPRQIERAIRNSTFDFELIYEITDENENQVEEDFKVIRDGLKLLELDYLGGSGSRGYGKVAFEKLKATTVFGNYDVKTLNELLTAEV\n\n### Answer:", "output": "Enhanced degradation of target ssRNA by the Csm complex, crRNA is longer, Csm complex remains unchanged.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00262", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "O=NN(c1ccccc1)c1ccccc1", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains aromatic ring. Its descriptor profile: MW=198.22, LogP=3.51, TPSA=32.67, HBD=0, HBA=2, QED=0.56.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=NN(c1ccccc1)c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 198.22, "LogP": 3.51, "TPSA": 32.67, "QED": 0.56, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 1.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00263", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "O=C(O)c1ccc[n+]([O-])c1", "output": "The SMILES represents an unnamed molecule with MW=139.11, LogP=0.02, TPSA=64.24. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "O=C(O)c1ccc[n+]([O-])c1", "compound_name": null, "pubchem_cid": null, "MW": 139.11, "LogP": 0.02, "TPSA": 64.24, "QED": 0.44, "functional_groups": ["hydroxyl", "pyridine"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": 0.0, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00264", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMNMG_SYNP2\n\n### Answer:", "output": "NAD-binding protein involved in the addition of a carboxymethylaminomethyl (cmnm) group at the wobble position (U34) of certain tRNAs, forming tRNA-cmnm(5)s(2)U34.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00265", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTPAGKRILVMAGGTGGHVFPALAVAKYLAQQGWQVRWLGTADRMEARLVPQYGFDIDFIDIKGVRGNGLIRKLAAPFKVVRSILQAKAVIAEFKPDVVLGMGGFASGPGGVAARLAGIPLVLHEQNAIPGMTNKLLSRIATQVLCAFKNTFTTVKAKVVGNPIRQELIALGAQPKPEADKALKVLVVGGSLGAKVFNDLMPEAVAILSQQQSVTVWHQVGKDNLAGVKAAYQQHGQDGGVNIAEFIDDMEAAYRWADVVLCRAGALTVSELAAVGLPSILVPYPHAVDDHQTRNGQVLVEAGAAFLLPQAILDVNKLAGKLQLLANDRTELARMGQRARDVAVLDATEQVAAVCISLAEKG\n\n### Sequence 2:\nMAFVASVPVFANASGLKTEAKVCQKPALKNSFFRGEEVTSRSFFASQAVSAKPATTFEVDTTIRAQAVDAKKKGDIPLNLFRPANPYIGKCIYNERIVGEGAPGETKHIIFTHEGKVPYLEGQSIGIIPPGTDKDGKPHKLRLYSIASTRHGDFGDDKTVSLSVKRLEYTDANGNLVKGVCSNYLCDLKPGDEVMITGPVGTTMLMPEDQSATIIMLATGTGIAPFRSFLRRMFEETHADYKFNGLAWLFLGVPTSSTLLYREELEKMQKANPNNFRLDYAISREQTDSKGEKMYIQNRIAEYANEFWNMIQKPNTFVYMCGLRGMEDGIQQCMEDIAKANGTTWDAVVKGLKKEKRWHVETY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00266", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "HBA1, HBB, GATA1", "output": "The expression pattern is consistent with erythroid progenitor in a context associated with COVID-19, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "lung", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["HBA1", "HBB", "GATA1"]}}
{"id": "eval_00267", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CCC(=O)Nc1ccc(O)cc1", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains aromatic ring, amide, hydroxyl, phenol. Its descriptor profile: MW=165.19, LogP=1.74, TPSA=49.33, HBD=2, HBA=2, QED=0.66.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "CCC(=O)Nc1ccc(O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 165.19, "LogP": 1.74, "TPSA": 49.33, "QED": 0.66, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "phenol"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00268", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: Clc1ccc2c(c1)Oc1cc(Cl)c(Cl)c(Cl)c1O2\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "FreeSolv", "canonical_smiles": "Clc1ccc2c(c1)Oc1cc(Cl)c(Cl)c(Cl)c1O2", "compound_name": null, "pubchem_cid": null, "MW": 321.97, "LogP": 6.2, "TPSA": 18.46, "QED": 0.46, "functional_groups": ["aromatic ring", "ether", "halide"], "moleculenet_labels": {"label": -3.84}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00269", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "SLC32A1, GAD2, GAD1", "output": "The expression pattern is consistent with inhibitory neuron in a context associated with type 2 diabetes, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "pancreas", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["SLC32A1", "GAD2", "GAD1"]}}
{"id": "eval_00270", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSINKIFEESWKNRLTKYEIARIISARALQLAMGASPLIDTNNLPFNDVISIAEEELKRGVLPITVRRVYPNGKIELVSVKKVEFK\n\n### Sequence 2:\nLHFNEVFVSLWQNKLTRYEIARVISARALQLAMGAPALIDINNISSTDVISIAEEEFKRGVLPITIRRRLPNGKIILLSLRK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00271", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nGATC_STRP2\n\n### Answer:", "output": "Belongs to the GatC family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00272", "category": "Mutation", "instruction": "Describe the effect of mutation A0A1D5NS60-C428F on protein Zinc finger protein 16-like.", "input": "### Wild-type:\nMSRKRNHCYMETGASSESQGAFVDSAGPFSRDEEDFSELEPDEQLVCSVTEITEHLGRNITVVLESALSEIRKLVGVRIRVLKMELREKSDEIELLKAKLESAEKDGRVSNFSSLDFRKSEHQKYGAEPKKAKTGTPVVKKENINAICDYLMKDKNQRGAAEVESDHSNQAFGSERDVRTEAQPHGSLSLWPDSGPADTDAETDIFSMLPSASKRMYDYEWMTGVELNSAEFKGDSETKCEDVPPMDEEDENEDSEEGRGSLRSVSDHFPLDTQGSPGEDRSSPAEDSMDRMEPGQQFTS\n\n### Mutant:\nMSRKRNHCYMETGASSESQGAFVDSAGPFSRDEEDFSELEPDEQLVCSVTEITEHLGRNITVVLESALSEIRKLVGVRIRVLKMELREKSDEIELLKAKLESAEKDGRVSNFSSLDFRKSEHQKYGAEPKKAKTGTPVVKKENINAICDYLMKDKNQRGAAEVESDHSNQAFGSERDVRTEAQPHGSLSLWPDSGPADTDAETDIFSMLPSASKRMYDYEWMTGVELNSAEFKGDSETKCEDVPPMDEEDENEDSEEGRGSLRSVSDHFPLDTQGSPGEDRSSPAEDSMDRMEPGQQFTS\n\n### Answer:", "output": "In st78; gross morphology is normal. Expression of myelin basic protein (mbp) in the central nervous system (CNS) is strongly reduced, whereas mbp expression in the peripheral nervous system is unaffected. Expression of sox10 and olig2 in oligodendrocyte precursor cells (OPCs) is significantly reduced. Migration of OPCs is impaired with delayed onset and aberrant pathfinding behavior. Myelination in the CNS is reduced at 5 and 9 days post-fertilization. Double knockouts with notch3 have a more severe phenotype.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00273", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVNYAINQFKNGLKLILDGNPCSIVNNEIVKPGKGQTFNRVKFKDLITGKTLIKTFKSGETLEGADVMELDLQYLYNDGNTWNFMDLDSFEQYTIDNATMNDAKGYLVEQDMCTVTLWNDNPISVIPPNHVILEVLNTDPGLKGDTAGTGGKPATMNTGVVVQVPLFVDIGDKVKVDTRTNEYVGRA\n\n### Sequence 2:\nMSEEKKDEILEQETVETKEEIKTEEAEQKTESLEEKVARLESELKESEEKFLRAYADFENMKKRLEKEKYQAIDYASEKFAKDLLTPLDTLEMALNSAKADVDANELLEKLKEGIELTLKNFITTFEKHNITKVETDGEFDPNVHNAVMQVDSAEHNSGQIVQELQKGYVLKDRLLRPSMVSIAN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00274", "category": "Mutation", "instruction": "Describe the effect of mutation O25119-Q325C on protein Flagellar motor switch protein FliG.", "input": "### Wild-type:\nMATKLTPKQKAQLDELSMSEKIAILLIQVGEDTTGEILRHLDIDSITEISKQIVQLNGTDKQIGAAVLEEFFAIFQSNQYINTGGLEYARELLTRTLGSEEAKKVMDKLTKSLQTQKNFAYLGKIKPQQLADFIINEHPQTIALILAHMEAPNAAETLSYFPDEMKAEISIRMANLGEISPQVVKRVSTVLENKLESLTSYKIEVGGLRAVAEIFNRLGQKSAKTTLARIESVDNKLAGAIKEMMFTFEDIVKLDNFAIREILKVADKKDLSLALKTSTKDLTDKFLNNMSSRAAEQFVE\n\n### Mutant:\nMATKLTPKQKAQLDELSMSEKIAILLIQVGEDTTGEILRHLDIDSITEISKQIVQLNGTDKQIGAAVLEEFFAIFQSNQYINTGGLEYARELLTRTLGSEEAKKVMDKLTKSLQTQKNFAYLGKIKPQQLADFIINEHPQTIALILAHMEAPNAAETLSYFPDEMKAEISIRMANLGEISPQVVKRVSTVLENKLESLTSYKIEVGGLRAVAEIFNRLGQKSAKTTLARIESVDNKLAGAIKEMMFTFEDIVKLDNFAIREILKVADKKDLSLALKTSTKDLTDKFLNNMSSRAAEQFVE\n\n### Answer:", "output": "No mobility shift in SDS-PAGE nor reduction in the fluorescence intensity upon 5-iodoacetamidofluorescein (5-IAF) haloalkylation after cysteine cross-linking. A mobility shift and reduction in the fluorescence intensity after cysteine cross-linking suggest formation of disulfide bonds, hence closeness of corresponding wild-type residues; when associated either with C-209; C-217; C-222 or C-243.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00275", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMATIFDILNSINNKNNNHSNYYSCKRLKSNNNNNNNNFRNNSIKRIDIIPSMDVTITDNNIIIETELAGISKDNIEIDIKDSILTIQGEKKINNNYKQQQQQHSDKTKTNNELPLIEENKKEFRKYLSERSFGNFKRCLDLTSILYQLDLSTIKSNFENGLLIITINKKSDLSNSSFKININ\n\n### Sequence 2:\nMRRAEVQRTTLETMINVEFTMDGNGEFSGTSGLGFFDHMLTLFCKFGGFNLNLSCQGDLEVDDHHTVEDIGLVLGEAFSKALGDKKGIARFASGFYPMDEALMLIAVDISGRPYLAFEADFPPVLAGKFNYQMVEEFLRGFVQKAGITLHVREISGKNLHHKAEAIFKGLGRTLKVAVSVQGDELPSTKGVI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00276", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPP2BB_HUMAN\n\n### Answer:", "output": "Calcium-dependent, calmodulin-stimulated protein phosphatase which plays an essential role in the transduction of intracellular Ca(2+)-mediated signals (PubMed:19154138, PubMed:25720963, PubMed:32753672, PubMed:26794871).  Dephosphorylates TFEB in response to lysosomal Ca(2+) release, resulting in TFEB nuclear translocation and stimulation of lysosomal biogenesis (PubMed:25720963, PubMed:32753672).  Dephosphorylates and activates transcription factor NFATC1 (PubMed:19154138).  Dephosphorylates and inactivates transcription factor ELK1 (PubMed:19154138).  Dephosphorylates DARPP32 (PubMed:19154138).  Negatively regulates MAP3K14/NIK signaling via inhibition of nuclear translocation of the transcription factors RELA and RELB (By similarity).  May play a role in skeletal muscle fiber type specification (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00277", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: NC(=O)c1cc[n+](CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)Cc4cccs4)[C@H]3SC2)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "NC(=O)c1cc[n+](CC2=C(C(=O)[O-])N3C(=O)[C@@H](NC(=O)Cc4cccs4)[C@H]3SC2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 458.52, "LogP": -1.22, "TPSA": 136.51, "QED": 0.39, "functional_groups": ["amide", "pyridine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 1.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00278", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nDVSGTVCLSALPPEATDTLNLIASDGPFPYSQDGVVFQNRESVLPTQSYGYYHEYTVITPGARTRGTRRIITGEATQEDYWTGDHYATFSLIDQTC\n\n### Answer:", "output": "<SEQ_2D>CCCCEEEGGGSCHHHHHHHHHHHHTCCCSSTTTTEECCCTTCCSCCCCTTSCEEEECCCTTCSSSCSCEEEECSSTTCEEEESSTTSCCEEEETTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00279", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nTNNI3 MYH7 TNNT2 MYH6\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as TNNI3, MYH6, MYH7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "spleen", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TNNI3", "MYH7", "TNNT2", "MYH6"]}}
{"id": "eval_00280", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMARDIVADFLSHLTNVRRKGEDKTAFFPVSNLIIKLVEIMKREGYIKNYRLVDSSRGKIIEIELSEYFNEANAIKPRFPVEYSELEKYEKRYLPALNFGRLILSTSKGLITNREAKEQKIGGVLIAYVY\n\n### Sequence 2:\nDPLANALSHITNSERVGKREVYIKPASKLIGEVLRVMQKYGYIGEFEFIDDGRAGVYRVQLLGRINKAGAIKPRFPVKVSEFEKWEKRFLPAFEFGILIVSTSQGVMSHKEAIEKGIGGRLIAYVY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00281", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMNQEKVLKTLLMPIVSEKMTMLSANNQYAFKVRMDSSKREIKAAVEILLGVNVENVTTLIVKGKKKIFKGRTGSRPNWKKAMVKVSAGQIIDVNRT\n\n### Sequence 2:\nMSKERLLKVLLAPHVSEKSALLADASEQYIFKVVPNATKPEVKQAVESLFDVKVQSVNMINIKGKTKVFKGRVGKRNGLRKAIVRLAPGQEID\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00282", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nSLC17A7:low RBFOX3:medium SATB2:medium\n</cell>", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as SLC17A7, SATB2, RBFOX3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "intestine", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SLC17A7", "RBFOX3", "SATB2"]}}
{"id": "eval_00283", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKCPFCGAEDTAVADTRLNDEADVVRRRRKCNACDKRFTTYERAEIQLPQVVKKNGLRTEFSRAKLRASLELALRKRPVSIESVDAAVADIEERLLSAGEREVTTQQLGELVMRELKKLDKVAYIRFASVYRNFEDVDAFSRAIREVSPAAKKK\n\n### Sequence 2:\nMKCPFCGNADTQVVDSRVSEEGDTIRRRRRCLSCDKRFTTYERIELAMPSVVKRNGSRSDYDTAKLRASLSLALRKRPVSTDQVDSVVARIEETLLASGQREVSTERIGELVMAELKKLDKVGYVRFASVYKNFEDIGEFVDAIREM\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00284", "category": "Mutation", "instruction": "Describe the effect of mutation Q99ZW2-H840A on protein CRISPR-associated endonuclease Cas9/Csn1.", "input": "### Wild-type:\nMDKKYSIGLDIGTNSVGWAVITDEYKVPSKKFKVLGNTDRHSIKKNLIGALLFDSGETAEATRLKRTARRRYTRRKNRICYLQEIFSNEMAKVDDSFFHRLEESFLVEEDKKHERHPIFGNIVDEVAYHEKYPTIYHLRKKLVDSTDKADLRLIYLALAHMIKFRGHFLIEGDLNPDNSDVDKLFIQLVQTYNQLFEENPINASGVDAKAILSARLSKSRRLENLIAQLPGEKKNGLFGNLIALSLGLTPNFKSNFDLAEDAKLQLSKDTYDDDLDNLLAQIGDQYADLFLAAKNLSDAI\n\n### Mutant:\nMDKKYSIGLDIGTNSVGWAVITDEYKVPSKKFKVLGNTDRHSIKKNLIGALLFDSGETAEATRLKRTARRRYTRRKNRICYLQEIFSNEMAKVDDSFFHRLEESFLVEEDKKHERHPIFGNIVDEVAYHEKYPTIYHLRKKLVDSTDKADLRLIYLALAHMIKFRGHFLIEGDLNPDNSDVDKLFIQLVQTYNQLFEENPINASGVDAKAILSARLSKSRRLENLIAQLPGEKKNGLFGNLIALSLGLTPNFKSNFDLAEDAKLQLSKDTYDDDLDNLLAQIGDQYADLFLAAKNLSDAI\n\n### Answer:", "output": "Target DNA complementary to the crRNA is not cleaved; nickase activity. In vivo, loss of Cas9-mediated CRISPR interference in plasmid transformation. Able to process and bind guide RNAs and target DNA but not cleave DNA; when associated with A-10.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00285", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMELQDPKMNGALPSDAVGYRQEREGFLPSRGPAPGSKPVQFMDFEGKTSFGMSVFNLSNAIMGSGILGLAYAMAHTGVIFFLALLLCIALLSSYSIHLLLTCAGIAGIRAYEQLGQRAFGPAGKVVVATVICLHNVGAMSSYLFIIKSELPLVIGTFLYMDPEGDWFLKGNLLIIIVSVLIILPLALMKHLGYLGYTSGLSLTCMLFFLVSVIYKKFQLGCAIGHNETAMESEALVGLPSQGLNSSCEAQMFTVDSQMSYTVPIMAFAFVCHPEVLPIYTELCRPSKRRMQAVANVSIGAMFCMYGLTATFGYLTFYSSVKAEMLHMYSQKDPLILCVRLAVLLAVTLTVPVVLFPIRRALQQLLFPGKAFSWPRHVAIALILLVLVNVLVICVPTIRDIFGVIGSTSAPSLIFILPSIFYLRIVPSEVEPFLSWPKIQALCFGVLGVLFMAVSLGFMFANWATGQSRMSGH\n\n### Sequence 2:\nMDSSLAISGGPRLSNREWPRWPQPGDRALKSLEDVLTSGRWTISCAYQGRDSYERQFASAFADYCGSAMCVPISTGTASLAIALEACGVGAGDEVIVPGLSWVASASAVLGINAVPVLVDVDPATYCLDPAATEAAITERTRAITVVHAYSAVADLDALLDIARRHGLPLIEDCAHAHGAGFRGRPVGAHGAAGVFSMQGSKLLTCGEGGALVTDDADVALRAEHLRADGRVVRREPVGVGEMELEETGRMMGSNACLSEFHAAVLLDQLELLDGQNARRTRAADHLTDRLSELGMTAQATAPGTTARAYYRYLVRLPDEVLAVAPVERFAHALTAELGFAVTQTHRPLNDNPLNRPSSRRRFATDARYLERVDPSRFDLPAAKRAHESVVSFSHEVLLAPLDAIDDIARAFRKVLDNVREVSR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00286", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nT170A_CHICK\n\n### Answer:", "output": "May regulate membrane morphogenesis in the endoplasmic reticulum (ER) by promoting ER sheet formation at the expense of ER tubules.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00287", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRS14_MYCMM\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uS14 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00288", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: HIV protease\nLigand SMILES: NCCNc1cccc2ccccc12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "NCCNc1cccc2ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 186.26, "LogP": 2.21, "TPSA": 38.05, "QED": 0.77, "functional_groups": ["aromatic ring", "amine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 1.0, "NR-Aromatase": 1.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 1.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00289", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRR18_CHLRE\n\n### Answer:", "output": "30S ribosomal protein S18, chloroplastic.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00290", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nIGHD TCL1A IGHM CD79B MS4A1 CD79A\n</cell>", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as IGHD, MS4A1, CD79B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IGHD", "TCL1A", "IGHM", "CD79B", "MS4A1", "CD79A"]}}
{"id": "eval_00291", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "COL1A1, COL1A2, DCN, PDGFRA, COL3A1", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as COL1A2, DCN, COL1A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "ovary", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["COL1A1", "COL1A2", "DCN", "PDGFRA", "COL3A1"]}}
{"id": "eval_00292", "category": "Mutation", "instruction": "Describe the effect of mutation Q86VP1-Q114A on protein Tax1-binding protein 1.", "input": "### Wild-type:\nMTSFQEVPLQTSNFAHVIFQNVAKSYLPNAHLECHYTLTPYIHPHPKDWVGIFKVGWSTARDYYTFLWSPMPEHYVEGSTVNCVLAFQGYYLPNDDGEFYQFCYVTHKGEIRGQSTPFQFRASSPVEELLTMEDEGNSDMLVVTTKAGLLELKIEKTMKEKEELLKLIAVLEKETAQLREQVGRMERELNHEKERCDQLQAEQKGLTEVTQSLKMENEEFKKRFSDATSKAHQLEEDIVSVTHKAIEKETELDSLKDKLKKAQHEREQLECQLKTEKDEKELYKVHLKNTEIENTKLMSE\n\n### Mutant:\nMTSFQEVPLQTSNFAHVIFQNVAKSYLPNAHLECHYTLTPYIHPHPKDWVGIFKVGWSTARDYYTFLWSPMPEHYVEGSTVNCVLAFQGYYLPNDDGEFYQFCYVTHKGEIRGASTPFQFRASSPVEELLTMEDEGNSDMLVVTTKAGLLELKIEKTMKEKEELLKLIAVLEKETAQLREQVGRMERELNHEKERCDQLQAEQKGLTEVTQSLKMENEEFKKRFSDATSKAHQLEEDIVSVTHKAIEKETELDSLKDKLKKAQHEREQLECQLKTEKDEKELYKVHLKNTEIENTKLMSE\n\n### Answer:", "output": "Complete gain of TBK1 and RB1CC1 binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00293", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SWW6-C1022S on protein Cellulose synthase A catalytic subunit 7 [UDP-forming].", "input": "### Wild-type:\nMEASAGLVAGSHNRNELVVIHNHEEPKPLKNLDGQFCEICGDQIGLTVEGDLFVACNECGFPACRPCYEYERREGTQNCPQCKTRYKRLRGSPRVEGDEDEEDIDDIEYEFNIEHEQDKHKHSAEAMLYGKMSYGRGPEDDENGRFPPVIAGGHSGEFPVGGGYGNGEHGLHKRVHPYPSSEAGSEGGWRERMDDWKLQHGNLGPEPDDDPEMGLIDEARQPLSRKVPIASSKINPYRMVIVARLVILAVFLRYRLLNPVHDALGLWLTSVICEIWFAVSWILDQFPKWFPIERETYLDR\n\n### Mutant:\nMEASAGLVAGSHNRNELVVIHNHEEPKPLKNLDGQFCEICGDQIGLTVEGDLFVACNECGFPACRPCYEYERREGTQNCPQCKTRYKRLRGSPRVEGDEDEEDIDDIEYEFNIEHEQDKHKHSAEAMLYGKMSYGRGPEDDENGRFPPVIAGGHSGEFPVGGGYGNGEHGLHKRVHPYPSSEAGSEGGWRERMDDWKLQHGNLGPEPDDDPEMGLIDEARQPLSRKVPIASSKINPYRMVIVARLVILAVFLRYRLLNPVHDALGLWLTSVICEIWFAVSWILDQFPKWFPIERETYLDR\n\n### Answer:", "output": "Impaired function as cellulose synthase. Abolishes S-acylation. Unable to localize to the plasma membrane; when associated with S-1026. Mutating the cysteine in CESA7 prevents localization of the cellulose synthase complex (CSC) to the plasma membrane.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00294", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSDNDELQQIAHLRREYTKGGLRRRDLPADPLTLFERWLSQACEAKLADPTAMVVATVDEHDQPYQRIVLLKHYDEKGMVFYTNLGSRKAHQIENNPRVSLLFPWHTLERQVMVIGKAERLSTLEVMKYFHSRPRDSQIGAWVSKQSSRISARGILESKFLELKQKFQQGEVPLPSFWGGFRVSLEQIEFWQGGEHRLHDRFLYQRENDAWKIDRLAP\n\n### Sequence 2:\nMSDNDELQQIAHLRREYTKGGLRRRDLPADPLTLFERWLSQACEAKLADPTAMVVATVDEHGQPYQRIVLLKHYDEKGMVFYTNLGSRKAHQIENNPRVSLLFPWHTLERQVMVIGKAERLSTLEVMKYFHSRPRDSQIGSWVSKQSSRISARGILESKFLELKQKFQQGEVPLPSFWGGFRVSLEQIEFWQGGEHRLHDRFLYQRENDAWKIDRLAP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00295", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTTVMKFGGTSVGSGERIRHVAKIVTKRKKEDDDVVVVVSAMSEVTNALVEISQQALDVRDIAKVGDFIKFIREKHYKAIEEAIKSEEIKEEVKKIIDSRIEELEKVLIGVAYLGELTPKSRDYILSFGERLSSPILSGAIRDLGEKSIALEGGEAGIITDNNFGSARVKRLEVKERLLPLLKEGIIPVVTGFIGTTEEGYITTLGRGGSDYSAALIGYGLDADIIEIWTDVSGVYTTDPRLVPTARRIPKLSYIEAMELAYFGAKVLHPRTIEPAMEKGIPILVKNTFEPESEGTLITNDMEMSDSIVKAISTIKNVALINIFGAGMVGVSGTAARIFKALGEEEVNVILISQGSSETNISLVVSEEDVDKALKALKREFGDFGKKSFLNNNLIRDVSVDKDVCVISVVGAGMRGAKGIAGKIFTAVSESGANIKMIAQGSSEVNISFVIDEKDLLNCVRKLHEKFIEKTNS\n\n### Answer:", "output": "<SEQ_2D>CCEEEEECTTTTSSHHHHHHHHHHHHHHHHHCSCEEEEECCCTTHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHHHHHHHHHCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHHHHHHHHHHHHHHHHHHTTCCEEEECTTTTTEEECSCTTSCCEEEECHHHHHHHHHHTTCEEEEETTEEECTTCCEEECCTTHHHHHHHHHHHHHTCSEEEEEESSSSCBSSCTTTCTTCCBCSEEEHHHHHHHCCTTCSSCCHHHHHHHHHHTCCEEEEESSCTTSCCEEEECCCCCCCCSCCEEEEEEEEEEEEEEESSCSTTSCHHHHHHHHHHTTTCCEEEEEECTTCSEEEEEEETTSHHHHHHHHHHHTCCCCCCCCCCSCSEEEEEEEEEEEEEEEECSTTTTCTTHHHHHHHHHHHTTCCCCEEEECSSTTEEEEEEEGGGHHHHHHHHHHHHTCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00296", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTNNI_ASTLP\n\n### Answer:", "output": "Troponin I is the actomyosin ATPase inhibitory subunit present in the thin filament regulatory complex.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00297", "category": "Mutation", "instruction": "Describe the effect of mutation Q5WNW9-E179K on protein HTH-type transcriptional activator BcrR.", "input": "### Wild-type:\nMEFNEKLQQLRTGKNLTQEQLAEQLYVSRTAISKWESGKGYPNMESLKCISKFFSVTIDELLSGEELITLAETENRSNLKKIYNYIYGILDMMAVAFIFLPLYGNSVGGYVYAVNLLSFTATTPFNLAVYWSAFAALIIIGIGKIISTHLDKEKWGGIATKCSLTITALAVCFFAAAREPYITVLVFLLLIGKIFVWIKQMGMK\n\n### Mutant:\nMEFNEKLQQLRTGKNLTQEQLAEQLYVSRTAISKWESGKGYPNMESLKCISKFFSVTIDELLSGEELITLAETENRSNLKKIYNYIYGILDMMAVAFIFLPLYGNSVGGYVYAVNLLSFTATTPFNLAVYWSAFAALIIIGIGKIISTHLDKEKWGGIATKCSLTITALAVCFFAAARKPYITVLVFLLLIGKIFVWIKQMGMK\n\n### Answer:", "output": "Loss of activity. Does not affect DNA-binding and cellular localization.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00298", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSSNSPSLETDVDDVENIVFQFQNSSLDFQSSDDFSILGIDQPHPIVRIGGMFFRGTWHQPIGTDIVVPSVNDSELSRDGLVLCKRRLMLEQIRLVPKNPSPSSSIHSPTQGEPENISEN\n\n### Sequence 2:\nMGAANKKQELRLKRKARIRKKIAGTPERPRLSIFRSARHIYAQLIDDTKGVTFVTASSNEPDVKNNTELSGKNKSEVAVFVGKLIAGRAKDKGISSVVFDRGGFVYHGRVKAVSDGAREGGLNF\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00299", "category": "Mutation", "instruction": "Describe the effect of mutation P13929-E374G on protein Beta-enolase.", "input": "### Wild-type:\nMAMQKIFAREILDSRGNPTVEVDLHTAKGRFRAAVPSGASTGIYEALELRDGDKGRYLGKGVLKAVENINNTLGPALLQKKLSVVDQEKVDKFMIELDGTENKSKFGANAILGVSLAVCKAGAAEKGVPLYRHIADLAGNPDLILPVPAFNVINGGSHAGNKLAMQEFMILPVGASSFKEAMRIGAEVYHHLKGVIKAKYGKDATNVGDEGGFAPNILENNEALELLKTAIQAAGYPDKVVIGMDVAASEFYRNGKYDLDFKSPDDPARHITGEKLGELYKSFIKNYPVVSIEDPFDQDD\n\n### Mutant:\nMAMQKIFAREILDSRGNPTVEVDLHTAKGRFRAAVPSGASTGIYEALELRDGDKGRYLGKGVLKAVENINNTLGPALLQKKLSVVDQEKVDKFMIELDGTENKSKFGANAILGVSLAVCKAGAAEKGVPLYRHIADLAGNPDLILPVPAFNVINGGSHAGNKLAMQEFMILPVGASSFKEAMRIGAEVYHHLKGVIKAKYGKDATNVGDEGGFAPNILENNEALELLKTAIQAAGYPDKVVIGMDVAASEFYRNGKYDLDFKSPDDPARHITGEKLGELYKSFIKNYPVVSIEDPFDQDD\n\n### Answer:", "output": "Our data suggest that ENO3 mutations result in increased stability of mutant beta-enolase.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00300", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "N=C(N)N[N+](=O)[O-]", "output": "The molecule contains the following functional groups: amine, nitro.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "N=C(N)N[N+](=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 104.07, "LogP": -1.34, "TPSA": 105.04, "QED": 0.17, "functional_groups": ["amine", "nitro"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amine", "nitro"]}}
{"id": "eval_00301", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTDTTYDNSTELDPRIAARLKRDAKGLVAAVIQQYDTREVLMVGYMNDEALRRTLTTGRVTFWSRSRQEYWRKGDTSGHVQYVKGVSLDCDGDALLVEVDQVGAACHTGKRSCFLEGGPLPVVEGHRPAEQ\n\n### Sequence 2:\nMTNEAYDNTTTLDPRIAQRLKHDDKGLVAAVIQQFDTKEVLMVGYMNDEAIRRTLTTGRVTFWSRSRQEYWRKGDTSGHAQYVKSFALDCDGDAILVEVDQVGAACHTGKRSCFEEGGQLPVVVGHRTKEQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00302", "category": "Mutation", "instruction": "Describe the effect of mutation P22301-A42R on protein Interleukin-10.", "input": "### Wild-type:\nMHSSALLCCLVLLTGVRASPGQGTQSENSCTHFPGNLPNMLADLRDAFSRVKTFFQMKDQLDNLLLKESLLEDFKGYLGCQALSEMIQFYLEEVMPQAENQDPDIKAHVNSLGENLKTLRLRLRRCHRFLPCENKSKAVEQVKNAFNKLQEKGIYKAMSEFDIFINYIEAYMTMKIRN\n\n### Mutant:\nMHSSALLCCLVLLTGVRASPGQGTQSENSCTHFPGNLPNMLRDLRDAFSRVKTFFQMKDQLDNLLLKESLLEDFKGYLGCQALSEMIQFYLEEVMPQAENQDPDIKAHVNSLGENLKTLRLRLRRCHRFLPCENKSKAVEQVKNAFNKLQEKGIYKAMSEFDIFINYIEAYMTMKIRN\n\n### Answer:", "output": "About 80% gain of IL10RA binding. The study unmasks the IL-10R2 binding site on IL-10 using surface plasmon resonance and cell-based assays. Critical IL-10R2 binding residues are located in helix A adjacent to the previously identified IL-10R1 recognition surface. Interestingly, IL-10R2 binding residues located in the N-terminal end of helix A exhibit small structural differences between unbound cIL-10 and cIL-10. IL-10R1 crystal structures. This suggests IL-10R1-induced conformational changes regulate IL-10R2 binding and disassembly of the ternary IL-10. IL-10R1. IL-10R2 complex.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00303", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTTLADLRINYSRASLDEADAAPDPFAQFDRWFKEALAAKLPEPNTMTLATVGADGRPSARIVLIKGVDERGFVFFTNYESRKGHDLAVHPQAALLFYWIELERQVRIEGRIEKTSAEESDRYFASRPLGSRIGAWASEQSAVIDSRATLEAREKAVSERYGDNPPRPPHWGGYRLVPDSIEFWQGRPSRLHDRLLYTRDAAASPGWTISRLSP\n\n### Sequence 2:\nMSNEAGGGIHKLICLLCDSQFSQDVTWRALFLLKPDEKVCYSCRSKLKKITGHICPLCGRPQSVHAVCRDCEVWRTRIRDSLLLRQNRSVYTYNDMMKETLSRFKFRGDAEIINAFKSDFSSTFSKVYPDKHFVLVPIPLSKEREEERGFNQAHLLAECLDRPSHHPLIRLNNEKQSKKKKTERLLSECIFDTKNNSAEGMNIILIDDLYTTGATLHFAARCLLEKGKAASVSSFTLIRS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00304", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: Cc1cc2n[nH]c(=O)n2c2cc(-c3cn[nH]c3)ccc12\nDescriptors: MW=265.28, LogP=1.87, TPSA=78.84, HBD=2, HBA=3", "output": "Descriptors: MW=265.28, LogP=1.87, TPSA=78.84, HBD=2, HBA=3, QED=0.55, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "Cc1cc2n[nH]c(=O)n2c2cc(-c3cn[nH]c3)ccc12", "compound_name": null, "pubchem_cid": null, "MW": 265.28, "LogP": 1.87, "TPSA": 78.84, "QED": 0.55, "functional_groups": ["aromatic ring", "pyridine"], "moleculenet_labels": {"label": 2.13}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00305", "category": "Mutation", "instruction": "Describe the effect of mutation C7E9W0-K101A on protein Antigenic protein SchS21.", "input": "### Wild-type:\nASVTFWTLDNVDRTLVFTGNPGSAAIETITVGPAENTTVEFPGSWVGNWYAYPTDAEDVPGMLGEVQFGGWNGLTYFDVSAIVNPTDHDNVKQMWPAESRKPMSGCEVFPCDNAYWLPDDIQTKVTHEVDLWTTLGAGSTGLTF\n\n### Mutant:\nASVTFWTLDNVDRTLVFTGNPGSAAIETITVGPAENTTVEFPGSWVGNWYAYPTDAEDVPGMLGEVQFGGWNGLTYFDVSAIVNPTDHDNVKQMWPAESRAPMSGCEVFPCDNAYWLPDDIQTKVTHEVDLWTTLGAGSTGLTF\n\n### Answer:", "output": "Significant decrease in IgE-binding of the 91-V--P-102 and 94-M--G-105 peptides. mutation is a critical amino acid for IgE binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00306", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCIP4_PONAB\n\n### Answer:", "output": "Cdc42-interacting protein 4, Thyroid receptor-interacting protein 10, TR-interacting protein 10, TRIP-10.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00307", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIIPALDLIEGQVVRLYQGDYGQVTEYKVDPAEQFNLYHQAGADWLHLVDLTGAKDTSARQLDLIAKLLASTPANIQIGGGVRTEQDVVDLLEAGAQRVVVGSTAVKQPELVKGWMEKYGAEKIVLALDINIDQDGTRKVAISGWQEDSGVTIEALINDYLTVGLQHVLCTDISRDGTLEGSNVELYVDLCKQYPQVQFQSSGGIGSLADIEALKGSGVAGVIVGRALLDGKFTAEEAFACWQSE\n\n### Sequence 2:\nMIIPALDLIEGQVVRLYQGDYGQVTEYKVDPAEQFNLYHQAGANWLHLVDLTGAKDTTARQLDLIAKLLASTPANIQIGGGVRTEQDVIDLLEAGAQRVVVGSTAVKQPELVKGWMEKYGAEKIVLALDINIDQDGTRKVAISGWQEDSGVTIEALINDYLTVGLQHVLCTDISRDGTLEGSNVELYVDLCKQYPQVQFQSSGGIGSLADIEALKGSGVAGVIVGRALLDGKFTAEEAFACWQSE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00308", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRNPH_AGRRK\n\n### Answer:", "output": "Phosphorolytic 3'-5' exoribonuclease that plays an important role in tRNA 3'-end maturation.  Removes nucleotide residues following the 3'-CCA terminus of tRNAs; can also add nucleotides to the ends of RNA molecules by using nucleoside diphosphates as substrates, but this may not be physiologically important.  Probably plays a role in initiation of 16S rRNA degradation (leading to ribosome degradation) during starvation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00309", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMDSQGPIILEKSIKIEEVIKIANTSIIDIVTKTVTPEIKAPYELVDVEYDKMGSDYILSILVDKEDGITVEDTSDLTNIISPLLDTIDPDPFPNQYMLEVSSPGLERPLKTADSLKAAVGSYINVSLYQAIDKVKVFQGDLLAFDGETLTIDYLDKTRHKIVNIPYQAVAKVRMAVKL\n\n### Sequence 2:\nMRVGLFAVGRLKSGPEKDLAARYFDRFAKAGPAVGLELARIAETAESRASNAETRKREEAAILLKSLAEGSVLILLDERGKALDSEAFANLLGAYRDQGKRDLMIAIGGADGLDPALYDRADVTLCLGKMTWPHQLVRTLIAEQLYRAVTILSGHPYHRV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00310", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSKVCEISGKRPIVANSIQRRGKAKREGGVGKKTTGISKRRQYPNLQKVRVRVAGQEITFRVAASHIPKVYELVERAKGLKLEGLSPKEIKKELLKLL\n\n### Answer:", "output": "<SEQ_2D>CCCCCTTTCCCCEEEEEECCCSCCSSSSSSCCCCCCEEEEEECCCCEEEEECCSSSCEEEEECSSSHHHHHHHHHHTTSSCCCSSCHHHHHHHCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00311", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nSFTPC SFTPB SFTPA1 ABCA3\n</cell>", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as ABCA3, SFTPB, SFTPA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SFTPC", "SFTPB", "SFTPA1", "ABCA3"]}}
{"id": "eval_00312", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNHYLLLLFVIFLTCAGQLSQKQATHSWKENAKRQTLIWLGLSVIFLAGGMLLWLKLLQYLPLSQAYPFLSINLILVTLSGHFFFKEKVTLQHWLGIGIMMVGILLLGQGI\n\n### Sequence 2:\nMAEKMEKTGQILQMQLKRFSRVEKAFYFSIAVTTLIVAISIIFMQTKLLQVQNDLTKINAQIEEKKTELDDAKQEVNELLRAERLKEIANSHDLQLNNENIRIAE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00313", "category": "Mutation", "instruction": "Describe the effect of mutation P09230-A397S on protein Alkaline extracellular protease.", "input": "### Wild-type:\nMKLATAFTILTAVLAAPLAAPAPAPDAAPAAVPEGPAAAAYSSILSVVAKQSKKFKHHKRDLDEKDQFIVVFDSSATVDQIASEIQKLDSLVDEDSSNGITSALDLPVYTDGSGFLGFVGKFNSTIVDKLKESSVLTVEPDTIVSLPEIPASSNAKRAIQTTPVTQWGLSRISHKKAQTGNYAYVRETVGKHPTVSYVVDSGIRTTHSEFGGRAVWGANFADTQNADLLGHGTHVAGTVGGKTYGVDANTKLVAVKVFAGRSAALSVINQGFTWALNDYISKRDTLPRGVLNFSGGGPKS\n\n### Mutant:\nMKLATAFTILTAVLAAPLAAPAPAPDAAPAAVPEGPAAAAYSSILSVVAKQSKKFKHHKRDLDEKDQFIVVFDSSATVDQIASEIQKLDSLVDEDSSNGITSALDLPVYTDGSGFLGFVGKFNSTIVDKLKESSVLTVEPDTIVSLPEIPASSNAKRAIQTTPVTQWGLSRISHKKAQTGNYAYVRETVGKHPTVSYVVDSGIRTTHSEFGGRAVWGANFADTQNADLLGHGTHVAGTVGGKTYGVDANTKLVAVKVFAGRSAALSVINQGFTWALNDYISKRDTLPRGVLNFSGGGPKS\n\n### Answer:", "output": "Highly enhances protease activity, but not maturation. After a change of the putative active site S to A, active AEP with the same mobility on SDS-PAGE as wild-type mature AEP was secreted.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00314", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nRAFSPGLTGVLPLRETRHLVEVLRARVGDRFTVFDGEREALAEVVDLGPPLRYRVLEERRPEREVGVEVVLYVALLKGDKLAEVVRAATELGATRIQPLVTRHSVPKEMGEGKLRRLRAVALEAAKQSGRVVVPEVLPPIPLKAVPQVAQGLVAHVGATARVREVLDPEKPLALAVGPEGGFAEEEVALLEARGFTPVSLGRRILRAETAALALLALCTAGEGR\n\n### Answer:", "output": "<SEQ_3Di>DAEDAPPPQFGDPVRLCCVCPVVNDDFQDKAKYDHQQWIFIWTFHASPVVTGIGGDGTDQDDQALQAAAEEEEEAEPDLLVLLQLLLLLLLHHQEYEYEQAPPHPDRFDDPVSLVSSQVSQCVSCVVNSHPHGYHYDTYYYLLPLAAAPAEEEADQPDDDACVVPQDSVGHYYYYWAYQVTHDPVSVVVVVVRHYDYHYQDNDDDHSRVSSNSVSCCRGVVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00315", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMDRFLSNVTNRIDAKGRVSVPSPFRSVLARRDIQELYCLQDFAFPAISVGGPDLLERYERQIASMDAFSPEANAMSLLVHGGGVFMKLDQEGRLMVTDFVREFTGISTEVTFVGRADHFQLWQPNAFLAAQAEARAGRGFSAARPA\n\n### Sequence 2:\nMSQNDLIKTIRMNYLFDFYQALLTKKQRNYLELFYLQDYSLSEIANTFDVSRQAVYDNIRRTGDLVEDYEEKLGLYRNFEQRQNLYEQIKQHINDPEQIKKYIQALEDLD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00316", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nFNLDAEAPAVLSGPPGSFFGFSVEFYRPGTDGVSVLVGAPKANTSQPGVLQGGAVYLCPWGASPTQCTPIEFDSKGSRLLESSLSSSEGEEPVEYKSLQWFGATVRAHGSSILACAPLYSWRTEKEPLSDPVGTCYLSTDNFTRILEYAPCRSDFSWAAGQGYCQGGFSAEFTKTGRVVLGGPGSYFWQGQILSATQEQIAESYYPEYLINLVQGQLQTRQASSIYDDSYLGYSVAVGEFSGDDTEDFVAGVPKGNLTYGYVTILNGSDIRSLYNFSGEQMASYFGYAVAATDVNGDGLDDLLVGAPLLMDRTPDGRPQEVGRVYVYLQHPAGIEPTPTLTLTGHDEFGRFGSSLTPLGDLDQDGYNDVAIGAPFGGETQQGVVFVFPGGPGGLGSKPSQVLQPLWAASHTPDFFGSALRGGRDLDGNGYPDLIVGSFGVDKAVVYRGRPIVSASASLTIFPAMFNPEERSCSLEGNPVACINLSFCLNASGKHVADSIGFTVELQLDWQKQKGGVRRALFLASRQATLTQTLLIQNGAREDCREMKIYLRNESEFRDKLSPIHIALNFSLDPQAPVDSHGLRPALHYQSKSRIEDKAQILLDCGEDNICVPDLQLEVFGEQNGGLENLYFQ\n\n### Answer:", "output": "<SEQ_2D>CCBCCSSCEEEECSTTSCTTSEEEEECSSSSCCEEEEEETTCCCCSTTCTTCCEEEEEESSCCSSCCEEECSCCCCCCBCSTTSSSSSCCCBSEECTTCCTTSEEEEETTEEEEEETTCEECCSSSCCCBCCCEEEEEETTTTEEEEECTTSSSCCSTTTTTTCCTTSEEEECTTCCEEEEETTHHHHTBEEEEECHHHHHHHCCTTSSCCCCTTCEECCCCCGGGCSCCBTSSEEEECCSSSSSCEEEEEEEEEETTEEEEEEECSSSCCEEEEEECSSTTSCTTSEEEEECCSSSSSCEEEEEETTCEEECTTCCEEECCEEEEECBCSSCBCSSCSEEEECCSTTCCGGGEEEEEESSSSSSSCEEEEEETTCSTTSSCEEEEECEETTEECSSCSEEECCSSCCCSSCCCTTSSEEESSCCSSSSCCCEEEEEGGGTEEEEECBCCEEEEEEEEEESSCEECTTCCCEESSSSEECEEEEEEEEEEECSSSCSEEEEEEEEEETTTTTTTSCCCEEETTTSSSEEEEEEEEETTCCCEEEEEEEEECCCSCCSCCSSCEEEEEEEEECSSSCCCSSSCCCEECTTSCCEEEEEEEECCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00317", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDNAJ_BURM1\n\n### Answer:", "output": "Chaperone protein DnaJ.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00318", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CLCA1, TFF3, MUC2", "output": "This cell is most consistent with goblet cell, an intestinal goblet cell. Key markers such as MUC2, CLCA1, TFF3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "goblet cell", "tissue": "spleen", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CLCA1", "TFF3", "MUC2"]}}
{"id": "eval_00319", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C(c1cc(Cc2n[nH]c(=O)c3ccccc23)ccc1F)N1CCN(c2ccccn2)CC1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, halide, pyridine. Its descriptor profile: MW=443.48, LogP=3.01, TPSA=82.19, HBD=1, HBA=5, QED=0.52.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "O=C(c1cc(Cc2n[nH]c(=O)c3ccccc23)ccc1F)N1CCN(c2ccccn2)CC1", "compound_name": null, "pubchem_cid": null, "MW": 443.48, "LogP": 3.01, "TPSA": 82.19, "QED": 0.52, "functional_groups": ["aromatic ring", "amide", "halide", "pyridine"], "moleculenet_labels": {"label": 2.55}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00320", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYIDC_SHEWM\n\n### Answer:", "output": "Membrane protein insertase YidC, Foldase YidC, Membrane integrase YidC, Membrane protein YidC.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00321", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD4, FOXP3, IL2RA, CD3D", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as CD3D, IL2RA, FOXP3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "blood", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD4", "FOXP3", "IL2RA", "CD3D"]}}
{"id": "eval_00322", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGLQIKKNAIIDDYKVTSQVLGLGINGKVLQIFNKRTQEKFALKMLQDCPKARREVELHWRASQCPHIVRIVDVYENLYAGRKCLLIVMECLDGGELFSRIQDRGDQAFTEREASEIMKSIGEAIQYLHSINIAHRDVKPENLLYTSKRPNAILKLTDFGFAKETTSHNSLTTPCYTPYYVAPEVLGPEKYDKSCDMWSLGVIMYILLCGYPPFYSNHGLAIKTRIRMGQYEFPNPEWSEVSEEVKMLIRNLLKTEPTQRMTITEFMNHPWIMQSTKVPQTPLHTSRVLKEDKERWEDVKEEMTSALATMRVDYEQIKIKKIEDASNPLLLKRRKKARALEAAALA\n\n### Answer:", "output": "<SEQ_3Di>DADAAADDPVVFKDWDPDFPDADDPGTKTWMAGPVVRAIWIKDWAFDDPLSVVLVVLLQVLQPQQQAWHWDHWYFDDDPNTGTIITITHDFAQAWPLVLLQDDDPDFDWVLNLQVVLLLVLSSLVVCVVQQKQQQPDDRRQWTWRHVPPPTRTHGDDSSVIDRADPDQDPDDDPDDDPDDDDDDDDQSLLRQLSVLLNSLQVSVCSFAVAGQDPPPDDLNCDPDPDDDPSNQNCVFCVPPDPLSSVLSCLSSPPDSNSHDGSVVVCVDCCRVPSVPDDGDGGCNNPNCNVDVVVVVVRVVVVVVNSPVSHDDPPDDDDDDPVVDDDPVRCVVVVVVVVVVVVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00323", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nPF4:high ITGA2B:high GP1BA:high PPBP:low\n</cell>", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as PF4, GP1BA, ITGA2B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "bone marrow", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["PF4", "ITGA2B", "GP1BA", "PPBP"]}}
{"id": "eval_00324", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL31B_STRPZ\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bL31 family. Type B subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00325", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMFKQVLGGMALMAAFSAPVLAAECSVDIAGTDQMQFDKKAIEVSKSCKQFTVNLKHTGKLPRNVMGHNWVLTKTADMQAVEKDGIAAGLDNQYLKAGDTRVLAHTKVLGGGESDSVTFDVAKLAAGDDYTFFCSFPGHGALMKGTLKLVD\n\n### Sequence 2:\nMSEQNNTEMNFQLLRIYTKDISFEAPNAPHVFQKDWQPEVKLDLDTASSQLADDVYEVVLRVTVTASLGEETAFLCEVQQAGIFNISGIEGTQMAHCLGAYCPNSLFPYARECITNLVSRGTFPQLNLAPVNFDALFMNYLQQQAAEGAPNHQDA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00326", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVKVKSKNSVIKLLSTAASGYSRYISIKKGAPLVTQVRYDPVVKRHVLFKEAKKRKVAERKPLDFLRTAK\n\n### Sequence 2:\nAVLDGDVRKCLPCGPRNRGRCFGPRICCGEELGCYLGTPETLRCQEESFLPTPCESGRKPCGGDGASCAAPGICCSSEGCVADPACEREALFA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00327", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: C[C@H]1C[C@H]2[C@@H]3CCC4=CC(=O)C=C[C@]4(C)[C@@]3(F)C(=O)C[C@]2(C)[C@@]1(O)C(=O)CCl\nDescriptors: MW=408.9, LogP=3.35, TPSA=71.44, HBD=1, HBA=4", "output": "Descriptors: MW=408.9, LogP=3.35, TPSA=71.44, HBD=1, HBA=4, QED=0.71, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "BBBP", "canonical_smiles": "C[C@H]1C[C@H]2[C@@H]3CCC4=CC(=O)C=C[C@]4(C)[C@@]3(F)C(=O)C[C@]2(C)[C@@]1(O)C(=O)CCl", "compound_name": null, "pubchem_cid": null, "MW": 408.9, "LogP": 3.35, "TPSA": 71.44, "QED": 0.71, "functional_groups": ["hydroxyl", "ketone", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00328", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nVLP21_BORHE\n\n### Answer:", "output": "The Vlp and Vsp proteins are antigenically distinct proteins, only one vlp or vsp gene is transcriptionally active at any one time.  Switching between these genes is a mechanism of host immune response evasion.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00329", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMQVDLLSSAQSAHTLHLFHQHSPLVHCMTNDVVQTFTANTLLALGASPAMVIETEEASQFAAIASALLINVGTLTQPRAQAMRAAVEQAKSSQTPWTLDPVAVGALDYRRHFCHELLSFKPAAIRGNASEIMALAGVANGGRGVDTTDAAVNAIPAAQTLARETGAIVVVTGEVDYVTDGHRAVGIHGGDPLMTKVVGTGCALSAVVAACCALPGDMLENVASACHWMKQAGERAVARSEGPGSFVPHFLDALWQLTQEVQA\n\n### Sequence 2:\nMTDSHYIGRFAPSPSGELHFGSLIAALGSYLQARAQRGIWRVRIEDIDPPREVPGAAATILRQLEHYGLHWDGEVLWQSQRHEAYREALAWLHEQGLSYYCTCPRSRIQRLGGIYDGHCRTLCHGPENAAVRIKQQHPVMHFHDALRGDIQADPQLASEDFIIHRRDGLFAYNLAVVVDDHFQGVTEIVRGADLIEPTVRQLSLYKQFGWRAPGYVHLPLALNEQGAKLSKQNHAPALATGDPRPVLVQALRFLGQRDVVAWQEMSVEELLRFAVAHWRLTAVPTSANVNPAFSNASR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00330", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SYM4-P27L on protein Alpha,alpha-trehalose-phosphate synthase [UDP-forming] 1.", "input": "### Wild-type:\nMPGNKYNCSSSHIPLSRTERLLRDREPREKRKSNRARNPNDVAGSSENSENDLRLEGDSSRQYVEQYLEGAAAAMAHDDACERQEVRPYNRQRLLVVANRLPVSAVRRGEDSWSLEISAGGLVSALLGVKEFEARWIGWAGVNVPDEVGQKALSKALAEKRCIPVFLDEEIVHQYYNGYCNNILWPLFHYLGLPQEDRLATTRSFQSQFAAYKKANQMFADVVNEHYEEGDVVWCHDYHLMFLPKCLKEYNSKMKVGWFLHTPFPSSEIHRTLPSRSELLRSVLAADLVGFHTYDYARHF\n\n### Mutant:\nMPGNKYNCSSSHIPLSRTERLLRDRELREKRKSNRARNPNDVAGSSENSENDLRLEGDSSRQYVEQYLEGAAAAMAHDDACERQEVRPYNRQRLLVVANRLPVSAVRRGEDSWSLEISAGGLVSALLGVKEFEARWIGWAGVNVPDEVGQKALSKALAEKRCIPVFLDEEIVHQYYNGYCNNILWPLFHYLGLPQEDRLATTRSFQSQFAAYKKANQMFADVVNEHYEEGDVVWCHDYHLMFLPKCLKEYNSKMKVGWFLHTPFPSSEIHRTLPSRSELLRSVLAADLVGFHTYDYARHF\n\n### Answer:", "output": "4-fold decrease in activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00331", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPLSY_BURO0\n\n### Answer:", "output": "Glycerol-3-phosphate acyltransferase, Acyl-PO4 G3P acyltransferase, Acyl-phosphate--glycerol-3-phosphate acyltransferase, G3P acyltransferase, GPAT, Lysophosphatidic acid synthase, LPA synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00332", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTKSELIERLASQQSHIPAKAVEDAVKEMLEHMASTLAQGERIEIRGFGSFSLHYRAPRTGRNPKTGDKVDLEGKYVPHFKPGKELRDRANIYGN\n\n### Sequence 2:\nMANIKSAIKRAKLSEERRAHNASIKSDMRSAVKTVEALVTNNDLENAKEAFKTASKKLDKAARKGLIHQNAAARQKSRLAKQVNA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00333", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPGIKVREGDAFDEAYRRFKKQTDRNLVVTECRARRFFESKTEKRKKQKISAKKKVLKRLYMLRRYESRL\n\n### Sequence 2:\nMPGIKVREGDAFDEAYRRFKKQTDRNLVVTECRARRFFESKTEKRKKQKISAKKEVLKRLYMLRRYESRL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00334", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSSLCLQRLQEERKKWRKDHPFGFYAKPVKKADGSMDLQKWEAGIPGKEGTNWAGGVYPITVEYPNEYPSKPPKVKFPAGFYHPNVYPSGTICLSILNEDQDWRPAITLKQIVLGVQDLLDSPNPNSPAQEPAWRSFSRNKAEYDKKVLLQAKQYSK\n\n### Sequence 2:\nMVKVIASSVRKGNVLDVDGKLYVVLTAQNFHPGKGTPVTQVDMRRIVDGVKVSERWRTTEQVERAFVEDVNFQYLYEDGEGFHFMNPANYDQVVVSQETMGDQKAYLQEGMTCILSIHEGIPLALELPRHVTLEIVETEPVVKGQTASSSYKPAMLSNGIRTSVPPHIDAGTRVVIATEDNSYVERAKD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00335", "category": "Mutation", "instruction": "Describe the effect of mutation O75603-Y282D on protein Chorion-specific transcription factor GCMb.", "input": "### Wild-type:\nMPAAAVQEAVGVCSYGMQLSWDINDPQMPQELALFDQFREWPDGYVRFIYSSDEKKAQRHLSGWAMRNTNNHNGHILKKSCLGVVVCTQACTLPDGSRLQLRPAICDKARLKQQKKACPNCHSALELIPCRGHSGYPVTNFWRLDGNAIFFQAKGVHDHPRPESKSETEARRSAIKRQMASFYQPQKKRIRESEAEENQDSSGHFSNIPPLENPEDFDIVTETSFPIPGQPCPSFPKSDVYKATCDLATFQGDKMPPFQKYSSPRIYLPRPPCSYELANPGYTNSSPYPTLYKDSTSIPN\n\n### Mutant:\nMPAAAVQEAVGVCSYGMQLSWDINDPQMPQELALFDQFREWPDGYVRFIYSSDEKKAQRHLSGWAMRNTNNHNGHILKKSCLGVVVCTQACTLPDGSRLQLRPAICDKARLKQQKKACPNCHSALELIPCRGHSGYPVTNFWRLDGNAIFFQAKGVHDHPRPESKSETEARRSAIKRQMASFYQPQKKRIRESEAEENQDSSGHFSNIPPLENPEDFDIVTETSFPIPGQPCPSFPKSDVYKATCDLATFQGDKMPPFQKYSSPRIYLPRPPCSYELANPGDTNSSPYPTLYKDSTSIPN\n\n### Answer:", "output": "Shows normal transcriptional activity. The mutation in GCM2 is a gain-of-function mutation that increases the transcriptional activity of GCM2, suggesting that GCM2 is a parathyroid proto-oncogene.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00336", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMARTKLKFRLHRAVIVLFCLALLVALMQGASWFSQNHQRQRNPQLEELARTLARQVTLNVAPLMRTDSPDEKRIQAILDQLTDESRILDAGVYDEQGDLIARSGESVEVRDRLALDGKKAGGYFNQQIVEPIAGKNGPLGYLRLTLDTHTLATEAQQVDNTTNILRLMLLLSLAIGVVLTRTLLQGKRTRWQQSPFLLTASKPVPEEEESEKKE\n\n### Sequence 2:\nMARAKLKFRLHRTAIVLICLALLVLLMQGASYFSLSHQLARSEQVEELAQTLAKQVAFSLSPIMDSGDDNIDSQKVDAILQQLTQSSRILDASVYQIDGTLVASAGENVKVRDRLALDGKRPGSYFNHQIVEMIAGNSGPSGFLRMTLDTHVLATESKQVDNTTNLLRLMILVALAVGIILARTLLQGRRSRWQQSPYLLTANIPVKEEDESE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00337", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLIHVRDVLSAAELQQTRELLSDAPWADGGVTAGVQSAQVKNNQQLPQAHPVTQALQQLVMAGLNRHPTFFSAALPKKVFPPLFNRYGGTANAFGNHVDNAVRFITSGPHTGQRVRTDVSCTLFLAEPETYDGGELVIEDTYGSQRVKLAAGDLVLYPGTSVHRVEPVTRGARVGCFFWIESMVRSDEQRRLLHEMDQHLMHLRRTVGETDAGVIGLTGTYHNLLRLWADT\n\n### Sequence 2:\nMLIHVRDVLSAAELQQTRELLSDAPWADGGVTAGVQSAQVKNNQQLPQAHPVTQALQQLVMAGLNRHPTFFSAALPKKVFPPLFNRYGGTANAFGNHVDNAVRFITSGPHTGQRVRTDVSCTLFLAEPETYDGGELVIEDTYGSQRVKLAAGDLVLYPGTSVHRVEPVTRGARVGCFFWIESMVRSDEQRRLLHEMDQHLMHLRRTVGETDAGVIGLTGTYHNLLRLWADT\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00338", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "CD163, CD206, APOE", "output": "This profile is most consistent with tumor-associated macrophage, which is commonly found in brain under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "brain", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CD163", "CD206", "APOE"]}}
{"id": "eval_00339", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD3D IL2RA IKZF2 CD4 FOXP3 CTLA4\n</cell>", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as IL2RA, CD4, FOXP3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "prostate", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD3D", "IL2RA", "IKZF2", "CD4", "FOXP3", "CTLA4"]}}
{"id": "eval_00340", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "O=c1c(O)cccc2cc(O)c(O)c(O)c12", "output": "The molecule contains the following functional groups: aromatic ring, hydroxyl, phenol.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "O=c1c(O)cccc2cc(O)c(O)c(O)c12", "compound_name": null, "pubchem_cid": null, "MW": 220.18, "LogP": 1.02, "TPSA": 97.99, "QED": 0.5, "functional_groups": ["aromatic ring", "hydroxyl", "phenol"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "hydroxyl", "phenol"]}}
{"id": "eval_00341", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "HNF4A, TTR, ALB", "output": "The expression pattern is consistent with hepatocyte in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "prostate", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["HNF4A", "TTR", "ALB"]}}
{"id": "eval_00342", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY4162_YERE8\n\n### Answer:", "output": "Putative transport protein YE4162.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00343", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMGCGGSRADAIEPRYYESWTRETESTWLTYTDSDALPSAAATDSGPEAGGLHAGVLEDGLSSNGVLRPAAPGGIANPEKKMNCGTQCPNSQNLSSGPLTQKQNGLWATEAKRDAKRMSAREVAINVTENIRQMDRSKRVTKNCIN\n\n### Sequence 2:\nMSRNPEVLWAQRSDKVYLTVALPDAKDISVKCEPQGLFSFSALGAQGERFEFSLELYGKIMTEYRKNVGLRNIIFSIQKEERSWWTRLLKSEEKPAPYIKVDWNKWCDEDEEVNSETASDDESAFVNQDSESSDDDGLLYLPDLEKARNK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00344", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMARVKTGVVRRRRHKKVLKLARGFFSARRKHFRKAKEQVERSLVYSYRDRRNKKRDFRRLWIVRINAACRLNDISYSKFINALKKANIELDRKVLADLAMNDAAAFSAIVAQAKKVM\n\n### Sequence 2:\nMARVKTGVVRRRRHKKVLKLARGFFSARRKHFRKAKEQVERSLVYSYRDRRNKKRDFRRLWIVRINAACRLNDISYSKFINALKKANIELDRKVLADLAMNDAAAFSAIVAQAKKVM\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00345", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nS100A9 CXCR2 CSF3R S100A8 FCGR3B\n</cell>", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as CXCR2, S100A8, FCGR3B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "intestine", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["S100A9", "CXCR2", "CSF3R", "S100A8", "FCGR3B"]}}
{"id": "eval_00346", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMFGPVSVEMIIERVDEVRGKVKAPPSKSYTHRAYFLSLLADSPSKVMNPLISEDTIASLDAISKFGAQVNGNKIIPPQELTPGKIDARESGTTARISLAVASLARGTSVITGKGRLVERPFKPLVDALRSLKVKISGEKLPIAVEGGNPVGEYVKVDCSLSSQFGTAMLILASKIGLTVEMLNPVSRPYIEVTLKVMESFGIEFERNGFKVKVHPGIRGSKFHVPGDYSSASFFLAAGALYGKVKVSNLVKDDPQADARIIDILEEFGADVKVGRKYVVVERNEMKPINVDCSNFPDLFPILAVLASYAEGKSVITGRQLRLKESDRVKAVAVNLRKAGIKVKELPNGLEIVGGKPRGFTVESFNDHRIVMAMAILGLGAEGKTIIKDPHVVSKSYPSFFLDLRRVLNEG\n\n### Sequence 2:\nMIIRPVDEVRGELNAPPSKSYTHRAYFLALLAEGESTIENPLVCDDTLATIEAIRSFGAGVDGKTVVPPEEPSPGFVYARESGTTARFSTALAGGIGGKTLIDGARRLRERPMDGLVKALKGLGAEVDGFSLPLTVKGPVKSGR-VSVDASKSSQFVSGLLLLGAEVGLKVEARNPVSKPYIEMTLRTMEAFGVEFEREGFSFEVYPGVKGTRYKVPGDYSTASFFLAAGALYGKVRVNNLLREDVQADMAFLDALEEFGARVKRGRDYVKVEGGELKAVALDCSDFPDSFPILAVVAAYAEGRSVIRARQLRFKESDRVRAMAVNLSRMGVKVRELEDGLEIEGGRPRGAKVETFNDHRIAMAMSIAALGATGPSIIEDTESVSKSHPGFFDDLRRLLE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00347", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMEFSVESLETPALTKAHLAELLFEQIGLNKRESKEMIDAFFELISDRLVEGKDVKISGFGNFQIRTKAPRPGRNPRTGEAIPIESRRVVTFHASHKLKEQIQGSALLGLKTANIFAEDNGPVT\n\n### Sequence 2:\nMGEIRLVNIGFGNIVSANRIVAIVSPESAPIKRIISEARERGTLIDATYGRRTRAVVITDSDHVVLSAVQPETVAHRLTARDASLAAAEEEEGEE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00348", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD79A, MS4A1, IGHA1", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as IGHA1, MS4A1, CD79A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "breast", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD79A", "MS4A1", "IGHA1"]}}
{"id": "eval_00349", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTSQRTRERLIQRLCEEGVSNTKVLDVIRRTPRHLFVDEALAHRAYEDTALPIGHNQTISQPFMVAHMSELLLEAGPLDKVLEIGTGSGYQTAILAQLVERVFSVERIKVLQDRAKERLVELNLRNVVFRWGDGCEGWPALAPYNGIIVTAVAPEVPQALLDQLAPGGRMVIPVGPAGEAQQLMLIVREEHGFSRRVLGAVRFVPLLNGPLA\n\n### Sequence 2:\nMFPMVTGFMSYGQQTIRATRYIGQSFITTLSHTNRLPITIHYPYEKSITPERFRGRIHFEFDKCIACEVCVRVCPIDLPVVDWRFEKDIKRKQLLNYSIDFGVCIFCGNCVEYCPTSCLSMTEEYELSTYDRHELNYNQIALSRLPISIMEDYTIQTIRNSTKSKIDEDKSSNSRTITDY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00350", "category": "Mutation", "instruction": "Describe the effect of mutation Q0DRX6-W52A on protein Cysteine-tryptophan domain-containing zinc finger protein 3.", "input": "### Wild-type:\nMPSNGGIIPGPANAASLPAPVLIEDNWVCCDMCHKWRLLPYGTNTSMLPKKWICSMLDWLPGMNKCDISEDETTNALNALYVTQIPAAGVSSGGPHTAHASVAASSTYNISGQLGQSRKRKNALKDENCYEHDQQAPAKMTLTSNQQAPAKNREVVDSEHYTNDRDPVSTHDLVPQSKSASERHKSKHKSRSSHSDGGDLTEKSKKHSKSKNRRGIDRDEHKTSKKTKKEDRHYFNKDWKNEYDLAGNKVRDETKALSAKAKMSKDSCEQDEFSLRKEKASRFDILEKTKRINDDDVAFH\n\n### Mutant:\nMPSNGGIIPGPANAASLPAPVLIEDNWVCCDMCHKWRLLPYGTNTSMLPKKAICSMLDWLPGMNKCDISEDETTNALNALYVTQIPAAGVSSGGPHTAHASVAASSTYNISGQLGQSRKRKNALKDENCYEHDQQAPAKMTLTSNQQAPAKNREVVDSEHYTNDRDPVSTHDLVPQSKSASERHKSKHKSRSSHSDGGDLTEKSKKHSKSKNRRGIDRDEHKTSKKTKKEDRHYFNKDWKNEYDLAGNKVRDETKALSAKAKMSKDSCEQDEFSLRKEKASRFDILEKTKRINDDDVAFH\n\n### Answer:", "output": "Abolishes binding to histone H3K4me1/2/3.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00351", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAVQQNKKSRSKRGMRRSHDALSTAQLSVDATSGELHLRHNVTADGFYRGKKVINK\n\n### Sequence 2:\nMQVLVRDNNVDQALRALKKKMQREGIFREMKMRGHYEKPSEKRAREKAEAVRRARKLARKRAQREGLIGGRTGAR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00352", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSKLLPIKVWGQGGPNPPRVAIILEELGLPYEFMPIQLSQVKEPEYLAINPNGRLPAIYDPNTDLTLWESGAIVEYLVERYDTAHDISFPRDTNDAQHARQWLFFQASGQGPYYGQACWFKKYHPEPVPSALERYIKEMNRVSGVVDGYLAKQPVPPGGDGPWLVGNKCSFADLAWVSWQMTLKKIIQTEDGYDVENFPHLKNWLDRMVAREPVKKVLSAVMPPPS\n\n### Sequence 2:\nMKRLVTGLLALSLFLAACGQDSDQQKDGNKEKDDKAKTEQQDKKTNDSSKDKKDNKDDSKDVNKDNKDNSANDNQQQSNSNATNNDQNQTNNNQSSNNQANNNQKSSYVAPYYGQNAAPVARQIYPFNGNKNQALQQLPNFQTALNAANNEANKFGSNNKVYNDYSIEEHNGNYKYVFSFKDPNANGKYSIVTVDYTGQAMVTDPNYQQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00353", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMGKIMKPGKVVLVLRGKYAGRKAVVVKQQDEGVSDRTYPHAIIAGIDRYPLKVTKDMGKKKIEKRNKLKPFLKVVSYTHLLPTRYSVDVAFDKTNINKEALKAPSKKRKALVEVKSKFEERYKTGKNKWFFTKLRF\n\n### Sequence 2:\nMTAKSRILVLNGPNLNLLGLREPTHYGSQTLEQIVATLRDQAQKADIELEHLQSNREYELIEAIHQAFGKVDFIIINPAAFTHTSVALRDALLGVAIPFIEVHLSNVHAREPFRHHSYLSDKAQGVICGLGAQGYEFALSAAIRALQAKQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00354", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL1_COLP3\n\n### Answer:", "output": "50S ribosomal protein L1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00355", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "O=C(N[C@@H](Cc1ccccc1)[C@@H](O)C(=O)N1C[C@@H](O)[C@@H](O)C1)c1cc2cc(Cl)ccc2[nH]1", "output": "This molecule is medium-sized and polar, with moderately lipophilic character. It contains aromatic ring, amide, hydroxyl, halide, indole. Its descriptor profile: MW=457.91, LogP=1.09, TPSA=125.89, HBD=5, HBA=5, QED=0.38.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=C(N[C@@H](Cc1ccccc1)[C@@H](O)C(=O)N1C[C@@H](O)[C@@H](O)C1)c1cc2cc(Cl)ccc2[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 457.91, "LogP": 1.09, "TPSA": 125.89, "QED": 0.38, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide", "indole"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00356", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNSLLVNKAAIVTGGSRGIGFGIAKLFAEHGANVQIWGINEEAGKSAAQDLSDKTGSKVSFALVDVSKNDMVSAQVQKFLAEYGTIDVVVNNAGITRDSLLMRMSEEEWSSVIDTNLGSIYNVCSAVIRPMIKARSGAIVNISSIVGLRGSPGQTNYAAAKAGIIGFSKALSKEVGSKNIRVNCIAPGFIDTDMTKGLSDNLKNEWLKGVPLGRVGTPEEIAMAALFLASNQSSYITGQVLSVDGGMA\n\n### Sequence 2:\nLEGKVCLITGAASGIGKATTLLFAQEGATVIAGDISKENLDSLVKEAEGLPG-KVDPYVLNVTDRDQIKEVVEKVVQKYGRIDVLVNNAGITRDALLVRMKEEDWDAVINVNLKGVFNVTQMVVPYMIKQRNGSIVNVSSVVGIYGNPGQTNYAASKAGVIGMTKTWAKELAGRNIRVNAVAPGFIETPMTEKLPEKARETALSRIPLGRFGKPEEVAQVILFLASDESSYVTGQVIGIDGGL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00357", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSKKVLLVDDSAVLRKIVSFNLKKEGYEVIEAENGQIALEKLSEFTPDLIVLDIMMPVMDGFTVLKKLQEKEEWKRIPVIVLTAKGGEEDESLALSLGARKVMRKPFSPSQFIEEVKHLLNE\n\n### Answer:", "output": "<SEQ_2D>CCCEEEEECSCHHHHHHHHHHHHHTTCEEEEESSHHHHHHHTTTCCCSEEEECSCCSSSCHHHHHHHHHTCTTTTTSCEEEEESCCCHHHHHHHHHTTCSEEEESSCCHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00358", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIRISDAAQAHFAKLLANQEEGTQIRVFVINPGTPNAECGVSYCPPDAVEATDTALKFDLLTAYVDELSAPYLEDAEIDFVTDQLGSQLTLKAPNAKMRKVADDAPLMERVEYALQSQINPQLAGHGGRVSLMEITDEGYAILQFGGGCNGCSMVDVTLKEGIEKQLLNEFPELKGVRDLTEHQRGEHSYY\n\n### Sequence 2:\nMAIKQPKGKTVRRLGVNIYGNPKYDKLLDRKPNGPGKERGARKRGKTSVYGEQLKEKQKFRFAYGISERQFRNLYKKASRMPGVTGDNMISLMEQRLDNTIYRMGFAISRAQARQMVTHAYFFINGKPVNIPSMCVSVNDVITTKNKKGIQNLIRHNMSTSQSARGSWLTIDDEKLSATVNILPVTTDIQPVGNIQNVVEYYSRNA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00359", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nCD206:medium MRC1:medium APOE:medium\n</cell>", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as CD206, APOE, MRC1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD206", "MRC1", "APOE"]}}
{"id": "eval_00360", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSPSLTWHDVIGQEKEQPYFKDTLAYVAAERRAGKTIYPPQKDIFNAFRLTELDQVKVVILGQDPYHGPNQAHGLSFSVLPGVPAPPSLGNIYKELVTDIPGFQRPNHGFLQSWAEQGVLLLNTVLTVEAGKAHSHANLGWETFTDRVIAALNEHREGVIFMLWGSHAQKKGRIINTERHYILKAPHPSPLSAHRGFLGCKHFSQANQLLQQQNQQPIDWQPKLPAVE\n\n### Sequence 2:\nMSPSLTWHDVIGQEKEQPYFKDTLAYVAAERRAGKTIYPPQKDIFNAFRLTELDQVKVVILGQDPYHGPNQAHGLSFSVLPGVPAPPSLGNIYKELVTDIPGFQRPNHGFLQSWAEQGVLLLNTVLTVEAGKAHSHANLGWETFTDKVIAALNEHREGVIFMLWGSHAQKKGRIINTERHYILKAPHPSPLSAHRGFLGCKHFSQANQLLQQQNQQPIDWQPKLPAVE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00361", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMRIVAPVMPRHFDEAQAIDISKYEDVNLIEWRADFLPKDEIVAVAPAIFEKFAGKEIIFTLRTVQEGGNITLSSQEYVDIIKEINAIYNPDYIDFEYFTHKSVFQEMLDFPNLILSYHNFEETPENLMEAFSEMTKLAPRVVKIAVMPQSEQDVLDLMNYTRGFKTLNPEQEFATISMGKLGRLSRFAGDVIGSSWTYVSLDHVSGPGQVTLNDMKRIIEVLEMDISN\n\n### Sequence 2:\nMKIVVPVMPRSLEEAQEIDLSKFDSVDIIEWRADALPKDDIINVAPAIFEKFAGHEIIFTLRTTREGGNIVLSDAEYVELIQKINSIYNPDYIDFEYFSHKEVFQEMLEFPNLVLSYHNFQETPENIMEIFSELTALAPRVVKIAVMPKNEQDVLDVMNYTRGFKTINPDQVYATVSMSKIGRISRFAGDVTGSSWTFAYLDSSIAPGQITISEMKRVKALLDAD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00362", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nKQRTLSIIKPDALKKKVVGKIIDRFESNGLEVIAMKRLHLSVKDAENFYAIHRERPFFKDLIEFMVSGPVVVMVLEGKDAVAKNRDLMGATDPKLAGTIRADFAESIDANAVHGSDSLENAHNEIAFFFAARDL\n\n### Answer:", "output": "<SEQ_3Di>DAKKKKKQAVLCVVVVCPVVVVVLQVVPFKDWLDKDKAADDLVLLCVLVVVCPPPPCSVLLSCRRRVGIMMMTMIDGVVSLVVVDVQQFDLFLVPTCHQQVVRPDGSSGRGMDIDNDRVSRVVVCVSPDDPVSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00363", "category": "Mutation", "instruction": "Describe the effect of mutation Q13609-R82G on protein Deoxyribonuclease gamma.", "input": "### Wild-type:\nMSRELAPLLLLLLSIHSALAMRICSFNVRSFGESKQEDKNAMDVIVKVIKRCDIILVMEIKDSNNRICPILMEKLNRNSRRRITYNYVISSRLGRNTYKEQYAFLYKEKLVSVKRSYHYHDYQDGDADVFSREPFVVWFQSPHTAVKDFVIIPLHTTPETSVKEIDELVEVYTDVKHRWKAENFIFMGDFNAGCSYVPKKAWKNIRLRTDPRFVWLIGDQEDTTVKKSTNCAYDRIVLRGQEIVSSVVPKSNSVFDFQKAYKLTEEEALDVSDHFPVEFKLQSSRAFTNSKKSVTLRKKT\n\n### Mutant:\nMSRELAPLLLLLLSIHSALAMRICSFNVRSFGESKQEDKNAMDVIVKVIKRCDIILVMEIKDSNNRICPILMEKLNRNSRRGITYNYVISSRLGRNTYKEQYAFLYKEKLVSVKRSYHYHDYQDGDADVFSREPFVVWFQSPHTAVKDFVIIPLHTTPETSVKEIDELVEVYTDVKHRWKAENFIFMGDFNAGCSYVPKKAWKNIRLRTDPRFVWLIGDQEDTTVKKSTNCAYDRIVLRGQEIVSSVVPKSNSVFDFQKAYKLTEEEALDVSDHFPVEFKLQSSRAFTNSKKSVTLRKKT\n\n### Answer:", "output": "Decreases enzymatic activity. SNP enhanced the activity of the enzyme, being dissimilar to . These findings permit us to assume that, although only R206 exhibits polymorphisms in a Caucasian-specific manner, at least SNPs and in DNASE1L3 might be potential protective factors against autoimmune disease.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00364", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRUVA_THEM4\n\n### Answer:", "output": "Belongs to the RuvA family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00365", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL101_ARATH\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uL16 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00366", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGPNSDVYAQEKQDFVQHFSQIVRVLTEDEMGHPEIGDAIARLKEVLEYNAIGGKYNRGLTVVVAFRELVEPRKQDADSLQRAWTVGWCVELLQAFFLVADDIMDSSLTRRGQICWYQKPGVGLDAINDANLLEACIYRLLKLYCREQPYYLNLIELFLQSSYQTEIGQTLDLLTAPQGNVDLVRFTEKRYKSIVKYKTAFYSFYLPIAAAMYMAGIDGEKEHANAKKILLEMGEFFQIQDDYLDLFGDPSVTGKIGTDIQDNKCSWLVVQCLQRATPEQYQILKENYGQKEAEKVARVKALYEELDLPAVFLQYEEDSYSHIMALIEQYAAPLPPAVFLGLARKIYKRRK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCTTHHHHHHHHHHHHHHHHHHTTTTCGGGHHHHHHHHHHHHHHSSSSCCHHHHHHHHHHHHHSCGGGCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTCSEETTEECGGGSTTTGGGHHHHHHHHHHHHHHHHHHHHTTSTTHHHHHHHHHHHHHHHHHHHHHHHHHSCTTCCCGGGCCHHHHHHHHHHHTHHHHHHHHHHHHHHHTTCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHSCHHHHSSCCCTTTTTCCCHHHHHHHHSCCHHHHHHHHHHTTCCCHHHHHHHHHHHHHTTHHHHHHHHHHHHHHHHHHHHHHHCTTSCHHHHHHHHHHHCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00367", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: N#C[SH+][Co-4]([NH2])([NH2])([NH2])([NH2])[NH2].O=S([O-])(O)=[OH+]\nDescriptors: MW=296.22, LogP=-4.05, TPSA=235.65, HBD=6, HBA=8", "output": "Descriptors: MW=296.22, LogP=-4.05, TPSA=235.65, HBD=6, HBA=8, QED=0.09, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "N#C[SH+][Co-4]([NH2])([NH2])([NH2])([NH2])[NH2].O=S([O-])(O)=[OH+]", "compound_name": null, "pubchem_cid": null, "MW": 296.22, "LogP": -4.05, "TPSA": 235.65, "QED": 0.09, "functional_groups": ["amine", "hydroxyl", "nitrile"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00368", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BZR6-A279R on protein Reticulon-4 receptor.", "input": "### Wild-type:\nMKRASAGGSRLLAWVLWLQAWQVAAPCPGACVCYNEPKVTTSCPQQGLQAVPVGIPAASQRIFLHGNRISHVPAASFRACRNLTILWLHSNVLARIDAAAFTGLALLEQLDLSDNAQLRSVDPATFHGLGRLHTLHLDRCGLQELGPGLFRGLAALQYLYLQDNALQALPDDTFRDLGNLTHLFLHGNRISSVPERAFRGLHSLDRLLLHQNRVAHVHPHAFRDLGRLMTLYLFANNLSALPTEALAPLRALQYLRLNDNPWVCDCRARPLWAWLQKFAGSSSEVPCSLPQRLAGRDLKR\n\n### Mutant:\nMKRASAGGSRLLAWVLWLQAWQVAAPCPGACVCYNEPKVTTSCPQQGLQAVPVGIPAASQRIFLHGNRISHVPAASFRACRNLTILWLHSNVLARIDAAAFTGLALLEQLDLSDNAQLRSVDPATFHGLGRLHTLHLDRCGLQELGPGLFRGLAALQYLYLQDNALQALPDDTFRDLGNLTHLFLHGNRISSVPERAFRGLHSLDRLLLHQNRVAHVHPHAFRDLGRLMTLYLFANNLSALPTEALAPLRALQYLRLNDNPWVCDCRARPLWAWLQKFRGSSSEVPCSLPQRLAGRDLKR\n\n### Answer:", "output": "Mildly increases interaction with MAG.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00369", "category": "Mutation", "instruction": "Describe the effect of mutation P10275-C764Y on protein Androgen receptor.", "input": "### Wild-type:\nMEVQLGLGRVYPRPPSKTYRGAFQNLFQSVREVIQNPGPRHPEAASAAPPGASLLLLQQQQQQQQQQQQQQQQQQQQQQQETSPRQQQQQQGEDGSPQAHRRGPTGYLVLDEEQQPSQPQSALECHPERGCVPEPGAAVAASKGLPQQLPAPPDEDDSAAPSTLSLLGPTFPGLSSCSADLKDILSEASTMQLLQQQQQEAVSEGSSSGRAREASGAPTSSKDNYLGGTSTISDNAKELCKAVSVSMGLGVEALEHLSPGEQLRGDCMYAPLLGVPPAVRPTPCAPLAECKGSLLDDSAG\n\n### Mutant:\nMEVQLGLGRVYPRPPSKTYRGAFQNLFQSVREVIQNPGPRHPEAASAAPPGASLLLLQQQQQQQQQQQQQQQQQQQQQQQETSPRQQQQQQGEDGSPQAHRRGPTGYLVLDEEQQPSQPQSALECHPERGCVPEPGAAVAASKGLPQQLPAPPDEDDSAAPSTLSLLGPTFPGLSSCSADLKDILSEASTMQLLQQQQQEAVSEGSSSGRAREASGAPTSSKDNYLGGTSTISDNAKELCKAVSVSMGLGVEALEHLSPGEQLRGDCMYAPLLGVPPAVRPTPCAPLAECKGSLLDDSAG\n\n### Answer:", "output": "Highlights in PAIS and prostate cancer; complete gain of androgen binding. Mutations and had no residual dose-independent transactivational potential and substantial N-/C-terminal interaction. Slight thermolability of the transfected receptor protein, however, Moreover, the ability of the receptor to suppress a reporter gene is notably enhanced only when both alterations are present, suggesting that the extended glutamine heteropolymeric segment mitigates the enhancement of receptor function caused by the tyrosine to cysteine substitution. Binding of the synthetic androgen ligand, methyltrienolone (R1881), by the mutant receptor was increased by 54% compared to the normal receptor. Transcriptional suppression of a mouse mammary tumor virus-chloramphenicol acetyltransferase (MMTV-CAT) reporter gene by AR mutant was significant at 0. 1 nM R1881 and only 55% at 10 nM R1881 when compared to the minimal response with the normal AR, as assessed by CAT activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00370", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nHPPEIIRKPQNQGVRVGGVASFYCAARGDPPPSIVWRKNGKKVSGTQSRYTVLEQPGGISILRIEPVRAGRDDAPYECVAENGVGDAVSADATLTIYEGDKTPAGFPVITQGPGTRVIEVGHTVLMTCKAIGNPTPNIYWIKNQTKVDMSNPRYSLKDGFLQIENSREEDQGKYECVAENSMGTEHSKATNLYVKV\n\n### Answer:", "output": "<SEQ_3Di>DAKDKPDFWAAFEEEQQFKDKTKTAMADVVGWAKFKAFVRHTDDCVDPQWHWDDDPRRMIMIMGGRDHPPPQFTKMKIWIGSVPDHIDIHIYTYHYDYDPPADAFEKDWPDKFEEEEAEAQAKDKGDTHITGVVGWDKFKAFPRHGDDCVPPQWDDDVRIIMGGRDDQVPWTWMWMWTDDPRHIGIGDIYTYGYDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00371", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nIL7R CD3D TCF7 SELL CCR7 CD3E LTB\n</cell>", "output": "This cell is most consistent with naive CD4+ T cell, a T-cell subtype with naive/central-memory-like profile. Key markers such as SELL, CD3D, LTB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "skin", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IL7R", "CD3D", "TCF7", "SELL", "CCR7", "CD3E", "LTB"]}}
{"id": "eval_00372", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nSELL:high CCR7:medium CD8B:medium\n</cell>", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as CD8B, SELL, CCR7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "intestine", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SELL", "CCR7", "CD8B"]}}
{"id": "eval_00373", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nY433_ENDTX\n\n### Answer:", "output": "Belongs to the RapZ-like family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00374", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGCSH_ALIFM\n\n### Answer:", "output": "Glycine cleavage system H protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00375", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "OCc1cccc(Oc2ccccc2)c1", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring, hydroxyl, ether. Its descriptor profile: MW=200.24, LogP=2.97, TPSA=29.46, HBD=1, HBA=2, QED=0.82.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "OCc1cccc(Oc2ccccc2)c1", "compound_name": null, "pubchem_cid": null, "MW": 200.24, "LogP": 2.97, "TPSA": 29.46, "QED": 0.82, "functional_groups": ["aromatic ring", "hydroxyl", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 1.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 1.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00376", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKIIAGLGNPGQKYDKTKHNTGFMTLDHYLNEKGLSLDKDKFEGHWTKQKINGEDVILLEPQTYMNESGRSVSQIANFFKVAPEDVLIIQDDMDMPIGKIRIRANGKSGGHNGIKSIIRDLGTEKFNRLKIGIRHPKNTTVVSWVLTPFNDEQQKLMDDAFDTSVKIIDDFIAGRDSQYLMNKYN\n\n### Sequence 2:\nMKLIVALGNPGLKYEKTKHNTGFMALDHYLDEKGLRLDRDKFTALYAKEKVAGEDVIFMEPQTYMNESGRAVGAAAKFFKIDPSDILVIHDDMDMPIAKLRIRAGGKSGGHNGIKSIIACLGTEKFNRLKIGIRHPDKQSVVSWVLTPFNPDQQKELEASFAKVDQIIDDFIAGKDAQYLMNRYN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00377", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: HIV protease\nLigand SMILES: CNCCCC1c2ccccc2C=Cc2ccccc21\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "CNCCCC1c2ccccc2C=Cc2ccccc21", "compound_name": null, "pubchem_cid": null, "MW": 263.38, "LogP": 4.3, "TPSA": 12.03, "QED": 0.81, "functional_groups": ["aromatic ring", "amine"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00378", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nNDHI_NUPAD\n\n### Answer:", "output": "NDH shuttles electrons from NAD(P)H:plastoquinone, via FMN and iron-sulfur (Fe-S) centers, to quinones in the photosynthetic chain and possibly in a chloroplast respiratory chain.  The immediate electron acceptor for the enzyme in this species is believed to be plastoquinone.  Couples the redox reaction to proton translocation, and thus conserves the redox energy in a proton gradient.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00379", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "N#Cc1ccc(OCC(=O)O)c(-c2ccccc2)c1", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, hydroxyl, ether, nitrile. Its descriptor profile: MW=253.26, LogP=2.69, TPSA=70.32, HBD=1, HBA=3, QED=0.91.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "N#Cc1ccc(OCC(=O)O)c(-c2ccccc2)c1", "compound_name": null, "pubchem_cid": null, "MW": 253.26, "LogP": 2.69, "TPSA": 70.32, "QED": 0.91, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "nitrile"], "moleculenet_labels": {"label": -1.08}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00380", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKILVDENMPYAEELFRRLGDVQAVPGRPIPRDALVDADALMVRSVTKVNEALLHGTSIGFVGTATAGTDHVDDTWLRQQGIGFSAAPGCNAIAVVEYVFSALMMMAERDGFQLRDKTVGIIGVGNVGSRLNARLQALGVRTLLCDPPRADRGDNEAFWPLEKLVREADVLTFHTPLNKTGAYQSLHMADDELLAALPDGRILINACRGAVVDNAALLRALEKGKKLSVVLDVWEPEPDLSLPLLARVDIGTPHIAGYTLEGKARGTTQVFEAFSQHLGQPQSVELASLLPVPEFSHLRLNGELDEGKLKRLMHLVYDVRRDDAPLRHVAGLPGEFDRLRKHYQERREWSSLCVQCDDATSAGLLQQLGFTTQLL\n\n### Sequence 2:\nMKNLLDLSIEELKELVSPSFRATQIYEWVYKKNATEFSQMLNLPKDMRQDLAEKFYLDPLKCVKFEQSSDGSIKYLFELKDGLKIESVLLPMKEEISDEDGKISRHARYTICVSSQVGCKMGCAFCLTAKGGLVRNLTAGEIVGQILWIKRENKIPYERRINVVYMGMGEPLDNLTNVSKAIKILALNEGLAISPRRQTVSTSGLGSQIKKLGEMDLGVLLAISLHAVTNELRSRLMPINKAYNIEAVMDAVRGFPIDMRKRVMFEYLVIKDLNDSVSDAKKLVKLLHGIKAKVNLIYFNPHEGSEFGRPELASMLKFQEYLRDHGVTCTIRQSKGLDISAACGQLKQRNENPKFRANVSGNSAAKTEEKPTNDKTNVSKK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00381", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nRGKNPEDWRQQWSGPGTTKRFPETVLARCVKYTEIHPEMRHVDCQSVWDAFKGAFISKHPCDITEEDYQPLMKLGTQTVPCNKILLWSRIKDLAHQFTQVQRDMFTLEDTLLGYLADDLTWCGEFATSKINYQSCPDWRKDCSNNPVSVFWKTVSRRFAEAACDVVHVMLDGSRSKIFDKDSTFGSVEVHNLQPEKVQTLEAWVIHGGREDSRDLCQDPTIKELESIISKRNIQFSCKNIYRPDKFLQCVKNPEDSSCTSEI\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCSEECCCTTHHHHHHHHHHHHHHHSGGGTTCCHHHHHHHHHHHHTTSCTTCCCHHHHHHHHHHTCCCCCGGGBEEESSCHHHHHHHHHHSCSCBCGGGSHHHHHHTTCCCCBCTTSCCBCSSEECCTTTTCSSCHHHHHHHHHHHHHHHHCCEEEEEEEETTSSCSSCTTSHHHHTTGGGCCTTTEEEEEEEEECCCCCCCCCGGGCHHHHHHHHHHHTTTCEEEEEEESSHHHHHHHCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00382", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "KRT5, KRT14, TP63", "output": "This cell is most consistent with basal epithelial cell, a basal epithelial cell. Key markers such as TP63, KRT5, KRT14 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "basal epithelial cell", "tissue": "prostate", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["KRT5", "KRT14", "TP63"]}}
{"id": "eval_00383", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMDAFIRVANQSQGRDRLFRATQHACMLLRYLLESKAGKEAVVTKLKNLETSVSTGRKWFRLGNVLHAIQATEQSIQATDLVPRLCLTLANLNRVVYYICDTVLWAKSVGLTSGINREKWQMRAARHYYYFLLLSLVRDLYEVLLHMGQVARDRAKREKSSGDPPKYSVANEESEWLQSFLLLLFQSLKRNPPLFLDTVKNFCDILIPLNQLGIYKSNLGVVGFGGLVSSVAGLITVVYPQLKLKAR\n\n### Sequence 2:\nMDAWVRFSAQSQARERLCRAAQYACSLLGHVLQRHGASPELQKQIRQLESHLSLGRKLLRLGNSADALESAKRAVHLSDVVLRFCITVSHLNRALYFACDNVLWAGKSGLAPRVDQEKWAQRSFRYYLFSLIMNLSRDAYEIRLLMEQESSACSRRLKGSGGGVPGGSETGGLGGPGTPGGHLPQLALKLRLQVLLLARVLRGHPPLLLDVVRNACDLFIPLDKLGLWRCGPGIVGLCGLVSSILSILTLIYPWLRLK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00384", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMVVIANAHNELIHDAVLDYYGKRLATCSSDKTIKIFEVEGETHKLIDTLTGHEGPVWRVDWAHPKFGTILASCSYDGKVLIWKEENGRWSQIAVHAVHSASVNSVQWAPHEYGPLLLVASSDGKVSVVEFKENGTTSPIIIDAHAIGVNSASWAPATIEEDGEHNGTKESRKFVTGGADNLVKIWKYNSDAQTYVLESTLEGHSDWVRDVAWSPTVLLRSYLASVSQDRTCIIWTQDNEQGPWKKTLLKEEKFPDVLWRASWSLSGNVLALSGGDNKVTLWKENLEGKWEPAGEVHQ\n\n### Answer:", "output": "<SEQ_2D>CCEETTSCSSCEEEEEECTTSCEEEEEETTSCEEEEEESSSCEEEEEEECCCSSCEEEEEECCTTSCSEEEEEETTSCEEEEEEETTEEEEEEEECCCSSCEEEEEECCTTTCSEEEEEETTSEEEEEECCTTSCCCCEEEECCSSCEEEEEECCCCCCCCCCCCCCCCCCEEEEEETTSCEEEEEEETTTTEEEEEEEECCCSSCEEEEEECCCCSSSEEEEEEETTSCEEEEEESSSSSCCEEEESCSSCCSSCEEEEEECSSSCCEEEEETTSCEEEEEECTTSCEEECCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00385", "category": "Mutation", "instruction": "Describe the effect of mutation Q09328-E297A on protein Alpha-1,6-mannosylglycoprotein 6-beta-N-acetylglucosaminyltransferase A.", "input": "### Wild-type:\nMALFTPWKLSSQKLGFFLVTFGFIWGMMLLHFTIQQRTQPESSSMLREQILDLSKRYIKALAEENRNVVDGPYAGVMTAYDLKKTLAVLLDNILQRIGKLESKVDNLVVNGTGTNSTNSTTAVPSLVALEKINVADIINGAQEKCVLPPMDGYPHCEGKIKWMKDMWRSDPCYADYGVDGSTCSFFIYLSEVENWCPHLPWRAKNPYEEADHNSLAEIRTDFNILYSMMKKHEEFRWMRLRIRRMADAWIQAIKSLAEKQNLEKRKRKKVLVHLGLLTKESGFKIAETAFSGGPLGELVQ\n\n### Mutant:\nMALFTPWKLSSQKLGFFLVTFGFIWGMMLLHFTIQQRTQPESSSMLREQILDLSKRYIKALAEENRNVVDGPYAGVMTAYDLKKTLAVLLDNILQRIGKLESKVDNLVVNGTGTNSTNSTTAVPSLVALEKINVADIINGAQEKCVLPPMDGYPHCEGKIKWMKDMWRSDPCYADYGVDGSTCSFFIYLSEVENWCPHLPWRAKNPYEEADHNSLAEIRTDFNILYSMMKKHEEFRWMRLRIRRMADAWIQAIKSLAEKQNLEKRKRKKVLVHLGLLTKESGFKIAETAFSGGPLGALVQ\n\n### Answer:", "output": "Loss of catalytic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00386", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMQNDWIKLKEFFIYIFLFIDKTNVESITMWNLTQNEYLTLMVGVWIVILFLTWFLLWMVFKIVGYFK\n\n### Sequence 2:\nMPKLKTNRGAVKRFKVTGSGKIKRAASNHNHMLTKKSQKRKRRLRKIHEVAPSDMRAVSEMLRD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00387", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nASSY_BACC2\n\n### Answer:", "output": "Argininosuccinate synthase, Citrulline--aspartate ligase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00388", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: VEGFR2\nLigand SMILES: FC(F)(F)c1nnc2ccc(N3CCCCCC3)nn12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "FC(F)(F)c1nnc2ccc(N3CCCCCC3)nn12", "compound_name": null, "pubchem_cid": null, "MW": 285.27, "LogP": 2.52, "TPSA": 46.32, "QED": 0.81, "functional_groups": ["halide"], "moleculenet_labels": {"label": 3.29}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_00389", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: C=CCNC(=S)NC=C(c1ccccc1)S(=O)Cc1ccccc1\nDescriptors: MW=356.52, LogP=3.58, TPSA=41.13, HBD=2, HBA=2", "output": "Descriptors: MW=356.52, LogP=3.58, TPSA=41.13, HBD=2, HBA=2, QED=0.59, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "C=CCNC(=S)NC=C(c1ccccc1)S(=O)Cc1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 356.52, "LogP": 3.58, "TPSA": 41.13, "QED": 0.59, "functional_groups": ["aromatic ring", "amine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00390", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: Cc1ncc(-c2ccnc(Nc3ccc(C(=O)N4CCN(C)CC4)cc3)n2)n1C(C)C\nDescriptors: MW=419.53, LogP=3.36, TPSA=79.18, HBD=1, HBA=6", "output": "Descriptors: MW=419.53, LogP=3.36, TPSA=79.18, HBD=1, HBA=6, QED=0.68, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "Cc1ncc(-c2ccnc(Nc3ccc(C(=O)N4CCN(C)CC4)cc3)n2)n1C(C)C", "compound_name": null, "pubchem_cid": null, "MW": 419.53, "LogP": 3.36, "TPSA": 79.18, "QED": 0.68, "functional_groups": ["aromatic ring", "amide", "amine"], "moleculenet_labels": {"label": 2.1}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00391", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nJAL29_ARATH\n\n### Answer:", "output": "Belongs to the jacalin lectin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00392", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAROB_ALIF1\n\n### Answer:", "output": "3-dehydroquinate synthase, DHQS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00393", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nABCA3:low SFTPB:high SFTPC:high\n</cell>", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as ABCA3, SFTPC, SFTPB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["ABCA3", "SFTPB", "SFTPC"]}}
{"id": "eval_00394", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "C=CCC1([C@H](C)CCC)C(=O)NC(=O)NC1=O", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains amide. Its descriptor profile: MW=238.29, LogP=1.35, TPSA=75.27, HBD=2, HBA=3, QED=0.56.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "C=CCC1([C@H](C)CCC)C(=O)NC(=O)NC1=O", "compound_name": null, "pubchem_cid": null, "MW": 238.29, "LogP": 1.35, "TPSA": 75.27, "QED": 0.56, "functional_groups": ["amide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00395", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nHES5_RAT\n\n### Answer:", "output": "Nucleus.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00396", "category": "Mutation", "instruction": "Describe the effect of mutation Q9EQC4-T22A on protein Elongation of very long chain fatty acids protein 4.", "input": "### Wild-type:\nMGLLDSEPGSVLNAMSTAFNDTVEFYRWTWTIADKRVADWPLMQSPWPTISISTLYLLFVWLGPKWMKDREPFQMRLVLIIYNFGMVLLNLFIFRELFMGSYNAGYSYICQSVDYSNDVNEVRIAGALWWYFVSKGVEYLDTVFFILRKKNNQVSFLHVYHHCTMFTLWWIGIKWVAGGQAFFGAQMNSFIHVIMYSYYGLTAFGPWIQKYLWWKRYLTMLQLVQFHVTIGHTALSLYTDCPFPKWMHWALIAYAISFIFLFLNFYTRTYNEPKQSKTGKTATNGISSNGVNKSEKALEN\n\n### Mutant:\nMGLLDSEPGSVLNAMSTAFNDAVEFYRWTWTIADKRVADWPLMQSPWPTISISTLYLLFVWLGPKWMKDREPFQMRLVLIIYNFGMVLLNLFIFRELFMGSYNAGYSYICQSVDYSNDVNEVRIAGALWWYFVSKGVEYLDTVFFILRKKNNQVSFLHVYHHCTMFTLWWIGIKWVAGGQAFFGAQMNSFIHVIMYSYYGLTAFGPWIQKYLWWKRYLTMLQLVQFHVTIGHTALSLYTDCPFPKWMHWALIAYAISFIFLFLNFYTRTYNEPKQSKTGKTATNGISSNGVNKSEKALEN\n\n### Answer:", "output": "Loss of N-glycosylation. No effect on fatty acid elongase activity. Mutations at site in ELOVL4 lead to truncation and mislocalization of the translated protein from the ER, resulting in loss of condensation activity in the elongation of VLC-FAs.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00397", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: HIV protease\nLigand SMILES: COC(=O)c1ccc(COC(=O)c2cc(C(=CCCC3CCC4(C)C(CCC5C4CCC4(C)C(C(C)CCCC(C)C)CCC54)C3)c3cc(Cl)c(OCc4ccc(C(=O)OC)cc4)c(C(=O)OCc4ccc(C(=O)OC)cc4)c3)cc(Cl)c2OCc2ccc(C(=O)OC)cc2)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "COC(=O)c1ccc(COC(=O)c2cc(C(=CCCC3CCC4(C)C(CCC5C4CCC4(C)C(C(C)CCCC(C)C)CCC54)C3)c3cc(Cl)c(OCc4ccc(C(=O)OC)cc4)c(C(=O)OCc4ccc(C(=O)OC)cc4)c3)cc(Cl)c2OCc2ccc(C(=O)OC)cc2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 1360.52, "LogP": 18.96, "TPSA": 176.26, "QED": 0.04, "functional_groups": ["aromatic ring", "ester", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00398", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "CD8B, PRF1, GZMB, GZMA, CD8A", "output": "This profile is most consistent with cytotoxic T cell, which is commonly found in pancreas under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "pancreas", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CD8B", "PRF1", "GZMB", "GZMA", "CD8A"]}}
{"id": "eval_00399", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDBPA_LISMO\n\n### Answer:", "output": "ATP-dependent RNA helicase DbpA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00400", "category": "Mutation", "instruction": "Describe the effect of mutation Q9X758-G83A on protein Serine-type anaerobic sulfatase-maturating enzyme.", "input": "### Wild-type:\nMLNIAALRQQQIPLAAEPRSPVPFHILMKPIGPACNLACRYCYYPQDETPVNKMDDARLEQFIRRYIAAQPAGAREINFVWQGGEPLLAGLSFYKKALALQARYAPDGVTISNSLQTNGTLINDAWCRLFREHGFIIGLGLEGNEALQDYHRPDKRGRSTWSAALRGIDLLHQHQVDFNLLVVVHNEMAAHAAAIYVRLVSLGARYLQFQPLMSEGAALREGYQLSADNWGRFMVGIWRQWRKRCDRGRVFVINIEQAWAQYFTHTSGSCVHSARCGSNLVMESDGQLYACDHLINTEHR\n\n### Mutant:\nMLNIAALRQQQIPLAAEPRSPVPFHILMKPIGPACNLACRYCYYPQDETPVNKMDDARLEQFIRRYIAAQPAGAREINFVWQAGEPLLAGLSFYKKALALQARYAPDGVTISNSLQTNGTLINDAWCRLFREHGFIIGLGLEGNEALQDYHRPDKRGRSTWSAALRGIDLLHQHQVDFNLLVVVHNEMAAHAAAIYVRLVSLGARYLQFQPLMSEGAALREGYQLSADNWGRFMVGIWRQWRKRCDRGRVFVINIEQAWAQYFTHTSGSCVHSARCGSNLVMESDGQLYACDHLINTEHR\n\n### Answer:", "output": "Drastic reduction in AtsA activity. Abolition of binding to AtsA. No binding to SAM. Mutation of G83 fully impaired FG formation, indicating that G83 is crucial for AtsB function.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00401", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSSVTMRDMLKAGVHFGHQTRYWNPKMNTYIFGARNKIHIINLEQTLPLFNDAMAFLNKLAASNNKILFVGTKRAAQKAVSEEAQRCGMPYVDHRWLGGMLTNWKTIRQSIKRYRELEAQFQDGTFDKLTKKEALMRKREMDKLERSIGGIKDMGGLPDALFVIDVDHEDIALQEARKLGIPVVAVVDTNSNPDGVDYVIPGNDDAIRAIQLYVKAAADVILEGKQYAQNQTGGESEFVEVADEAAGEKAEG\n\n### Sequence 2:\nMAHQAHAYHMVDPSPWPLTGAIAALLLTSGTAVWFHFHSLTLLTMGNILLLLTMYQWWRDIIREGTFQGHHTPPVQKGLRYGMILFITSEVFFFLGFFWAFYHSSLSPTPELGGCWPPTGIITLDPFEVPLLNTAVLLASGVTVTWAHHSIMEGERKQTIQALTLTILLGFYFTFLQGMEYYEAPFTIADGVYGSTFFVATGFHGLHVIIGSTFLAICLLRQIQYHFTSEHHFGFEAAAWYWHFVDVVWLFLYVSIYWWGS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00402", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nYBEY_YERPE\n\n### Answer:", "output": "Single strand-specific metallo-endoribonuclease involved in late-stage 70S ribosome quality control and in maturation of the 3' terminus of the 16S rRNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00403", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMSTQSNRNALVVAQLKGDFVAFLFVLWKALNLPVPTKCQIDMAKVLANGDNKKFILQAFRGIGKSFITCAFVVWSLWRDPQLKILIVSASKERADANSIFIKNIIDLLPFLSELKPRPGQRDSVISFDVGPANPDHSPSVKSVGITGQLTGSRADIIIADDVEIPSNSATMGAREKLWTLVQEFAALLKPLPSSRVIYLGTPQTEMTLYKELEDNRGYTTIIWPALYPRTREENLYYSQRLAPMLRAEYDENPEALAGTPTDPVRFDRDDLRERELEYGKAGFTLQFMLNPNLSDAEKYPLRLRDAIVAALDLEKAPMHYQWLPNRQNIIEDLPNVGLKGDDLHTYHDCSNNSGQYQQKILVIDPSGRGKDETGYAVLYTLNGYIYLMEAGGFRDGYSDKTLELLAKKAKQWGVQTVVYESNFGDGMFGKVFSPILLKHHNCAMEEIRARGMKEMRICDTLEPVMQTHRLVIRDEV\n\n### Answer:", "output": "<SEQ_3Di>DLVVVLVVCLPPLLVVQQPQFWQADPPPGTDRDDAFPQLSVVLVCQDLVHDFAAEEAEFAPDVNLLSLLSNLLSQQLNPPQAAEEEEEAQDPVLVVSPVSNVSRLVRGDVSSHQDWDDDCPPVDAKTQPVRSHYIYGRDYLLVPVVVLVVDDRHAEYEYEQVQPRPPCVSNVVSCCQCPPVHCNVVCSVVVRRGHYYYYHFHEACPPVVVVLLVCLLQFFDDYDFPVSPVVCQADPVRHGDSNPVVVVVVCPVHHPVVCCNHTNSDHDPAFFFLAGVVLLVVAAADDADDDVFQKDFQDFADPPWAKAKFWECDPDLQAWIWIWIWGPRDQAIETGMITTDPPCPDLPCVLVVVQVVCVRRVLHAYAYEPDPSGQVSCQCCPCPRNVNSPRQDAPDPSGRHDDPPDRRLLSNLSVLVNVSSVVRSYHYHHVVLSVQSVAQGDPGSHTGGPPPGHCNHVSRSSSVSSVVVDPVVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00404", "category": "Mutation", "instruction": "Describe the effect of mutation P10644-D213A on protein cAMP-dependent protein kinase type I-alpha regulatory subunit.", "input": "### Wild-type:\nMESGSTAASEEARSLRECELYVQKHNIQALLKDSIVQLCTARPERPMAFLREYFERLEKEEAKQIQNLQKAGTRTDSREDEISPPPPNPVVKGRRRRGAISAEVYTEEDAASYVRKVIPKDYKTMAALAKAIEKNVLFSHLDDNERSDIFDAMFSVSFIAGETVIQQGDEGDNFYVIDQGETDVYVNNEWATSVGEGGSFGELALIYGTPRADTVKAKTNVKLWGIDRDSYRRILMGSTLRKRKMYEEFLSKVSILESLDKWERLTVADALEPVQFEDGQKIVVQGEPGDEFFIILEGSA\n\n### Mutant:\nMESGSTAASEEARSLRECELYVQKHNIQALLKDSIVQLCTARPERPMAFLREYFERLEKEEAKQIQNLQKAGTRTDSREDEISPPPPNPVVKGRRRRGAISAEVYTEEDAASYVRKVIPKDYKTMAALAKAIEKNVLFSHLDDNERSDIFDAMFSVSFIAGETVIQQGDEGDNFYVIDQGETDVYVNNEWATSVGEGGSFGELALIYGTPRAATVKAKTNVKLWGIDRDSYRRILMGSTLRKRKMYEEFLSKVSILESLDKWERLTVADALEPVQFEDGQKIVVQGEPGDEFFIILEGSA\n\n### Answer:", "output": "Decreases CNC1 activity, leading to reduced PKA activity attributed to increased binding to cAMP and/or the catalytic subunit; enhances protein degradation. The expressed mutation p. in the RIalpha protein exhibited decreased PKA activity, which we attribute to increased binding to cAMP and/or the catalytic subunit. The mutation in the PRKAR1A gene has been observed in acrodysostosis with hormonal resistance. This mutation is located in one of the cAMP-binding domains of the PRKAR1A protein. Expression of the mutant results in an increase of forskolin-induced PKA activation and an enhanced ability of cAMP to dissociate PRKAR1A from the catalytic PKA subunits. The mutation enhances cAMP binding locally in the domain containing the mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00405", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "APOB, FABP2, ALPI", "output": "The expression pattern is consistent with enterocyte in a context associated with type 2 diabetes, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "enterocyte", "tissue": "bone marrow", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["APOB", "FABP2", "ALPI"]}}
{"id": "eval_00406", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKTSLQIAAEARLEPIAAIAERLGLPVRYLEPYGRYRGKIDLTFLDDYHDRPLGRYVLVSAITPTPLGEGKTTTAIGLAMALNRIGKRAAVTLRQSSLGPVFGIKGGGAGGGYSQIVPLVESILHLNGDIHAVSQAHNQLAALTDNSWYHGNPLDIDPDRIEIRRVVDVNDRFLRQVMIGLGGKQNGFPRQTGFDISVASELMAILAMVNGVGARAALRDLRSRIGRMVVAFRRDGTPITAEDVRGAGAATVLMREALKPNLMQTIENTPALIHAGPFANIAQGNSSILADLIALRCADYVVTEAGFGVDIGAEKFFNLKCRASGLWPDVAVIVATIRALKAHSGKYDIVAGKPLPPALLHENPDDVISGGANLRRQIENLHQFKVPVIVALNAYPEDTPAEIDAVAHIATTAGAAGMAVSNVYAAGSAGGVDLARLVIEIAERPGPRPVQFLYPLEWSLADKITTIAHRIYGAAAVTFSPTAAAQLAALEDAGFGNLPICMAKTHLSLSHDPALRGAPEGFTFPIREVRLSAGAGFILPIAGTTVTMPGLGAHPAAHQIDIDDEGNIVGLF\n\n### Sequence 2:\nMMLQTSTILQLLLFLVGSVSAYNILVFSPATSKSHLISNGRIADELARAGHNVTLLEIDFLGIVDTTKSAKLVKKTIVRTPKGMQGFRNVLQGFSEIVMEDPGLWGLVEGNIMYQNAYNALCEEFLEMDDIFQELKAQNFDGFFAEQLNICGFGYAKALGIERRFLISSCPYFSHVYDYTSHPAPYASVPFVADMSPEPTYFERAQNLLNGFTCNMLFRYMHTRLSFIFRNKFGQDFPSVPEIVRNADIIFLATDEIIDFSAPTLPNLVNIGGLGVDDDTTEMEPVFEAEMKKGEKGVILFSLGTIANTSTIDKKVMESFLGIVKKFPDYHFLIRADKYDKNTKERAKGISNVFVSDWLPQPAILHHPRLRTFITHAGYNGLVEAARAGVPLITIPFMFDQNLNSRAIEKKGWGIRSDKKKLLNDPDSFEADLKEMLTNPSYTKNAHRIRDLIKSKPLGARDRFIKTTEWVIQNGGVRELLTEGRDLSIISSYNLDIIVPVLFVLLYCLIIPFFKLIGGFYYYSCFGHIESKHKLD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00407", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nFABP2 VIL1 ALPI APOB\n</cell>", "output": "This cell is most consistent with enterocyte, an intestinal enterocyte. Key markers such as FABP2, APOB, VIL1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "enterocyte", "tissue": "liver", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["FABP2", "VIL1", "ALPI", "APOB"]}}
{"id": "eval_00408", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCHLI_SYNY3\n\n### Answer:", "output": "Magnesium-chelatase subunit ChlI, Mg-protoporphyrin IX chelatase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00409", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "c1ccc(NCCCOc2cccc(CN3CCCCC3)c2)nc1", "output": "This molecule is medium-sized and relatively nonpolar, with lipophilic character. It contains aromatic ring, amine, ether, pyridine. Its descriptor profile: MW=325.46, LogP=3.95, TPSA=37.39, HBD=1, HBA=4, QED=0.75.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "c1ccc(NCCCOc2cccc(CN3CCCCC3)c2)nc1", "compound_name": null, "pubchem_cid": null, "MW": 325.46, "LogP": 3.95, "TPSA": 37.39, "QED": 0.75, "functional_groups": ["aromatic ring", "amine", "ether", "pyridine"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00410", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nCLDN5:medium PECAM1:low KDR:high\n</cell>", "output": "This cell is most consistent with endothelial cell, a vascular endothelial cell. Key markers such as KDR, CLDN5, PECAM1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "endothelial cell", "tissue": "brain", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CLDN5", "PECAM1", "KDR"]}}
{"id": "eval_00411", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BXI2-Q179R on protein Mitochondrial ornithine transporter 2.", "input": "### Wild-type:\nMKSGPGIQAAIDLTAGAAGGTACVLTGQPFDTIKVKMQTFPDLYKGLTDCFLKTYAQVGLRGFYKGTGPALMAYVAENSVLFMCYGFCQQFVRKVAGMDKQAKLSDLQTAAAGSFASAFAALALCPTELVKCRLQTMYEMEMSGKIAKSHNTIWSVVKGILKKDGPLGFYHGLSSTLLQEVPGYFFFFGGYELSRSFFASGRSKDELGPVHLMLSGGVAGICLWLVVFPVDCIKSRIQVLSMYGKQAGFIGTLLSVVRNEGIVALYSGLKATMIRAIPANGALFVAYEYSRKMMMKQLEA\n\n### Mutant:\nMKSGPGIQAAIDLTAGAAGGTACVLTGQPFDTIKVKMQTFPDLYKGLTDCFLKTYAQVGLRGFYKGTGPALMAYVAENSVLFMCYGFCQQFVRKVAGMDKQAKLSDLQTAAAGSFASAFAALALCPTELVKCRLQTMYEMEMSGKIAKSHNTIWSVVKGILKKDGPLGFYHGLSSTLLREVPGYFFFFGGYELSRSFFASGRSKDELGPVHLMLSGGVAGICLWLVVFPVDCIKSRIQVLSMYGKQAGFIGTLLSVVRNEGIVALYSGLKATMIRAIPANGALFVAYEYSRKMMMKQLEA\n\n### Answer:", "output": "Ornithine homo-exchange is enhanced 18-fold. Vmax value 33-fold higher than wild-type. The recombinant, purified mitochondrial solute carrier SLC25A2 efficiently transports ADMA in addition to its known substrates arginine, lysine, and ornithine. The SLC25A2-mediated ADMA transport followed first-order kinetics and was highly specific as symmetric dimethylarginine (SDMA) was not transported at all. ADMA inhibited SLC25A2 activity with an inhibition constant of 0. 38 ± 0. 04 mM.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00412", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNLKEFVVDVKDFPQKGIIFKDITPLLNNKDAFKYMIDTIAEFVKQLDVDVIVAPEARGFLLASAVAYATNKRFVLVRKPNKLPREVYDVEYSLEYGTNHQQIHKEDLKPNDKVVVIDDVLATGGTMQAIIDLVKLSKAEVVGMSFLIDLTFLHKEDLFSEYQVQKLIKY\n\n### Sequence 2:\nMAEKENTKRNRREEILQALAQMLESSDGSQRITTAKLAANVGVSEAALYRHFPSKTRMFDSLIEFIEDSLMSRINLILQDEKETFNRLRLILLLVLGFAERNPGLTRIMTGHALMFEQDRLQGRINQLFERIEVQLRQVLREKKLRDGQGFIHDEALLATQLLAFCEGMLSRFVRSEFRYRPTQEFDARWPLIVAQLQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00413", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nIF3_PARMW\n\n### Answer:", "output": "Translation initiation factor IF-3.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00414", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "CCC[N+]1(C)CCCCC1.F[B-](F)(F)F", "output": "The SMILES represents an unnamed molecule with MW=229.07, LogP=3.33, TPSA=0.0. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "CCC[N+]1(C)CCCCC1.F[B-](F)(F)F", "compound_name": null, "pubchem_cid": null, "MW": 229.07, "LogP": 3.33, "TPSA": 0.0, "QED": 0.39, "functional_groups": ["halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00415", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMFKSPFGANANPFGGASDGATVGGNAMHQVIEEEENDTVTSPTSPNFGMNAQSMFSGPFGGDASDDVLPSALRSPPNPESYPAQYNFSRRTSVSAESLKPSADTYDNWTPPVHDKTNEQLSRLKTAIAGNFLFSHLDDEQSAQILGALIEKPIPAKDIKVISQGDAGDYFYVVEKGSFDVYVNEKGTLQPGPEGMGEKVGTIQAGGSFGELALMYNAPRAATVISAEPGCTLWALDRLTFRRILMESTFSRRRMYEDFLREVPLLQTLTPYERSKIADALETQKYTPGATIIKEGDPGHSFYLLESGEADAYLGDGKESVKHYSKGDFFGELALLNDAPRAASIVATTDVKVASLGKSAFQRLLGPVEGIMRRTKYDDIKTGVEEMDPLQV\n\n### Sequence 2:\nAAFQKISEEDEYEVTSPTDPTFRSANAAAASSSTGSPFFGGSYGENSGPIRFNRSPFDNGPREEDEEGADEFPPEDIRPTGAANQGFPNNYALGRRTSVSAESLNPTSAGSDSWTPPYHEKTEEQLSRLKTAVSSNFLFSHLDDDQFKSVLDALVEKPIPAKGIKVISQGDAGDYFYIVENGHFDFMIHPSGSVQPGPDGMGNKVGSVGPGGSFGELALMYNAPRAATVVSVDPKSTLWALDRITFRRILMDSAFQRRRMYEAFLEEVPLLSSLKPYERAKIADALDAIKYPAGSTIIAEGDPGDAFYLLESGEADAFKNGVEGPVKSYKRGDYFGELALLDDKPRAASIVAKTDVKVAKLGRDGFKRLLGPVEDIMRRAEYE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00416", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: BACE1\nLigand SMILES: CCc1ccc(O)c(OC)c1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CCc1ccc(O)c(OC)c1", "compound_name": null, "pubchem_cid": null, "MW": 152.19, "LogP": 1.96, "TPSA": 29.46, "QED": 0.7, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_00417", "category": "Mutation", "instruction": "Describe the effect of mutation Q9FNE9-I95A on protein Histone-lysine N-methyltransferase ATXR6.", "input": "### Wild-type:\nMVAVRRRRTQASNPRSEPPQHMSDHDSDSDWDTVCEECSSGKQPAKLLLCDKCDKGFHLFCLRPILVSVPKGSWFCPSCSKHQIPKSFPLIQTKIIDFFRIKRSPDSSQISSSSDSIGKKRKKTSLVMSKKKRRLLPYNPSNDPQRRLEQMASLATALRASNTKFSNELTYVSGKAPRSANQAAFEKGGMQVLSKEGVETLALCKKMMDLGECPPLMVVFDPYEGFTVEADRFIKDWTIITEYVGDVDYLSNREDDYDGDSMMTLLHASDPSQCLVICPDRRSNIARFISGINNHSPEGR\n\n### Mutant:\nMVAVRRRRTQASNPRSEPPQHMSDHDSDSDWDTVCEECSSGKQPAKLLLCDKCDKGFHLFCLRPILVSVPKGSWFCPSCSKHQIPKSFPLIQTKAIDFFRIKRSPDSSQISSSSDSIGKKRKKTSLVMSKKKRRLLPYNPSNDPQRRLEQMASLATALRASNTKFSNELTYVSGKAPRSANQAAFEKGGMQVLSKEGVETLALCKKMMDLGECPPLMVVFDPYEGFTVEADRFIKDWTIITEYVGDVDYLSNREDDYDGDSMMTLLHASDPSQCLVICPDRRSNIARFISGINNHSPEGR\n\n### Answer:", "output": "Loss of methylation activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00418", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "S100A4, CD3D, CD4, CD45RO", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD45RO, CD3D, S100A4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "spleen", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["S100A4", "CD3D", "CD4", "CD45RO"]}}
{"id": "eval_00419", "category": "Mutation", "instruction": "Describe the effect of mutation O35627-F175N on protein Nuclear receptor subfamily 1 group I member 3.", "input": "### Wild-type:\nMTAMLTLETMASEEEYGPRNCVVCGDRATGYHFHALTCEGCKGFFRRTVSKTIGPICPFAGRCEVSKAQRRHCPACRLQKCLNVGMRKDMILSAEALALRRARQAQRRAEKASLQLNQQQKELVQILLGAHTRHVGPMFDQFVQFKPPAYLFMHHRPFQPRGPVLPLLTHFADIFTFMVQQIIKFTKDLPLFRSLTMEDQISLLKGAAVEILHISLNTTFCLQTENFFCGPLCYKMEDAVHAGFQYEFLESILHFHKNLKGLHLQEPEYVLMAATALFSPDRPGVTQREEIDQLQEEMAL\n\n### Mutant:\nMTAMLTLETMASEEEYGPRNCVVCGDRATGYHFHALTCEGCKGFFRRTVSKTIGPICPFAGRCEVSKAQRRHCPACRLQKCLNVGMRKDMILSAEALALRRARQAQRRAEKASLQLNQQQKELVQILLGAHTRHVGPMFDQFVQFKPPAYLFMHHRPFQPRGPVLPLLTHFADINTFMVQQIIKFTKDLPLFRSLTMEDQISLLKGAAVEILHISLNTTFCLQTENFFCGPLCYKMEDAVHAGFQYEFLESILHFHKNLKGLHLQEPEYVLMAATALFSPDRPGVTQREEIDQLQEEMAL\n\n### Answer:", "output": "Enhanced binding of coactivator NCOA2 in the absence of TCPOBOP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00420", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nEFP_BACC3\n\n### Answer:", "output": "Elongation factor P, EF-P.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00421", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "CCCCC(C)(C)C(O)C=CC1C(O)CC(=O)C1CCCCC=CC(=O)OC", "output": "The molecule contains the following functional groups: ester, hydroxyl, ether, ketone.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "SIDER", "canonical_smiles": "CCCCC(C)(C)C(O)C=CC1C(O)CC(=O)C1CCCCC=CC(=O)OC", "compound_name": null, "pubchem_cid": null, "MW": 394.55, "LogP": 3.98, "TPSA": 83.83, "QED": 0.23, "functional_groups": ["ester", "hydroxyl", "ether", "ketone"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 0, "Product issues": 0, "Eye disorders": 0, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 1, "Vascular disorders": 1, "Blood and lymphatic system disorders": 0, "Skin and subcutaneous tissue disorders": 0, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 0, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 0, "Renal and urinary disorders": 0, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 0, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["ester", "hydroxyl", "ether", "ketone"]}}
{"id": "eval_00422", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMERKIEHFQMTYQSFKDLSVEAKLSYTFNWFGDYSSGGFTAKKGDKHYFDLFLKIKPDAKKSFTIANFKTEEDNSTTIGGKETTRNLEWIEFSASLKFSLKGKDDVSQKSVNKFLSTFANNTEGYSSNINLFIYLEYLIK\n\n### Sequence 2:\nMERKVDHFQMTYQSFKDLLVEARLSYTFNWFGDYSSGDFTAKRGDKHYFYLFLKIKSDPKKQFSAKKFLTEGEAFTDQEGKQTTRNLEWIEFSASISWWTKGKDDVSQKSLKKFLESFATNT-GYSSDINLFSYLEYLIK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00423", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNVPSSAERWRSAMLYAWVSAGSVVGGLTRYLVGLALDTGPGFPFATLFINATGSLIIGFYATLTGPDGRMLARPEHRQFVMTGFCGGYTTFSAFSLETFRLFHGGMKYIALAYVASSVVCWLVSVWLGHIMASRYNRLKRS\n\n### Sequence 2:\nMGEHQNVKLNRTVKAAVLTVSDTRDFDTDKGGQCVRQLLQADDVEVSDAHYTIVKDEKVAITTQVKKWLEEDIDVIITTGGTGIAQRDVTIEAVKPLLTKEIEGFGELFRYLSYVEDVGTRALLSRAVAGTVNNKLIFSIPGSTGAVKLALEKLIKPELNHLIHELTK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00424", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "COc1ccc(Oc2ncccc2C(=O)NCc2ccccc2Cl)cc1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, ether, halide, pyridine. Its descriptor profile: MW=368.82, LogP=4.47, TPSA=60.45, HBD=1, HBA=4, QED=0.7.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "COc1ccc(Oc2ncccc2C(=O)NCc2ccccc2Cl)cc1", "compound_name": null, "pubchem_cid": null, "MW": 368.82, "LogP": 4.47, "TPSA": 60.45, "QED": 0.7, "functional_groups": ["aromatic ring", "amide", "ether", "halide", "pyridine"], "moleculenet_labels": {"label": 3.59}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00425", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNEEVSRPTRVALEIRKTLMEILHRDTKDPRLEKVNITECKISKDLSIATVYFTIIGVNEKDQPAQDAIKALEKAKGFFRSEIGKRMNLRITPEVRFFYDTVAENASHIEELIFKALHQKK\n\n### Sequence 2:\nRAERVGEQMKKELMDIINNKVKDPRIGFITITDVQVTNDLSLAKVYLTVLG-NEKE--VDDTFKALEKAKGFIKSELGSRMRLRIVPELQFEYDHSIEYGNKIEKMI-QDLHKK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00426", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "N[C@@H](CCC(=O)N[C@@H](CSSC[C@H](NC(=O)CC[C@H](N)C(=O)O)C(=O)NCC(=O)O)C(=O)NCC(=O)O)C(=O)O", "output": "This molecule is relatively large and polar, with hydrophilic character. It contains amide, amine, hydroxyl. Its descriptor profile: MW=612.64, LogP=-3.88, TPSA=317.64, HBD=10, HBA=12, QED=0.04.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "N[C@@H](CCC(=O)N[C@@H](CSSC[C@H](NC(=O)CC[C@H](N)C(=O)O)C(=O)NCC(=O)O)C(=O)NCC(=O)O)C(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 612.64, "LogP": -3.88, "TPSA": 317.64, "QED": 0.04, "functional_groups": ["amide", "amine", "hydroxyl"], "moleculenet_labels": {"NR-AR": 1.0, "NR-AR-LBD": 1.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 1.0, "NR-ER-LBD": 1.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00427", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: Cc1ccccc1OP(=S)(Oc1ccccc1C)Oc1ccccc1C\nDescriptors: MW=384.44, LogP=6.37, TPSA=27.69, HBD=0, HBA=4", "output": "Descriptors: MW=384.44, LogP=6.37, TPSA=27.69, HBD=0, HBA=4, QED=0.46, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "Cc1ccccc1OP(=S)(Oc1ccccc1C)Oc1ccccc1C", "compound_name": null, "pubchem_cid": null, "MW": 384.44, "LogP": 6.37, "TPSA": 27.69, "QED": 0.46, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00428", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMPKKNEAPASFEKALSELEQIVTRLESGDLPLEEALNEFERGVQLARQGQAKLQQAEQRVQILLSDNEDASLTPFTPDNE\n\n### Sequence 2:\nQXPSTAELCKINSNACSVPFSXIPCQKXFLAACDRHDTCYHCGKHFGFKQDDCDDAFFRDMTALCAHGTDDEGXCPX\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00429", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMARGNPRKRAAAAKDKWKMKEWYIVYAPDFFGSKEIGLTPADDPEKVIGRVIETTLKDLTGDFTKGHVKLYFQVYDVKGQNAYTKFKGHTLARSYIRSLVRRRTTRVDGIFNITTKDGYKLRVMGMVIAYRRIQTSQERAIRKIIQDIIYKKAEELNFADFVLQSVNGQIASEIAKEARKIYPIKRAEVRKIKVLAEPSA\n\n### Sequence 2:\nMARGNPRKRAAAAKDKWKMKEWYIVYAPDFFGSKEIGLTPADDPEKVIGRVIETTLKDLTGDFTKGHVKLYFQVYDVKGQNAYTKFKGHTLARSYIRSLVRRRTTRVDGIFNITTKDGYKLRVMGMVIAYRRIQTSQERAIRKIIQDIIYKKAEELNFADFVLQSVNGQIASEIAKEARKIYPIKRAEVRKIKVLAEPSA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00430", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "COc1ccc(C(C)=O)cc1", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring, ether, ketone. Its descriptor profile: MW=150.18, LogP=1.9, TPSA=26.3, HBD=0, HBA=2, QED=0.6.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "COc1ccc(C(C)=O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 150.18, "LogP": 1.9, "TPSA": 26.3, "QED": 0.6, "functional_groups": ["aromatic ring", "ether", "ketone"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00431", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UNH5-C278S on protein Dual specificity protein phosphatase CDC14A.", "input": "### Wild-type:\nMAAESGELIGACEFMKDRLYFATLRNRPKSTVNTHYFSIDEELVYENFYADFGPLNLAMVYRYCCKLNKKLKSYSLSRKKIVHYTCFDQRKRANAAFLIGAYAVIYLKKTPEEAYRALLSGSNPPYLPFRDASFGNCTYNLTILDCLQGIRKGLQHGFFDFETFDVDEYEHYERVENGDFNWIVPGKFLAFSGPHPKSKIENGYPLHAPEAYFPYFKKHNVTAVVRLNKKIYEAKRFTDAGFEHYDLFFIDGSTPSDNIVRRFLNICENTEGAIAVHCKAGLGRTGTLIACYVMKHYRFT\n\n### Mutant:\nMAAESGELIGACEFMKDRLYFATLRNRPKSTVNTHYFSIDEELVYENFYADFGPLNLAMVYRYCCKLNKKLKSYSLSRKKIVHYTCFDQRKRANAAFLIGAYAVIYLKKTPEEAYRALLSGSNPPYLPFRDASFGNCTYNLTILDCLQGIRKGLQHGFFDFETFDVDEYEHYERVENGDFNWIVPGKFLAFSGPHPKSKIENGYPLHAPEAYFPYFKKHNVTAVVRLNKKIYEAKRFTDAGFEHYDLFFIDGSTPSDNIVRRFLNICENTEGAIAVHSKAGLGRTGTLIACYVMKHYRFT\n\n### Answer:", "output": "Loss of phosphatase activity. Phosphorylation of hCdh1 at alters its conformation and prevents it from activating APC.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00432", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGLPK_BACAA\n\n### Answer:", "output": "Key enzyme in the regulation of glycerol uptake and metabolism.  Catalyzes the phosphorylation of glycerol to yield sn-glycerol 3-phosphate.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00433", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nXMQVWPIEGIKKFETLSYLPPLTVEDLLKQIEYLLRSKWVPCLEFSKVGFVYRENHRSPGYYDGRYWTMWKLPMFGCTDATQVLKELEEAKKAYPDAFVRIIGFDNVRQVQCISFIAYKPPGCEESGGN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCSSSCCCCSTTTTSCCCCHHHHHHHHHHHHHTTCEEEEEEESCCSCBCSSCCSTTCCBSSCCEEESCCCTTCCCHHHHHHHHHHHHHHCTTSEEEEEEEETTTTEEEEEEEEECCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00434", "category": "Mutation", "instruction": "Describe the effect of mutation P54845-T96M on protein Neural retina-specific leucine zipper protein.", "input": "### Wild-type:\nMALPPSPLAMEYVNDFDLMKFEVKREPSEGRPGPPTASLGSTPYSSVPPSPTFSEPGMVGATEGTRPGLEELYWLATLQQQLGAGEALGLSPEEATELLQGQGPVPVDGPHGYYPGSPEETGAQHVQLAERFSDAALVSMSVRELNRQLRGCGRDEALRLKQRRRTLKNRGYAQACRSKRLQQRRGLEAERARLAAQLDALRAEVARLARERDLYKARCDRLTSSGPGSGDPSHLFL\n\n### Mutant:\nMALPPSPLAMEYVNDFDLMKFEVKREPSEGRPGPPTASLGSTPYSSVPPSPTFSEPGMVGATEGTRPGLEELYWLATLQQQLGAGEALGLSPEEAMELLQGQGPVPVDGPHGYYPGSPEETGAQHVQLAERFSDAALVSMSVRELNRQLRGCGRDEALRLKQRRRTLKNRGYAQACRSKRLQQRRGLEAERARLAAQLDALRAEVARLARERDLYKARCDRLTSSGPGSGDPSHLFL\n\n### Answer:", "output": "Decreased transactivational activity. The novel sequence variation (c. 287T>C) in the NRL gene results in the p. protein change. The functional test showed downregulation of the RHO promoter by p. protein, dissimilar to other missense NRL mutations that cause adRP. Affected RP patients carry the nucleotide change, while two other family members also carry the c. 287T>C variation but remain asymptomatic. The pathogenic mechanisms associated with adRP NRL mutations appear to be caused by a loss of function. The enhancement and retention strategy is suggested as a therapeutic approach for NRL-linked adRP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00435", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMFSQIGAFLDSALLLLVAFFAVYRLVLVLCRWQRRQLDIPIHI\n\n### Sequence 2:\nMVRYRCCRSQSRSRCRRRRRRCRRRRRRCCQRRRVRKCCRRTYTLRCRRY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00436", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKFIDEATIIVAAGDGGNGCISFRREKYIPFGPAEGGDGGNGGNVWLQADENLNTLIDYHFQHNFHAENGKHGQGKNFTGKCGKDLTIKVPIGTRVVDQNTNEILGDLIVHQQYLLVAKGGLRGLGNNHFKSSANCTPRKKTNGTKGEIRRLQLELILLADVGLLGLPNVGKSTLIRAVSAAKPKVANYPFTTLVPNLGVVQVHKKQSFIIADIPGLIKGAADGAGLGIRFLKHLERCRILLHLIDLAPADQSSPVENASIIINELKRYSEKLATKPSWLVFNKLDLIDKREALNIAQTISDALNQKHNYYLISAMNHQGIKTLCRDIMLFINKNK\n\n### Sequence 2:\nMKFVDEASILVVAGDGGNGCVSFRREKYIPKGGPDGGDGGDGGDVWLEADENLNTLIDYRFEKSFRAERGQNGASRDCTGKRGKDVTIKVPVGTRVIDQGTGETMGDMTKHGQRLMVAKGGWHGLGNTRFKSSVNRTPRQKTMGTPGDKRDLMLELMLLADVGMLGMPNAGKSTFIRAVSAAKPKVADYPFTTLVPSLGVVRMDNEKSFVVADIPGLIEGAAEGAGLGIRFLKHLERCRVLLHLIDIDPIDGSDPVENARIIIGELEKYSQDLAAKPRWLVFNKIDLLDQAEAEEKAKAIAQALGWEDKYYMISAASQTGVKDLCWDVMTFIIEN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00437", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMENLSPLLIEYQGLLKLIRNIKAMALNGLWDDVVEQEIVYIQSIERISQINVPANIPSTVQLQFRQLLQDILDTESQVKELLQNRMQELAVLIQQSQNQKSINNTYAEFSDDILPGKPQP\n\n### Sequence 2:\nLLKDYQQLLSLSQKILHLAVNGQWDTLVEQEIVYVQSVEGLVNTPIPDEIDSVMRLHLRQILQEVMDNEAKVKQLLQKRMDELSSLMGQSLKQKSINTTYNEFAGQRMLPGDALP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00438", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "C1QA, MRC1, C1QB, CD68", "output": "The expression pattern is consistent with macrophage in a context associated with inflammatory bowel disease, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["C1QA", "MRC1", "C1QB", "CD68"]}}
{"id": "eval_00439", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKGKKYYSDYHVWIEPIHSRIVKLGLSSQMREHLGNILHIDLPSLGAFIKEGEKLCILESSKSAIEVLSPVSGEVLEVNTALEDDILPVNNATESEGWFVVLQLTEDFRSESFSLEP\n\n### Sequence 2:\nMLAAALRRCTAAAAARGLLHPASAPSPAAAVCSIRCYSHGSHETDEEFDARWVTYFNKPDIDAWELRKGMNTLVGYDLVPEPKIIDAALRACRRLNDFASAVRILEVVKDKAGPHKEIYPYVIQELRPTLNELGISTPEELGLDKV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00440", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRS19_ORITB\n\n### Answer:", "output": "30S ribosomal protein S19.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00441", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "KIT, CD34, PROM1", "output": "This cell is most consistent with hematopoietic stem cell, a hematopoietic stem/progenitor cell. Key markers such as CD34, KIT, PROM1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hematopoietic stem cell", "tissue": "lung", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["KIT", "CD34", "PROM1"]}}
{"id": "eval_00442", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQVQLQVSGPGVVRPSETLSLTCEVSSGSTSRDFFYWSWVRQTPGKGLEWIGGMYSNSEETNHNPSLKSRVIISKDTSKNEFSLRLTSVTAADTAVYFCSSRAKIYYSASYSGGRIDVWGPGLLVTVS\n\n### Answer:", "output": "<SEQ_3Di>DDAQDKDDDQEDEAQAWDKIKTQHADDPDPPDKKKKWKWWADPVDGIGTAWTGMPVDADIGGDCVLPPQWDWGDDNVVRITMITRGRDHQVPQTWMKMWIWAFDDDDDPDGNGTDIDIDGTRGHGYD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00443", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nMS4A7:high LST1:medium CDKN1C:low FCGR3A:low\n</cell>", "output": "This cell is most consistent with CD16+ monocyte, a non-classical CD16+ monocyte. Key markers such as FCGR3A, MS4A7, CDKN1C support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "ovary", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MS4A7", "LST1", "CDKN1C", "FCGR3A"]}}
{"id": "eval_00444", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRPOC_MYCVP\n\n### Answer:", "output": "DNA-directed RNA polymerase subunit beta', RNAP subunit beta', RNA polymerase subunit beta', Transcriptase subunit beta'.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00445", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL22_LISMH\n\n### Answer:", "output": "The globular domain of the protein is located near the polypeptide exit tunnel on the outside of the subunit, while an extended beta-hairpin is found that lines the wall of the exit tunnel in the center of the 70S ribosome.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00446", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nEIF3B_DROSI\n\n### Answer:", "output": "Belongs to the eIF-3 subunit B family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00447", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKALHFGAGNIGRGFIGKLLADAGIQLTFADVNQVVLDALNARHSYQVHVVGETEQVDTVSGVNAVSSIGDDVVDLIAQVDLVTTAVGPVVLERIAPAIAKGLVKRKEQGNESPLNIIACENMVRGTTQLKGHVMNALPEDAKAWVEEHVGFVDSAVDRIVPPSASATNDPLEVTVETFSEWIVDKTQFKGALPNIPSMELTDNLMAFVERKLFTLNTGHAITAYLGKLAGHQTIRDAILDEKIRAVVKGAMEESGAVLIKRYGFDADKHAAYIQKILGRFENPYLKDDVERVGRQPLRKLSAGDRLIKPLLGTLEYSLPHKNLIQGIAGAMHFRSEDDPQAQELAALIADKGPQAALAQISGLDANSEVVSEAVTAYKAMQ\n\n### Sequence 2:\nMEKFFIVGSHGLKGEIKVSGAKNSALPIIAATLLTREPCVLENIPKLEDVNIMLSILKRLGSKVEFGQLLTIQNNRINELIIPEELARKIRASNLFLGPLVARFQEGVVPLPGGCNIGNRPMDLHLKGLRLMGAEVEEKSGFIRARAKKLKGAEIHLDFPSVGATENLMMAAALAQGTTIIRNAAREPEIVDLQNFLNLMGAKVKGAGTDIIKITGVNQLKGVTHKIIPDRIEAGTHMVMAAATQSDIIISGVIPEHLEAVMAKLKEAGALITSGGDWVRVTGKSIIKPVDIKTMPYPGFPTDMQPQILALLTLAQGTSIISEGVFDNRFKHVEELRRMGADIRLESRIAVIKGVPKLTGASVIAHDLRAAAALVIAGLAAEGMTVLEGIKNLDRGYENLDLKYQLIGAQIKRVNGEEKY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00448", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nTKIAMANFKSAMPIFKSHAYLKELEKTLKPQHFDRVFVFPDFFGLLPNSFLHFTLGVQNAYPRDCGAFTGEITSKHLEELKIHTLLIGHSERRTLLKESPSFLKEKFDFFKSKNFKIVYCIGEELTTREKGFKAVKEFLSEQLENIDLNYPNLVVAYEPIWAIGTKKSASLEDIYLTHGFLKQILNQKTPLLYGGSVNTQNAKEILGIDSVDGLLIGSASWELENFKTIISFL\n\n### Answer:", "output": "<SEQ_2D>CCEEEEECTBCSCHHHHHHHHHHHHHHSCGGGTTTEEEECCTTTCCCSCCSSSEECBSCCBSSSSBSCTTCCBHHHHHHTTCCEEEECCHHHHHTSCCCHHHHHHHHHHHHHTTCEEEEEECCCHHHHHHCHHHHHHHHHHHHTTSCTTCTTEEEEECCGGGTTCCCCCCHHHHHHHHHHHTTTSCTTSCEEEESSCCTTTHHHHHTSTTCCEEEECGGGGSHHHHHHHHTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00449", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKTIAVNSFKGGTAKTSTTLHLGAALAQYHKARVLLIDFDAQANLTAGLGLDPDCYDSLAVVLQGEKQISEVIRSIDSSGLDLIPADTWLERVEVSGSLAADRYSHERLKTILSTIEHQYDYVIIDTPPSLCWLTESALIAAQHALICATPEFYSVKGLERLATFIQGISSRHPLNILGVTLSFWNYRGKNNAAFTELIQKTFPGKLLNTRIRRDITISEAAIHGKPVFSTAPSARASEDYLKLTEELLFLLRDI\n\n### Sequence 2:\nMKTIAVNSFKGGTAKTSTTLHLGAALAQYHKARVLLIDFDAQANLTAGLGLDPDCYDSLAVVLQGEKQISEVIRSIDSSGLDLIPADTWLERVEVSGSLAADRYSHERLKTILSTIEHQYDYVIIDTPPSLCWLTESALIAAQHALICATPEFYSVKGLERLATFIQGISSRHPLNILGVTLSFWNYRGKNNAAFTELIQKTFPGKLLNTRIRRDITISEAAIHGKPVFSTAPSARASEDYLKLTEELLFLLRDI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00450", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCD20_MOUSE\n\n### Answer:", "output": "Belongs to the MS4A family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00451", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL18_SALTY\n\n### Answer:", "output": "50S ribosomal protein L18.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00452", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL14E_DESA1\n\n### Answer:", "output": "Belongs to the eukaryotic ribosomal protein eL14 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00453", "category": "Mutation", "instruction": "Describe the effect of mutation Q02169-K53A on protein dTTP/UTP pyrophosphatase.", "input": "### Wild-type:\nMTKPLILASQSPRRKELLDLLQLPYSIIVSEVEEKLNRNFSPEENVQWLAKQKAKAVADLHPHAIVIGADTMVCLDGECLGKPQDQEEAASMLRRLSGRSHSVITAVSIQAENHSETFYDKTEVAFWSLSEEEIWTYIETKEPMDKAGAYGIQGRGALFVKKIDGDYYSVMGLPISKTMRALRHFDIRA\n\n### Mutant:\nMTKPLILASQSPRRKELLDLLQLPYSIIVSEVEEKLNRNFSPEENVQWLAKQAAKAVADLHPHAIVIGADTMVCLDGECLGKPQDQEEAASMLRRLSGRSHSVITAVSIQAENHSETFYDKTEVAFWSLSEEEIWTYIETKEPMDKAGAYGIQGRGALFVKKIDGDYYSVMGLPISKTMRALRHFDIRA\n\n### Answer:", "output": "Loss of activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00454", "category": "Mutation", "instruction": "Describe the effect of mutation Q55721-F252C on protein Folate-biopterin transporter.", "input": "### Wild-type:\nMLVAMSMTPIAILFSTPLKRFLREKVLLGNAPSWELLAILSIYFVQGVLGLSRLAVSFFLKDELGLSPAAMGALIGLGAAPWILKPVLGLMSDTVPLFGYRRRSYLWLSGLMGSAGWLLFAAWVSSGTQAGLVLLFTSLSVAIGDVIVDSLVVERAQRESLAQVGSLQSLTWGAAAVGGIITAYASGALLEWFSTRTVFAITAIFPLLTVGAAFLISEVSTAEEEEKPQPKAQIKLVWQAVRQKTILLPTLFIFFWQATPSAESAFFYFTTNELGFEPKFLGRVRLVTSVAGLIGVGLYQ\n\n### Mutant:\nMLVAMSMTPIAILFSTPLKRFLREKVLLGNAPSWELLAILSIYFVQGVLGLSRLAVSFFLKDELGLSPAAMGALIGLGAAPWILKPVLGLMSDTVPLFGYRRRSYLWLSGLMGSAGWLLFAAWVSSGTQAGLVLLFTSLSVAIGDVIVDSLVVERAQRESLAQVGSLQSLTWGAAAVGGIITAYASGALLEWFSTRTVFAITAIFPLLTVGAAFLISEVSTAEEEEKPQPKAQIKLVWQAVRQKTILLPTLCIFFWQATPSAESAFFYFTTNELGFEPKFLGRVRLVTSVAGLIGVGLYQ\n\n### Answer:", "output": "Affects 5-formyltetrahydrofolate transporter activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00455", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nBBRD_ORYSJ\n\n### Answer:", "output": "Transcriptional regulator that specifically binds to GA-rich elements (GAGA-repeats) present in regulatory sequences of genes involved in developmental processes.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00456", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMEFMRVERVLLKDYTTLGVGGPAELWTVETREELKRATEAPYRVLGNGSNLLVLDEGVPERVIRLAGEFQTYDLKGWVGAGTLLPLLVQEAARAGLSGLEGLLGIPAQVGGAVKMNAGTRFGEMADALEAVEVFHDGAFHVYCPEELGFGYRKSHLPPGGIVTRVRLKLKERPKEEILRRMAEVDRARKGQPKRKSAGCAFKNPPGQSAGRLIDERGLKGLRVGDAMISLEHGNFIVNLGQARAKDVLELVRRVQEELPLELEWEVWP\n\n### Answer:", "output": "<SEQ_2D>CCCCEEEEEEGGGGSSSCCCCEEEEEEECSHHHHHHHTTSCEEECSSSSSEEECTTCCSSEEEEECGGGGCBCTTSEEETTSBHHHHHHHHHHTTEESCGGGTTCCCBHHHHHHTTCEETTEEGGGGEEEEEEEETTEEEEECGGGSCCBTTBCCCSTTCEEEEEEECCEECCHHHHHHHHHHHHHHTTTSCSSCSSSCCBCCCTTCCHHHHHHHTTCTTCEETTEEECSSCTTCEEECSSCCHHHHHHHHHHHHHHSCCCBSSEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00457", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPAP3_SPOEX\n\n### Answer:", "output": "Paralytic peptide 3, Paralytic peptide III, PP III.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00458", "category": "Mutation", "instruction": "Describe the effect of mutation P02533-Y366A on protein Keratin, type I cytoskeletal 14.", "input": "### Wild-type:\nMTTCSRQFTSSSSMKGSCGIGGGIGGGSSRISSVLAGGSCRAPSTYGGGLSVSSSRFSSGGACGLGGGYGGGFSSSSSSFGSGFGGGYGGGLGAGLGGGFGGGFAGGDGLLVGSEKVTMQNLNDRLASYLDKVRALEEANADLEVKIRDWYQRQRPAEIKDYSPYFKTIEDLRNKILTATVDNANVLLQIDNARLAADDFRTKYETELNLRMSVEADINGLRRVLDELTLARADLEMQIESLKEELAYLKKNHEEEMNALRGQVGGDVNVEMDAAPGVDLSRILNEMRDQYEKMAEKNRK\n\n### Mutant:\nMTTCSRQFTSSSSMKGSCGIGGGIGGGSSRISSVLAGGSCRAPSTYGGGLSVSSSRFSSGGACGLGGGYGGGFSSSSSSFGSGFGGGYGGGLGAGLGGGFGGGFAGGDGLLVGSEKVTMQNLNDRLASYLDKVRALEEANADLEVKIRDWYQRQRPAEIKDYSPYFKTIEDLRNKILTATVDNANVLLQIDNARLAADDFRTKYETELNLRMSVEADINGLRRVLDELTLARADLEMQIESLKEELAYLKKNHEEEMNALRGQVGGDVNVEMDAAPGVDLSRILNEMRDQYEKMAEKNRK\n\n### Answer:", "output": "No effect on interaction with KRT5 or keratin intermediate filament networks. The mutation in keratin 5 affects the specific contacts between interacting 2B domains, which play a crucial role during 10-nm intermediate filament assembly.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00459", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "KDR, CLDN5, PECAM1, VWF, CDH5", "output": "This cell is most consistent with endothelial cell, a vascular endothelial cell. Key markers such as PECAM1, CDH5, CLDN5 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "endothelial cell", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["KDR", "CLDN5", "PECAM1", "VWF", "CDH5"]}}
{"id": "eval_00460", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGLLSTNFDMIQALPLNVKQRVCALKNLQMKTIQIESDFYKRVHELEIEFEGKFKSTFDQRKAIVAGEVEPTKEQIDTPILEGLEGDQLAELYKAAEADPSAKGIKDFWLTALRTHDLVAEAIEEHDVPILSYLTDVTTAASKDPAGFKIEFHFATNPYFKNQVLTKTYLLGFDPDAEAPLQFDGPHVIRAVGDTIEWEDGKNVTKKATKTVKADSFFNFFEPPDEQAEEFLELDYEMGQAIRDTIIPRAVLFYTGELQ\n\n### Answer:", "output": "<SEQ_3Di>DPPDPCPVVLVPDDPVVNVVVVVVVVVLLVVLQVVLVVQVVVLVVVQVCLVVVVVVQVVLLCQQQLNDDADPVRVPDQLDPPDDPVRSVCVVVVDDDDPPDRHNFLLQVQLQCLDPLSVVQADPLCSVQSSQFSHKDKHFDDPLQWIKIKGQGDDDLFFPDRIWMKTWHKDSAADPVCSSPDDGIGTDFIAADDGRGDPPRDQQWCDVVTHGDRHNNCLRVADPDPCPVVNVSSVVSVCCVHPPCNVPVVCSSVSVDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00461", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CCC(C)SSC(C)CC", "output": "No standard functional groups from the preset library were detected in this molecule.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "CCC(C)SSC(C)CC", "compound_name": null, "pubchem_cid": null, "MW": 178.37, "LogP": 3.96, "TPSA": 0.0, "QED": 0.58, "functional_groups": [], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": []}}
{"id": "eval_00462", "category": "Mutation", "instruction": "Describe the effect of mutation P0ACC7-C296A on protein Bifunctional protein GlmU [Includes: UDP-N-acetylglucosamine pyrophosphorylase.", "input": "### Wild-type:\nMLNNAMSVVILAAGKGTRMYSDLPKVLHTLAGKAMVQHVIDAANELGAAHVHLVYGHGGDLLKQALKDDNLNWVLQAEQLGTGHAMQQAAPFFADDEDILMLYGDVPLISVETLQRLRDAKPQGGIGLLTVKLDDPTGYGRITRENGKVTGIVEHKDATDEQRQIQEINTGILIANGADMKRWLAKLTNNNAQGEYYITDIIALAYQEGREIVAVHPQRLSEVEGVNNRLQLSRLERVYQSEQAEKLLLAGVMLRDPARFDLRGTLTHGRDVEIDTNVIIEGNVTLGHRVKIGTGCVIKN\n\n### Mutant:\nMLNNAMSVVILAAGKGTRMYSDLPKVLHTLAGKAMVQHVIDAANELGAAHVHLVYGHGGDLLKQALKDDNLNWVLQAEQLGTGHAMQQAAPFFADDEDILMLYGDVPLISVETLQRLRDAKPQGGIGLLTVKLDDPTGYGRITRENGKVTGIVEHKDATDEQRQIQEINTGILIANGADMKRWLAKLTNNNAQGEYYITDIIALAYQEGREIVAVHPQRLSEVEGVNNRLQLSRLERVYQSEQAEKLLLAGVMLRDPARFDLRGTLTHGRDVEIDTNVIIEGNVTLGHRVKIGTGAVIKN\n\n### Answer:", "output": "No effect. No effect on enzyme activity", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00463", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "CX3CR1, TMEM119, P2RY12", "output": "The expression pattern is consistent with microglia in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "brain", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CX3CR1", "TMEM119", "P2RY12"]}}
{"id": "eval_00464", "category": "Mutation", "instruction": "Describe the effect of mutation Q01827-M481A on protein Synaptic vesicular amine transporter.", "input": "### Wild-type:\nMALSDLVLLRWLRDSRHSRKLILFIVFLALLLDNMLLTVVVPIIPSYLYSIKHEKNSTEIQTTRPELVVSTSESIFSYYNNSTVLITGNATGTLPGGQSHKATSTQHTVANTTVPSDCPSEDRDLLNENVQVGLLFASKATVQLLTNPFIGLLTNRIGYPIPMFAGFCIMFISTVMFAFSSSYAFLLIARSLQGIGSSCSSVAGMGMLASVYTDDEERGKPMGIALGGLAMGVLVGPPFGSVLYEFVGKTAPFLVLAALVLLDGAIQLFVLQPSRVQPESQKGTPLTTLLKDPYILIAAG\n\n### Mutant:\nMALSDLVLLRWLRDSRHSRKLILFIVFLALLLDNMLLTVVVPIIPSYLYSIKHEKNSTEIQTTRPELVVSTSESIFSYYNNSTVLITGNATGTLPGGQSHKATSTQHTVANTTVPSDCPSEDRDLLNENVQVGLLFASKATVQLLTNPFIGLLTNRIGYPIPMFAGFCIMFISTVMFAFSSSYAFLLIARSLQGIGSSCSSVAGMGMLASVYTDDEERGKPMGIALGGLAMGVLVGPPFGSVLYEFVGKTAPFLVLAALVLLDGAIQLFVLQPSRVQPESQKGTPLTTLLKDPYILIAAG\n\n### Answer:", "output": "Does not alter sorting to large dense granules. The mutation in VMAT2 affects the trafficking of the protein to large dense-core vesicles (LDCVs) within the biosynthetic pathway.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00465", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMASSESRYSPTSLEPKWQQRWQEMGLDQTPEPSAERAENFYALSMFPYPSGNLHMGHVRNYVITDVIARLMRLKGKRVLHPMGWDAFGLPAENAAIERGVDPADWTDRNIAQMREQLQRLGLSIDWSREVATCHSDYYRWTQWLFLQFLNADLAYRKEATVNWDPIDQTVLANEQVDADGRSWRSGALAEKRKLRQWFLKITAVADELLDDLEQLKGWPERVRTMQANWIGRSSGATLRFAIEADGKQAIEVFTTRPDTVFGVSYVVLAPEHDLVDQLTSADQRQAVEAFRQSLQSISEQDRVADDRPKRGVATGGTVQHPFTGQAVPVWIADYVLPDYGTGAVMGVPAHDSRDFAFAKQYDLPITTVVVEPGQAPDSGEPSEAFTGLGELVGSADFDGLQGEEAKTAIIQAAEQRGVGAAKITFRLRDWLISRQRYWGCPIPVIHCDDCGVVPVPESDLPVELPRDVDLSGSGGSPLERATEWKQVRCPKCGKPATRETDTMDTFMCSSWYYLRYTDANNDSAAFTNASIDAWMPVDQYVGGVEHAILHLLYSRFFTKVLQQRNLVSCSEPFQKLLTQGMVQGVTYRNPKTRKYIAPSAVSDPNQPTDPDDGEALEVFFEKMSKSKYNGVDPGAVVDRYGADTARMFILFKAPPEKDLEWDDADVEGQFRFLQRIWRLCDGAKASGLQLQSPLPLPAELTPAETDLRRAVHTAIEAVSDDLQGDELQFNTAVSELMKLSNAMAGQLEAVSTAVAQEAVRSLLLLLAPFAPHLADELWEQLQGSGSIHQQRWPEVDASALVRDTITVVLQVKGKVRGNLEVPAAISKDELEQVALASDVAQKWLEGNAPKRVIVVPGKLVNLVP\n\n### Sequence 2:\nRYDPIELETRWQKEWLRQGLDRTPVAETNQ-KRFYALSMFPYPSGKLHMGHVRNYVITDVIARVQRMRGDAVLHPMGWDAFGLPAENAAIARNVDPGDWTDQNIAQMRAQLDRLGLSIDWDRQQATCHQDYYRWTQWLFLELFAGGLAYQKEATVNWDPVDKTVLANEQVDGEGRSWRSGALVEQRQLKQWFLRITDYADALIDDLDELTGWPERVRTMQANWIGRSHGAEIKFRVAGVANSIITVFTTRPDTLHGASYVVLAPEHPLVESLTSPEQRLAVTAFCDLISQLSVKDRTAEDQPKRGVPIGAQVINPVNGESLPVWIADYVLADYGSGAVMGVPAHDERDFIFARSHELPIRIVV----QLPDTDEHHNDGQAWTGAGVLVNSGAFDGLSTEEAKVAITTHGASEGWAQSKVQYRLRDWLISRQRYWGCPIPIIHCASCGIVPVPQEDLPVTLPRDIDLSGKGGSPIAQEQAWVEVKCPICGEKAHRETDTMDTFMCSSWYYLRFADPLNSQRPFDKDIVDEWLPVDQYVGGIEHAILHLLYARFFTKALHDRNLIGFKEPFNRLLTQGMVQGLTYRNAKNGSYISPELVSDDSDPRDPESGDRLEVLFEKMSKSKYNGVDPAVVIDRYGADTARMFILFKAPPEKDLEWDDADVEGQFRFLQRLIRLIDSFAWPKTDGENASISSANLIIDSADLSEEEINMRRATHKAIEAITEDLSGD-IQLNTAISELMKLSNSLSGKLDKVRNEVAAEALSVLVRLMAPFAPHLAEEFWLKLHGHGSIHQQSWPVIDPSALVLETIELVIQVKGKVRGKIQVPANADKKTLEELALNSDIAVKWLEGQSPRRIIIVPGKLVNLVP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00466", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCRCB_SALEP\n\n### Answer:", "output": "Important for reducing fluoride concentration in the cell, thus reducing its toxicity.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00467", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "IRF7, GZMB, CLEC4C, TCF4, LILRA4", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as GZMB, CLEC4C, LILRA4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "skin", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IRF7", "GZMB", "CLEC4C", "TCF4", "LILRA4"]}}
{"id": "eval_00468", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nAROE_CHLTE\n\n### Answer:", "output": "Belongs to the shikimate dehydrogenase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00469", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "SLC32A1, GAD2, GAD1", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as GAD2, SLC32A1, GAD1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "blood", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SLC32A1", "GAD2", "GAD1"]}}
{"id": "eval_00470", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTTLLNPYFGEFGGMYVPQILMPALNQLEEAFVSAQKDPEFQAQFADLLKNYAGRPTALTKCQNITAGTRTTLYLKREDLLHGGAHKTNQVLGQALLAKRMGKSEIIAETGAGQHGVASALASALLGLKCRIYMGAKDVERQSPNVFRMRLMGAEVIPVHSGSATLKDACNEALRDWSGSYETAHYMLGTAAGPHPYPTIVREFQRMIGEETKAQILDKEGRLPDAVIACVGGGSNAIGMFADFINDTSVGLIGVEPGGHGIETGEHGAPLKHGRVGIYFGMKAPMMQTADGQIEESYSISAGLDFPSVGPQHAYLNSIGRADYVSITDDEALEAFKTLCRHEGIIPALESSHALAHALKMMREQPEKEQLLVVNLSGRGDKDIFTVHDILKARGEI\n\n### Sequence 2:\nMFKVGMINLGDKQSTIVVAMSGGVDSSAVAAMLHEQGHNVIGITLQLYDHGMAVGKKNACCAGQDIYDAKMVANKLGIPHYVLDYENKFKESVIDNFVDSYLQGETPLPCVQCNKSVKFRDLIKTARELGADKLATGHYVRKINGDNGAELHTGLDPAKDQSYFLFTTTKEQLEYLRFPLGGLTKDETRKLASKFGLEVADKPDSQDICFIPDGNYKSVINKIRPNSSESGKIIHVNGFELGEHSGIINYTIGQRRGLGIAYNEPLYVVKIDPKDNIVYVGPESALNVQEFIIRDVNWLADEIKDNEKLEVAVKIRSTRPPRLAEISKLGDDKMKVKFLCEEKAVAPGQACVIYAGERVLGGGWITREIR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00471", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIQPQSYLNVADNSGARKLMCIRVLGSSNRKYAHIGDMVIAVVKETVPNMPLKKSEIVRAVIVRTRKGLKRDNGMVLRFDDNAAVVINQEGNPRGTRVFGPVARELRDLNFTKIVSLAPEVL\n\n### Sequence 2:\nMAQLSKDDILEAVANMSVMDVVDLVKAMEEKFGVSAQAAIAVAGPVAGGEAAAAEEKTEFNVKMVSFGDNKIGVIKAIRTITGLGLKEAKDLVESVPSVVKESVSKEEAEKIKKELEEAGAKVELE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00472", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C1C2C=CC(n3[nH]c(=O)n(Cc4ccccc4)c3=O)CC2C(=O)N1c1cccc(Cl)c1", "output": "This molecule is medium-sized and polar, with moderately lipophilic character. It contains aromatic ring, amide, halide. Its descriptor profile: MW=450.88, LogP=2.35, TPSA=97.17, HBD=1, HBA=4, QED=0.49.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "O=C1C2C=CC(n3[nH]c(=O)n(Cc4ccccc4)c3=O)CC2C(=O)N1c1cccc(Cl)c1", "compound_name": null, "pubchem_cid": null, "MW": 450.88, "LogP": 2.35, "TPSA": 97.17, "QED": 0.49, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00473", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLPPAYTNSFTRGVYYPDKVFRSSVLHSTQDLFLPFFSNVTWFHVIFDNPVLPFNDGVYFASIEKSNIIRGWIFGTTLDSKTQSLLIVNNATNVVIKVCEFQFCNDPFLEFRVYSSANNCTFEYVSQPFLMQGNFKNLREFVFKNIDGYFKIYSKHTPIIVREPEDLPQGFSALEPLVDLPIGINITRFQTLLALHRSGAAAYYVGYLQPRTFLLKYNENGTITDAVDCALDPLSETKCTLKSFTVEKGIYQTSNFRVQPTESIVRFPNITNLCPFDEVFNATRFASVYAWNRKRISNCVADYSVLYNLAPFFTFKCYGVSPTKLNDLCFTNVYADSFVIRGDEVRQIAPGQTGNIADYNYKLPDDFTGCVIAWNSNKLDSKVSGNYNYLYRLFRKSNLKPFERDISTEIYQAGNKPCNGVAGFNCYFPLRSYSFRPTYGVGHQPYRVVVLSFELLHAPATVCGPKKSTNLVKNKCVNFNFNGLKGTGVLTESNKKFLPFQQFGRDIADTTDAVRDPQTLEILDITPCSFGGVSVITPGTNTSNQVAVLYQGVNCTEVNVFQTRAGCLIGAEYVNNSYECDIPIGAGICASYQTSQSIIAYTMSLGAENSVAYSNNSIAIPTNFTISVTTEILPVSMTKTSVDCTMYICGDSTECSNLLLQYGSFCTQLKRALTGIAVEQDKNTQEVFAQVKQIYKTPPIKYFGGFNFSQILPDPSKPSKRSPIEDLLFNKVTRDLICAQKFKGLTVLPPLLTDEMIAQYTSALLAGTITSGWTFGAGPALQIPFPMQMAYRFNGIGVTQNVLYENQKLIANQFNSAIGKIQDSLSSTPSALGKLQDVVNHNAQALNTLVKQLSSKFGAISSVLNDIFSRLDPPEAEVQIDRLITGRLQSLQTYVTQQLIRAAEIRASANLAATKMSECVLGQSKRVDFCGKGYHLMSFPQSAPHGVVFLHVTYVPAQEKNFTTAPAICHDGKAHFPREGVFVSNGTHWFVTQRNFYEPQIITTDNTFVSGNCDVVIGIVNNTVYDPLQPELDS\n\n### Answer:", "output": "<SEQ_3Di>DQFDKDWAADEWAADPDPDFDAFFKDKDWDWTADGGDIWRKDQDVDDWDKAWADFKKKKKKWAADAWFFKKKWADPLAQVFWIWIQGDDDQWTWTWTGSFGQDDACADVRPGGDDGDDTPDTPRDRHHQDPDLWIKMWIWMWGDDPQFTWIWIDIDIDNDPPDDDDDDAYDYTDTDGGHGGDGDTTIMTTHTYCPPCNHIIMIMDHIDGWMKIFGAHRVRDTHDIATCPDDPLSQVCVVDVDPKDAFFKDFRWKDFDAADDAAEAEADAPAEWPPCCQPVNPDHAAQLTKDKDKDFQHEYACVVVVPDPDFPQWPQPADDSPPAQAAFFAIKIKIKYKDWQQCPVLQHADDDDFSCWAGAHAHNPFHWTKTKDFPLVPQADPPFRQPYKDFGDDDGTDDHPDRDPDWDWDQPDPDDDDPDDDDNTHGGIDTQGAHPPDDRGNHMMMMMIMRGPRDPFLTTYHHDDDRGDDDAFDFHWYCQNNDTDTAGWHQDDDDDDPGDAWDADDPRHTQWGAHNHPGGIIGGDHTDMATKMWRDNMRPPDSFIWIFGFPEQVPPGVWDQDPGGITGREEEDAPADAADYDDDDQKHWDFDPVIGIYIYGHDPDDDDDPDDDQWKDWAFAFWFKAWDKDWFFQAAWLKAFDLCCQLPPPDPVVVVVVCVVPPLRVVLRVLSVVNLVVSLVLLCLQLVQFDDQDAADPDDCQLQADCQQQAFDPPDPLRGGVLQVLQAVLQDVDPQVVLVVNGTDTDDHPQDSSNSRRSLNSLVRQCSAQNQVRNVDNRDGDGNLVNLQVLCVPQFDHSVPCVVCVVVVNVSNNVNVVVVSVVPVPDSCPCVVVSVSVVVVVVLSVVLVVQQQPQQPAPGSDLVCLPVPDDPVVSVVRSVSSSVSSSVSSVVSSVVVVVVSVVSVVSSVLSSVLCCDASNHQHPDAPSQPHARWRHWRWHDHDSIIMITTTHIGGDRIDMWTFFQFEADPQWTWGFQQFTWTDDPPDIATATNHHDDGDAAEPVGTDTDHGPDPGPRYYYDDHHYPCVVVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00474", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nCD14:low VCAN:low S100A8:low FCN1:high S100A9:medium LYZ:low\n</cell>", "output": "This cell is most consistent with CD14+ monocyte, a classical CD14+ monocyte. Key markers such as CD14, S100A9, VCAN support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "CD14+ monocyte", "tissue": "colon", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD14", "VCAN", "S100A8", "FCN1", "S100A9", "LYZ"]}}
{"id": "eval_00475", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nYRHRHLIQDLSGLLPHSRKLHTMDELNFTGNCLKGSRPVLNFGVPPNARKSKPFIDHVMSFVRTYEISHSTKNKEEYEDGEEDISLVEIGPRF\n\n### Answer:", "output": "<SEQ_3Di>DCVVQVQVVLCQQEVPDDGQWWDPDDPADADWDRAEAEADVNGHDCDDPPHDDHGQWYAYVHFIWGWDDDDDDPVCRPPDDHDTDIGTHTHDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00476", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSYSGERDNFAPHMALVPMVIEQTSRGERSFDIYSRLLKERVIFLTGQVEDHMANLIVAQMLFLEAENPEKDIYLYINSPGGVITAGMSIYDTMQFIKPDVSTICMGQAASMGAFLLTAGAKGKRFCLPNSRVMIHQPLGGYQGQATDIEIHAREILKVKGRMNELMALHTGQSLEQIERDTERDRFLSAPEAVEYGLVDSILTHRN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCBCCCCCCCCCCCCBHHHHHHTTTEEEEESSBCHHHHHHHHHHHHHHHHHCSSSCEEEEEEECCBCHHHHHHHHHHHHHSSSCEEEEEEEEEETHHHHHHHTSCTTCEEECTTCEEEECCCEEEEEEEHHHHHHHHHHHHHHHHHHHHHHHHHHCCCHHHHHHHTSSCEEEEHHHHHHHTSSSEECCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00477", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEYSDVPKFIRDVSIKQHEWMRESIKLIASENITSLAVREACATDFMHRYAEGLPGKRLYQGCKYIDEVETLCIELSKELFKAEHANVQPTSGVVANLAVFFAETKPGDKLMALSVPDGGHISHWKVAAGIRGLKVINHPFDPEEMNIDADAMVKKILKPKLILFGGSLFPFPHPVADAYEAAQEVGAKIAYDGAHVLGLIAGKQFQDPLREGAEYLMGSTHKTFFGPQGGVILTTKENADKIDSHVFPGVVHHLHHKAGLAIALAEMLEFGEAYAKQVIKNAKALAQALYERGFNVLCEHKDFTESHQVIIDIESSPDIEFSASELAKMYEEANIILNKNLLPWDDSDNPSGIRLGTQECTRLGMKEKEMEEIAEFMKRIAIDKEKPEKVREDVKEFAKEYSTIHYSFDEGDGFKYLRFY\n\n### Answer:", "output": "<SEQ_3Di>DVVCVVVPVVVVLLVVLVLVLLFAEFAQLFFFAFPLLVVLLPDPLQVDFAAFPQGPDPDDPCVVVSVLQVLLFVLVCVQLVFQGKGFFFQALLSQLVLVCLLDAAAAFEEEEADQLLAFAPSQAPVPVNPGHHDYDYFDADQQAGEGDLVVRLVVPCAGAEYEEHGSRQFAAYQLLSCAVSAPVNPHAYEYECAQHVLLVSLPARDSNQVSHHQKYKYFCGHNFRFFGGMMIGGHNVCHVSSSVSPDVRDGGRRSRVSSSSLRSVLSRVLSNVLNVQQVQQLLLLLVLCCVVPWAWHPPPVSQGSDQKHKGFQPVIPDDPDFPVVVQVLVVLLSYHWDDGAHSPRPDDGHRMTITGCRNVPQLPDGSVLSVVVNVLSCCCRPPNHRSNVSSVVSSCRSNVRQWGHSRPDTDRSSDDDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00478", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRIMP_THESQ\n\n### Answer:", "output": "Ribosome maturation factor RimP.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00479", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMYQPDFPTVPFRLGLYPVVDSVEWIERLLEAGVRTIQLRIKDKRDEEVEADVIAAIALGRRYDARLFINDYWRLAMKHNAYGVHLGQEDLETTDLKAIQAAGLRLGVSTHDDMEIDIALAAKPSYIALGHVFPTQTKQMPSAPQGLAQLVRHIERLADYPTVAIGGISLEHAPAVLATGVGSIAVVSAITQAADWRDATAQLLAIAGVGDE\n\n### Sequence 2:\nMAKNRFNQHWLHDHINDPYVKMAQREGYRARAAYKLKEIDEQDKLIRPGQVIVDLGAAPGSWSQYARNKLAQGKQRDAQREGGIDGTIIALDLLPMEPIADVHFLQGDFREDDVLHQLEGVLEGRPVDLVISDMAPNLSGVASADAARIEHICDLALEFAQNHLKPEGALLVKCFHGSGYSQIVEKFKQQFKVVAPRKPKASRDKSSETFILGRQLKHPR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00480", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "IL7R, CD3E, CCR7, SELL, LTB, TCF7, CD3D", "output": "This cell is most consistent with naive CD4+ T cell, a T-cell subtype with naive/central-memory-like profile. Key markers such as CD3E, TCF7, CD3D support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "ovary", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IL7R", "CD3E", "CCR7", "SELL", "LTB", "TCF7", "CD3D"]}}
{"id": "eval_00481", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nBACR_HALS4\n\n### Answer:", "output": "Light-driven proton pump.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00482", "category": "Mutation", "instruction": "Describe the effect of mutation P9WGG1-A173Q on protein Phosphoheptose isomerase.", "input": "### Wild-type:\nMCTARTAEEIFVETIAVKTRILNDRVLLEAARAIGDRLIAGYRAGARVFMCGNGGSAADAQHFAAELTGHLIFDRPPLGAEALHANSSHLTAVANDYDYDTVFARALEGSARPGDTLFAISTSGNSMSVLRAAKTARELGVTVVAMTGESGGQLAEFADFLINVPSRDTGRIAESHIVFIHAISEHVEHALFAPRQ\n\n### Mutant:\nMCTARTAEEIFVETIAVKTRILNDRVLLEAARAIGDRLIAGYRAGARVFMCGNGGSAADAQHFAAELTGHLIFDRPPLGAEALHANSSHLTAVANDYDYDTVFARALEGSARPGDTLFAISTSGNSMSVLRAAKTARELGVTVVAMTGESGGQLAEFADFLINVPSRDTGRIQESHIVFIHAISEHVEHALFAPRQ\n\n### Answer:", "output": "24-fold increase in catalytic efficiency.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00483", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKAKEIRDLTTSEIEEQIKSSKEELFNLRFQLATGQLEETARIRTVRKTIARLKTVAREREIEQSKANQ\n\n### Sequence 2:\nMKAKEIRDLTTAEIEQKIKSLKEELFNLRFQLATGQLENTARIRQVRKDIARMKTIIRERELAANK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00484", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEGQVLVLDGRGHLLGRLAAIVAKQVLLGRKVVVVRCEGINISGNFYRNKLKYLAFLRKRMNTNPSRGPYHFRAPSRIFWRTVRGMLPHKTKRGQAALERLKVLDGIPPPYDKKKRMVVPAALKVVRLKPTRKFAYLGRLAHEVGWKYQAVTATLEEKRKEKAKMHYRKKKQILRLRKQAEKNVEKKICKFTEVLKTNGLLV\n\n### Answer:", "output": "<SEQ_3Di>DPAEAEAEQAQFALLQVLLVLLVCLLPHHLYEYEALQRHKHFAAPVVVLVVLVVVVPDADDVDRVVTDDSDQDSQVVSLVSSLVVFPCVDPSSVSSSVSYHYYHHADPVGLPDDHDYDLSGGCVNPDDPPTHMGGSNVSSVVVDDDCPPVSVVVNVVSVVVVVVVVVVVVVVVVVVVVVCVPCPVVCVVVQVVCVVVPHDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00485", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRR18_EUCGG\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bS18 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00486", "category": "Mutation", "instruction": "Describe the effect of mutation P04439-A5P on protein HLA class I histocompatibility antigen, A alpha chain.", "input": "### Wild-type:\nMAVMAPRTLLLLLSGALALTQTWAGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAASQRMEPRAPWIEQEGPEYWDQETRNVKAQSQTDRVDLGTLRGYYNQSEAGSHTIQIMYGCDVGSDGRFLRGYRQDAYDGKDYIALNEDLRSWTAADMAAQITKRKWEAAHEAEQLRAYLDGTCVEWLRRYLENGKETLQRTDPPKTHMTHHPISDHEATLRCWALGFYPAEITLTWQRDGEDQTQDTELVETRPAGDGTFQKWAAVVVPSGEEQRYTCHVQHEGLPKPLTLRWEL\n\n### Mutant:\nMAVMPPRTLLLLLSGALALTQTWAGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAASQRMEPRAPWIEQEGPEYWDQETRNVKAQSQTDRVDLGTLRGYYNQSEAGSHTIQIMYGCDVGSDGRFLRGYRQDAYDGKDYIALNEDLRSWTAADMAAQITKRKWEAAHEAEQLRAYLDGTCVEWLRRYLENGKETLQRTDPPKTHMTHHPISDHEATLRCWALGFYPAEITLTWQRDGEDQTQDTELVETRPAGDGTFQKWAAVVVPSGEEQRYTCHVQHEGLPKPLTLRWEL\n\n### Answer:", "output": "In allele A*80:01.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00487", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAAKIRRDDEVIVLTGKDKGKRGKVKNVLSSGKVIVEGINLVKKHQKPVPALNQPGGIVEKEAAIQVSNLALFNATTGKADRVGFRFEDGKKVRFFKSTSETIK\n\n### Sequence 2:\nMDGQFEQKKKQKDETYDIEHLIACFSPMIRKKLSNTSYQEREDLEQELKIKMFEKADMLLCQDVPGFWEFILYMVDENS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00488", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "COCCOC(=O)C1=C(C)NC(C)=C(C(=O)OC(C)C)C1c1cccc([N+](=O)[O-])c1", "output": "This molecule is medium-sized and polar, with moderately lipophilic character. It contains aromatic ring, carboxylic acid, ester, amine, ether. Its descriptor profile: MW=418.45, LogP=2.97, TPSA=117.0, HBD=1, HBA=8, QED=0.3.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "ClinTox", "canonical_smiles": "COCCOC(=O)C1=C(C)NC(C)=C(C(=O)OC(C)C)C1c1cccc([N+](=O)[O-])c1", "compound_name": null, "pubchem_cid": null, "MW": 418.45, "LogP": 2.97, "TPSA": 117.0, "QED": 0.3, "functional_groups": ["aromatic ring", "carboxylic acid", "ester", "amine", "ether", "nitro"], "moleculenet_labels": {"FDA_APPROVED": 1, "CT_TOX": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00489", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SM23-Y171A on protein Acyl-CoA-binding domain-containing protein 1.", "input": "### Wild-type:\nMADWYQLAQSIIFGLIFAYLLAKLISILLAFKDENLSLTRNHTTQSEYENLRKVETLTGISGETDSLIAEQGSLRGDEDESDDDDWEGVESTELDEAFSAATAFVAAAASDRLSQKVSNELQLQLYGLYKIATEGPCTAPQPSALKMTARAKWQAWQKLGAMPPEEAMEKYIDLVTQLYPAWVEGGSKRRNRSGEAAGPMGPVFSSLVYEEESDNELKIDAIHAFAREGEVENLLKCIENGIPVNARDSEGRTPLHWAIDRGHLNVAEALVDKNADVNAKDNEGQTSLHYAVVCEREALA\n\n### Mutant:\nMADWYQLAQSIIFGLIFAYLLAKLISILLAFKDENLSLTRNHTTQSEYENLRKVETLTGISGETDSLIAEQGSLRGDEDESDDDDWEGVESTELDEAFSAATAFVAAAASDRLSQKVSNELQLQLYGLYKIATEGPCTAPQPSALKMTARAKWQAWQKLGAMPPEEAMEKAIDLVTQLYPAWVEGGSKRRNRSGEAAGPMGPVFSSLVYEEESDNELKIDAIHAFAREGEVENLLKCIENGIPVNARDSEGRTPLHWAIDRGHLNVAEALVDKNADVNAKDNEGQTSLHYAVVCEREALA\n\n### Answer:", "output": "Normal arachidonyl-CoA-binding activity. Replacement in ACBP1 did not affect arachidonyl-CoA binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00490", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAKKNKNVLVRLVSTAGTGVFWVKKRNPKTQTEKLSFRKYDKVVRKHVLFKEEKIK\n\n### Sequence 2:\nMAKKNKNILVRLVSTAGTGFFLVKKRNPKTQTEKLSFRKYDPKVRKHVLFKEEKIK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00491", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nACTA2 FAP S100A4 PDGFRB\n</cell>", "output": "This cell is most consistent with cancer-associated fibroblast, a cancer-associated fibroblast (CAF). Key markers such as ACTA2, PDGFRB, FAP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "breast", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["ACTA2", "FAP", "S100A4", "PDGFRB"]}}
{"id": "eval_00492", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CCCCC1CCN(C(C)C(=O)NC(Cc2cc(F)cc(F)c2)C(O)C2Cc3ccccc3C[NH2+]2)C1=O", "output": "This molecule is relatively large and moderately polar, with moderately lipophilic character. It contains aromatic ring, amide, hydroxyl, halide. Its descriptor profile: MW=514.64, LogP=2.47, TPSA=86.25, HBD=3, HBA=3, QED=0.46.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BACE", "canonical_smiles": "CCCCC1CCN(C(C)C(=O)NC(Cc2cc(F)cc(F)c2)C(O)C2Cc3ccccc3C[NH2+]2)C1=O", "compound_name": null, "pubchem_cid": null, "MW": 514.64, "LogP": 2.47, "TPSA": 86.25, "QED": 0.46, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00493", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKLIIFTGLILFPIVSLIEAQAKNEKACLPQYQVCTDAPGNCCSNLICDCYGRYKNGARIGRNCFCLQKGVLYKKEN\n\n### Sequence 2:\nMVDSVFPAPSLVDETFLAPSSSETNDTAMAADGTLAKMNMFQFNSMLKEYSPPRFEDLFQDLDASKAGWEHYRPRLVSEVDLIDAEHYSLFSANTVSPL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00494", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSRIESLTRARTERSRELASKNREKEKMKEAKDARYTNGHLFTTISVSGMTMCYACNKSITAKEALICPTCNVTIHNRCKDTLANCTKVKQKQQKAALLKNSTALQSVSLRSKTTPRERPSSAIYPSDSFRQSLLGSRRGRSSLSLAKSVSTTNIAGHFSDESPLGLRRILSQSTDSLNMRNRALSVESLIDEGAEVIYSELMSDFETDERDFAADSWSLAVDSSFLQQQKKEVMKQQDVIYELIQTELHHVRTLKIMTRLFRTGMLEELQLEPGVVQGLFPCVDELSDIHTRFLSQLLERRRQALCPGSTRNFVIHRLADLLISQFSGPSAERMRKAYSEFCSRHTKALKLYKELYARDKRFQQFIRKVTRSAVLKRHGVQECILLVTQRITKYPVLINRILQHSHGTDEERQDLTTALGLVKELLSNVDQDVHELEKGARLQEIYHRMDPRAQAPVPSKGPFGREELLRRKLIHDGCLLWKTATGRFKDVLMLLMTDVLVFLQEKDQKYIFPALDKPSVVSLQNLIVRDIANQEKGMFLISAAPPEMYEVHTASRDDRSTWVRVIQQSVRVCPSREDFPLIETEDEAYLRRIKMELQQKDKALVELLREKVGLFAEMTHFQVEEDSGGVALPALPRGLFRSESLECPRGERLLQDAIREVEGLKDLLVGPGVELLLTPRDPALLVDPDSGGSTSPGVTANGEARNFNGSIELCRTDSDSSQKDRNGNQLRAPQEEALQRLVNLYGLLHGLQAAVAQQDTLMEARFPEGPERREKLARANSRDGEAGRVGPAPVAPDKQATELALLQRQHALLQEELRRCRRLGEERATEAGNLEARLRESEQARALLEREAEEARRQLAILGQSEPPPAEAPWARRPLDPRRRSLPAGDALYLSFTPPQPSRGHDRLDLSVTIRSVHRPFEDRERQELGSPEERLQDSSDPDTGSEEEGGGRLSPPHSPRDFTRMQDIPEETESRDGEPMVSES\n\n### Sequence 2:\nMSRIESLTRARTERSRELASKNREKEKMKEAKDARYTNGHLFTTISVSGMTMCYACNKSITAKEALICPTCNVTIHNRCKDTLANCTKVKQKQQKAALLKNSTALQSVSLRSKTTPRERPSSAIYPSDSFRQSLLGSRRGRSSLSLAKSVSTTNIAGHFSDESPLGLRRILSQSTDSLNMRNRALSVESLIDEGAEVIYSELMSDFETDERDFAADSWSLAVDSSFLQQQKKEVMKQQDVIYELIQTELHHVRTLKIMTRLFRTGMLEELQLEPGVVQGLFPCVDELSDIHTRFLSQLLERRRQALCPGSTRNFVIHRLADLLISQFSGPSAERMRKAYSEFCSRHTKALKLYKELYARDKRFQQFIRKVTRSAVLKRHGVQECILLVTQRITKYPVLINRILQHSHGTDEERQDLTTALGLVKELLSNVDQDVHELEKGARLQEIYHRMDPRAQAPVPSKGPFGREELLRRKLIHDGCLLWKTATGRFKDVLMLLMTDVLVFLQEKDQKYIFPALDKPSVVSLQNLIVRDIANQEKGMFLISAAPPEMYEVHTASRDDRSTWVRVIQQSVRVCPSREDFPLIETEDEAYLRRIKMELQQKDKALVELLREKVGLFAEMTHFQVEEDSGGVALPALPRGLFRSESLECPRGERLLQDAIREVEGLKDLLVGPGVELLLTPRDPALLVDPDSGGSTSPGVTANGEARNFNGSIELCRTDSDSSQKDRNGNQLRAPQEEALQRLVNLYGLLHGLQAAVAQQDTLMEARFPEGPERREKLARANSRDGEAGRVGPAPVAPDKQATELALLQRQHALLQEELRRCRRLGEERATEAGNLEARLRESEQARALLEREAEEARRQLAILGQSEPPPAEAPWARRPLDPRRRSLPAGDALYLSFTPPQPSRGHDRLDLSVTIRSVHRPFEDRERQELGSPEERLQDSSDPDTGSEEEGGGRLSPPHSPRDFTRMQDIPEETESRDGEPMVSES\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00495", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nRBFOX3:low SATB2:low SLC17A7:medium\n</cell>", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as SLC17A7, SATB2, RBFOX3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "tumor", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["RBFOX3", "SATB2", "SLC17A7"]}}
{"id": "eval_00496", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRIMI_HAEIN\n\n### Answer:", "output": "Acetylates the N-terminal alanine of ribosomal protein S18.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00497", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nVCAN:high CD14:high LYZ:medium FCN1:low S100A9:low S100A8:high\n</cell>", "output": "This cell is most consistent with CD14+ monocyte, a classical CD14+ monocyte. Key markers such as S100A9, LYZ, S100A8 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "CD14+ monocyte", "tissue": "colon", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["VCAN", "CD14", "LYZ", "FCN1", "S100A9", "S100A8"]}}
{"id": "eval_00498", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nMOBA_METJA\n\n### Answer:", "output": "Belongs to the MobA family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00499", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nBAKOR_RAT\n\n### Answer:", "output": "Beclin 1-associated autophagy-related key regulator, Barkor, Autophagy-related protein 14-like protein, Atg14L.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00500", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMVKLILVRHAESEWNPVGRYQGLLDPDLSERGKKQAKLLAQELSREHLDVIYSSPLKRTYLTALEIAEAKNLEVIKEDRIIEIDHGMWSGMLVEEVMEKYPEDFRRWVEEPHKVEFQGGESLASVYNRVKGFLEEVRKRHWNQTVVVVSHTVPMRAMYCALLGVDLSKFWSFGCDNASYSVIHMEERRNVILKLNITCHLGEFYVEAHKAI\n\n### Answer:", "output": "<SEQ_2D>CEEEEEEECCCBTTTTTTBCCTTCCCCBCHHHHHHHHHHHHHHTTSCCCEEEECSSHHHHHHHHHHHTTTTCCEEECGGGSCCCCGGGTTCBHHHHHHHSHHHHHHHHHCGGGCBCTTSCBHHHHHHHHHHHHHHHHHHTTTSEEEEEECHHHHHHHHHHHHTCCGGGGGGBCCCTTCEEEEEECSSCEEEEEEEECGGGGGGCCCCCTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00501", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nHHHATIKQRTLKNIIRATGVGLHSGEKVYLTLKPAPVDTGIVFCRTDLDPVVEIPARAENVGETTMSTTLVKGDVKVDTVEHLLSAMAGLGIDNAYVELSASEVPIMDGSAGPFVFLIQSAGLQEQEAAKKFIRIKREVSVEEGDKRAVFVPFDGFKVSFEIDFDHPVFRTQQASVDFTSFVKEVSRARTFGFMRDIEYLRSQNLALGGSVENAIVVDENRVLNEDGLRYEDEFVKHKILDAIGDLYLLGNSLIGEFRGFKSGHALNNQLLRTLIADKDAWEVVTFEDARTAPISYMRP\n\n### Answer:", "output": "<SEQ_3Di>DDFAWAFAKWFQDKFWDWAAFQQQRDIKIKIKFTDDPPAFEWEWQVLDVVIDIHHLALVQFDDFPPATWGDDVPAIATHQQQLLLLCLQQRHQDIYMYMNDRYDGLQQQFQQVVNVRRVVRNMDGDDHHAKAKFFAAKFKDDDDPWIKIWAGAQAEKEKEWEDDPAQLLCTDIDIDTPVGDNPQARPQGAEDEPVCVVVSVVVSTSPNDDNRRHQYHYRHDGPNPVDGPDSRSHVSVVVSSLSSLVCSNSHHYRTYMYTYNDGNVVSSVRSVVVVVPVVGIDMDIDRGPVPGSHDHDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00502", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "COc1ccc(N(C(=O)c2ccco2)C(C(=O)NC2CCCC2)c2ccccc2F)c(OC)c1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, ether, halide. Its descriptor profile: MW=466.51, LogP=4.88, TPSA=81.01, HBD=1, HBA=5, QED=0.51.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "COc1ccc(N(C(=O)c2ccco2)C(C(=O)NC2CCCC2)c2ccccc2F)c(OC)c1", "compound_name": null, "pubchem_cid": null, "MW": 466.51, "LogP": 4.88, "TPSA": 81.01, "QED": 0.51, "functional_groups": ["aromatic ring", "amide", "ether", "halide"], "moleculenet_labels": {"label": 3.09}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00503", "category": "Mutation", "instruction": "Describe the effect of mutation Q13190-Q79H on protein Syntaxin-5.", "input": "### Wild-type:\nMIPRKRYGSKNTDQGVYLGLSKTQVLSPATAGSSSSDIAPLPPPVTLVPPPPDTMSCRDRTQEFLSACKSLQTRQNGIQTNKPALRAVRQRSEFTLMAKRIGKDLSNTFAKLEKLTILAKRKSLFDDKAVEIEELTYIIKQDINSLNKQIAQLQDFVRAKGSQSGRHLQTHSNTIVVSLQSKLASMSNDFKSVLEVRTENLKQQRSRREQFSRAPVSALPLAPNHLGGGAVVLGAESHASKDVAIDMMDSRTSQQLQLIDEQDSYIQSRADTMQNIESTIVELGSIFQQLAHMVKEQEET\n\n### Mutant:\nMIPRKRYGSKNTDQGVYLGLSKTQVLSPATAGSSSSDIAPLPPPVTLVPPPPDTMSCRDRTQEFLSACKSLQTRQNGIHTNKPALRAVRQRSEFTLMAKRIGKDLSNTFAKLEKLTILAKRKSLFDDKAVEIEELTYIIKQDINSLNKQIAQLQDFVRAKGSQSGRHLQTHSNTIVVSLQSKLASMSNDFKSVLEVRTENLKQQRSRREQFSRAPVSALPLAPNHLGGGAVVLGAESHASKDVAIDMMDSRTSQQLQLIDEQDSYIQSRADTMQNIESTIVELGSIFQQLAHMVKEQEET\n\n### Answer:", "output": "In a breast cancer sample; somatic mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00504", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "IL2RA, CD3D, CD4", "output": "The expression pattern is consistent with a healthy regulatory T cell without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["IL2RA", "CD3D", "CD4"]}}
{"id": "eval_00505", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nDYLLSANGSKAIDFVWSNFLKHPYNPKLKQPDANVKAAWEKHADWEHWSNMGQDWSLNTPSGLVPRPNHGIAHTLRVAQLVPVIAEFLKAYSGDPKFQKLTQKEIQKAQYMMLFSVIGRENDMSWTDANHYQQAFKEAYNGQYSKHIYATFKENAQKGFLNHVVTNKSSLIPSLFSDESELQYWAEALDTISSASGILMALAHDLDLMRCYDKGKFNSLKMKDLVARLGGNEDAAKKLADYAHDLIVATGDRCMGYGVTQDYNYSLFGKCSLDPNECLKQLQSIPKPET\n\n### Answer:", "output": "<SEQ_3Di>DCCPPPLNVLLLVLLQVPFQQDFFAPVQQFDDPLLQVLLVVDPVDDHRNLSNPCRWDDFPLGIFGSLVQHLLLLLQLLLLLVLLLVLLCVFQPDVLSVVCDPLLSSLLSSLSSLLQRLGRTRHDQVSLVSSQVSVCVRPVHPGDDRSVLSSLVVSLVSLVVVCVVCVVVCDPRRPPDVVSVVVSSVVSRPAPHPSNPSSVLSVLLLCQLPDAAVCSCVVRLVVSCNSRVVDNVSSVLSSVLSVQLCVLQQGAGHHDDPTDHGPRHSVSVVRVDVVSSVVSNVPDDRDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00506", "category": "Mutation", "instruction": "Describe the effect of mutation Q540U1-K177R on protein Class B acid phosphatase.", "input": "### Wild-type:\nMKKITLALSAVCLLFTLNHSANALVSSPSTLNPGTNVAKLAEQAPVHWVSVAQIENSLTGRPPMAVGFDIDDTVLFSSPGFWRGKKTYSPDSDDYLKNPAFWEKMNNGWDEFSIPKEVARQLIDMHVRRGDSIYFVTGRSQTKTETVSKTLADNFHIPAANMNPVIFAGDKPEQNTKVQWLQEKNMRIFYGDSDNDITAARDCGIRGIRILRAANSTYKPLPQAGAFGEEVIVNSEY\n\n### Mutant:\nMKKITLALSAVCLLFTLNHSANALVSSPSTLNPGTNVAKLAEQAPVHWVSVAQIENSLTGRPPMAVGFDIDDTVLFSSPGFWRGKKTYSPDSDDYLKNPAFWEKMNNGWDEFSIPKEVARQLIDMHVRRGDSIYFVTGRSQTKTETVSKTLADNFHIPAANMNPVIFAGDKPEQNTRVQWLQEKNMRIFYGDSDNDITAARDCGIRGIRILRAANSTYKPLPQAGAFGEEVIVNSEY\n\n### Answer:", "output": "20-fold reduction in phosphatase activity. Nearly 2-fold increase in KM for pNPP in presence of 1 mM MgCl(2).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00507", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMFSB_ASPFU\n\n### Answer:", "output": "Major facilitator superfamily multidrug transporter mfsB.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00508", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMASGAARWLVLAPVRSGALRSGPSLRKDGDVSAAWSGSGRSLVPSRSVIVTRSGAILPKPVKMSFGLLRVFSIVIPFLYVGTLISKNFAALLEEHDIFVPEDDDDDD\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCTTSCCCCCCCCCTTCHHHHHHHHHHHHHHHHHHHHHHHHHHHTTCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00509", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDPS_SALSV\n\n### Answer:", "output": "During stationary phase, binds the chromosome non-specifically, forming a highly ordered and stable dps-DNA co-crystal within which chromosomal DNA is condensed and protected from diverse damages.  It protects DNA from oxidative damage by sequestering intracellular Fe(2+) ion and storing it in the form of Fe(3+) oxyhydroxide mineral, which can be released after reduction.  One hydrogen peroxide oxidizes two Fe(2+) ions, which prevents hydroxyl radical production by the Fenton reaction.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00510", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMGFPIPDPYVWDPSFRTFYSIIDDEHKTLFNGIFHLAIDDNADNLGELRRCTGKHFLNEQVLMQASQYQFYDEHKKEHETFIHALDNWKGDVKWAKSWLVNHIKTIDFKYKGKI\n\n### Sequence 2:\nGFPIPDPYCWDISFRTFYTIVDDEHKTLFNGILLLSQADNADHLNELRRCTGKHFLNEQQLMQASQYAGYAEHKKAHDDFIHKLDTWDGDVTYAKNWLVNHIKTIDFKYRGKI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00511", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "COc1ccc(CC(C)NCC(O)c2ccc(O)c(NC=O)c2)cc1", "output": "The SMILES represents an unnamed molecule with MW=344.41, LogP=2.22, TPSA=90.82. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "SIDER", "canonical_smiles": "COc1ccc(CC(C)NCC(O)c2ccc(O)c(NC=O)c2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 344.41, "LogP": 2.22, "TPSA": 90.82, "QED": 0.41, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 1, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 0, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00512", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: BRAF\nLigand SMILES: CCCN(CCCl)c1c([N+](=O)[O-])cc(C(F)(F)F)cc1[N+](=O)[O-]\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and B-Raf kinase (BRAF) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CCCN(CCCl)c1c([N+](=O)[O-])cc(C(F)(F)F)cc1[N+](=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 355.7, "LogP": 3.98, "TPSA": 89.52, "QED": 0.42, "functional_groups": ["aromatic ring", "nitro", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BRAF", "has_evidence": false}}
{"id": "eval_00513", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPURA_FRASN\n\n### Answer:", "output": "Belongs to the adenylosuccinate synthetase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00514", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: ALK\nLigand SMILES: CCCCCCCCCCCCCCCCC(=O)O\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CCCCCCCCCCCCCCCCC(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 270.46, "LogP": 5.94, "TPSA": 37.3, "QED": 0.37, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_00515", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nITGA2B PF4 PPBP GP1BA\n</cell>", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as GP1BA, PF4, ITGA2B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "heart", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["ITGA2B", "PF4", "PPBP", "GP1BA"]}}
{"id": "eval_00516", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMAHHHHHHGHHHQLPTLKVAYIPEHFSTPLFFAQQQGYYKAHDLSIEFVKVPEGSGRLINLLNSNEVDIAIGLTEAFIADIAKGNENIHVLDTYVKSPLLWAVSTGSNRDDVTDAKQLKRIGVSRIGSGSYVMSFVLAHQLGVPSFDQFQVLSNFKNLRDSVNLKDGVEGSDAFMWEYFTSKKYYDNHEIKQIDQIYTPWSSWVVATSSDSLQAKSDVIKNFIDAVNQGIQYYNEHVDEAIEYISSNLDYSAEDAKEWTKTVEFNSRIGKTPLDWDTIVVKTKDTLKLAGVLAESDDVILKRLNSNVKKTNLQLDGDLEAA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCEEEEECCSGGGHHHHHHHHTTHHHHTTCCEEEEECTTCHHHHHHHHHTTSCSEEEEEHHHHHHHHHTTCTTEEEEEEEECSCCEEEEEEESSCTTCSSGGGCCEEEESSTTSHHHHHHHHHHHHHTSCCCCEEEECCSHHHHHHHHTTCTTSCCCSEEEEEHHHHHHHHHTTSEEEEEEEECSSCSEEEEEEHHHHHHSHHHHHHHHHHHHHHHHHHHHCHHHHHHHHHHHSSCCHHHHHHHHTTCCBCTTBTTBCCCHHHHTHHHHHHHHHHTSCCSCHHHHHHHHHHHBCSSCCCCCCTTCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00517", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMAMKKLLIASLLFSSATVYGAEGFVVKDIHFEGLQRVAVGAALLSMPVRTGDTVNDEDISNTIRALFATGNFEDVRVLRDGDTLLVQVKERPTIASITFSGNKSVKDDMLKQNLEASGVRVGESLDRTTIADIEKGLEDFYYSVGKYSASVKAVVTPLPRNRVDLKLVFQEGVSAEIQQINIVGNHAFTTDELISHFQLRDEVPWWNVVGDRKYQKQKLAGDLETLRSYYLDRGYARFNIDSTQVSLTPDKKGIYVTVNITEGDQYKLSGVEVSGNLAGHSAEIEQLTKIEPGELYNGTKVTKMEDDIKKLLGRYGYAYPRVQSMPEINDADKTVKLRVNVDAGNRFYVRKIRFEGNDTSKDAVLRREMRQMEGAWLGSDLVDQGKERLNRLGFFETVDTDTQRVPGSPDQVDVVYKVKERNTGSFNFGIGYGTESGVSFQAGVQQDNWLGTGYAVGINGTKNDYQTYAELSVTNPYFTVDGVSLGGRLFYNDFQADDADLSDYTNKSYGTDVTLGFPINEYNSLRAGLGYVHNSLSNMQPQVAMWRYLYSMGEHPSTSDQDNSFKTDDFTFNYGWTYNKLDRGYFPTDGSRVNLTGKVTIPGSDNEYYKVTLDTATYVPIDDDHKWVVLGRTRWGYGDGLGGKEMPFYENFYAGGSSTVRGFQSNTIGPKAVYFPHQASNYDPDYDYECATQDGAKDLCKSDDAVGGNAMAVASLEFITPTPFISDKYANSVRTSFFWDMGTVWDTNWDSSQYSGYPDYSDPSNIRMSAGIALQWMSPLGPLVFSYAQPFKKYDGDKAEQFQFNIGKTW\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCBCCCEEEECCTTSCHHHHHHHCSCCTTCBCCHHHHHHHHHHHHHTTCEEEEEEEEETTEEEEEEEECCCEEEEEEESCSSSCHHHHHHHHHTTSCCSSSCCCGGGHHHHHHHHHHHTTTTTCSSCCCEEEEEECSSSCEEEEEECCCCCCCCEEEEEEESCSSSCHHHHHTTTCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHHHHTTCTTCEEEEEEEEECTTSSSEEEEEEEECCCCCEEEEEEEEEECTTCHHHHHHHTCCCTTCBCCHHHHHHHHHHHHHHHHHHTCCSCEEEEEEEECCSSCEEEEEEEEECCCCCCEEEEEEESCSSSCHHHHHTTCSSCTTSCCCHHHHHHHHHHHHHHTCEEEEEEEEEECTTCSSCEEEEEEEEECCCCEECCCCCCCSSSSCCCCCCEECSSSSSSCCBCEEEEEECSSEEEEECCEEESCSSTTCCEEEEEEEEEEECCCSSSSSSCEEEEEEEEEEEEEEEETTEEEEEEEEEEEEEEESCCCBHHHHHHHHHHTCCCCSSCCCCCEEEEEEEEEEEEEEECCSCSSSCCSEEEEEEEEEEECTTCSSCEEEEEEEEEEEEECSTTSCSEEEEEEEEEEEEESSSCCCCSSSCBCBCSTTSCTTBCTTCBSCEEEECCBCCSSCCCTTSSCSSBTTTCCSCEEEEEECCBSEEEEEEEEEEECCSSCCCCCCCCEEEEEEEEEEEEECTTCCTTTTTTSCCTTCTTCCEEEEEEEEEEEETTEEEEEEEEEEEECCTTCCBCSEEEECCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00518", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD3E, CD8A, LEF1, CD3D, CCR7, SELL, CD8B", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as LEF1, CD8A, CCR7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "pancreas", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD3E", "CD8A", "LEF1", "CD3D", "CCR7", "SELL", "CD8B"]}}
{"id": "eval_00519", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "FCER1A, CLEC10A, HLA-DQA1", "output": "This profile is most consistent with conventional dendritic cell, which is commonly found in intestine under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "intestine", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["FCER1A", "CLEC10A", "HLA-DQA1"]}}
{"id": "eval_00520", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSRPAKIGATAALSQLQGWAVVSEKDAIGKTYLFADFNAAFGFMTRVALMADKLDHHPEWFNVYNRVEVVLTTHDADGVSDLDVTMARFMDGIATA\n\n### Sequence 2:\nMNPLIQSLTEGQLRTDIPSFRPGDTVRVHAKVVEGNRERIQIFEGVVIARKGAGISENYTVRKISNGIGVERIFPIHTPRVEKIEVVRYGKVRRAKLYYLRALQGKAARIKEIRR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00521", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMTETANLFVLGAAGGVEWGTVIVQVLTFIVLLALLKKFAWGPLKDVMDKRERDINRDIDDAEQAKLNAQKLEEENKQKLKETQEEVQKILEDAKVQARQQQEQIIHEANVRANGMIETAQSEINSQKERAIADINNQVSELSVLIASKVLRKEISEQDQKALVDKYLKEAGDK\n\n### Sequence 2:\nMAPKKKVSALIKLQIQAGKANPAPPLGPALGSHGVNIMDFCKAYNAQTQDKMGQVIPVEITVYEDRSFTFVLKTPPAAALLKKAAGIEKGTENPLTHKVGSVTKAQVREIAETKMEDLSARDIEAGMKIIEGTARSMGITVTD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00522", "category": "Mutation", "instruction": "Describe the effect of mutation P33535-T279D on protein Mu-type opioid receptor.", "input": "### Wild-type:\nMDSSTGPGNTSDCSDPLAQASCSPAPGSWLNLSHVDGNQSDPCGLNRTGLGGNDSLCPQTGSPSMVTAITIMALYSIVCVVGLFGNFLVMYVIVRYTKMKTATNIYIFNLALADALATSTLPFQSVNYLMGTWPFGTILCKIVISIDYYNMFTSIFTLCTMSVDRYIAVCHPVKALDFRTPRNAKIVNVCNWILSSAIGLPVMFMATTKYRQGSIDCTLTFSHPTWYWENLLKICVFIFAFIMPVLIITVCYGLMILRLKSVRMLSGSKEKDRNLRRITRMVLVVVAVFIVCWTPIHIYV\n\n### Mutant:\nMDSSTGPGNTSDCSDPLAQASCSPAPGSWLNLSHVDGNQSDPCGLNRTGLGGNDSLCPQTGSPSMVTAITIMALYSIVCVVGLFGNFLVMYVIVRYTKMKTATNIYIFNLALADALATSTLPFQSVNYLMGTWPFGTILCKIVISIDYYNMFTSIFTLCTMSVDRYIAVCHPVKALDFRTPRNAKIVNVCNWILSSAIGLPVMFMATTKYRQGSIDCTLTFSHPTWYWENLLKICVFIFAFIMPVLIITVCYGLMILRLKSVRMLSGSKEKDRNLRRIDRMVLVVVAVFIVCWTPIHIYV\n\n### Answer:", "output": "Receptor inactivation. In contrast, the T6. 34(279)D substitution did not increase basal [35S]GTPgammaS binding, greatly reduced agonist-promoted [35S]GTPgammaS binding, and markedly decreased affinity for DAMGO. Thus, the T6. 34(279)D mutant adopts conformations corresponding to inactive states of the receptor. The results were interpreted in the structural context of a model for the mu opioid receptor that incorporates the information from the crystal structure of rhodopsin. The interaction of T6. 34(279) with R3. 50(165) in the mu opioid receptor is considered to stabilize the inactive conformations. The T6. 34(279)K substitution would then disrupt this interaction and support agonist-free activation, while T6. 34(279)D mutation should strengthen this interaction which keeps the receptor in inactive states. T6. 34(279) may, in addition, interact with the neighboring R6. 35(280) to help constrain the receptor in inactive states, and T6. 34(279)K and T6. 34(279)D mutations would affect this interaction by disrupting or strengthening it, respectively. The T6. 34(279)D mutant produced an inactive form of the mu opioid receptor.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00523", "category": "Mutation", "instruction": "Describe the effect of mutation E0X9C7-A46F on protein Sensor histidine kinase TodS.", "input": "### Wild-type:\nMSSLDRKKPQNRSKNNYYNICLKEKGSEELTCEEHARIIFDGLYEAVGLLDAHGNVLEVNQVALEGAGITLEEIRGKPFWKARWWQISKKTEATQKRLVETASSGEFVRCDVEILGKSGGREVIAVDFSLLPICNEEGSIVYLLAEGRNITDKKKAEAMLALKNQELEQSVERIRKLDNAKSDFFAKVSHELRTPLSLILGPLEAVMAAEAGRESPYWKQFEVIQRNAMTLLKQVNTLLDLAKMDARQMGLSYRRANLSQLTRTISSNFEGIAQQKSITFDTKLPVQMVAEVDCEKYERI\n\n### Mutant:\nMSSLDRKKPQNRSKNNYYNICLKEKGSEELTCEEHARIIFDGLYEFVGLLDAHGNVLEVNQVALEGAGITLEEIRGKPFWKARWWQISKKTEATQKRLVETASSGEFVRCDVEILGKSGGREVIAVDFSLLPICNEEGSIVYLLAEGRNITDKKKAEAMLALKNQELEQSVERIRKLDNAKSDFFAKVSHELRTPLSLILGPLEAVMAAEAGRESPYWKQFEVIQRNAMTLLKQVNTLLDLAKMDARQMGLSYRRANLSQLTRTISSNFEGIAQQKSITFDTKLPVQMVAEVDCEKYERI\n\n### Answer:", "output": "Increases affinity for agonists and antagonists.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00524", "category": "Mutation", "instruction": "Describe the effect of mutation O59747-A373R on protein Palmitoyl-protein thioesterase-dolichyl pyrophosphate phosphatase fusion 1 [Cleaved into: Palmitoyl-protein thioesterase.", "input": "### Wild-type:\nMLSCSSFLIFFLFSWVLLPMKSFAIPIISLDKVRLAINDGASEQLPVVIWHGLGDTPTSFTLTEVSQRVQKLTKGAVYAIRVGDNEFEDIKAGYLGKLEDQLDEVCDLIGNEDSLSNGFYALGLSQGGLFLRALAQTCDAAKIRSLITLGSPHSGINTIPGCSPTNLICKAVVHSILGLGIWHSWIQNHVVQAQYYRTEKQYDKYLENNKFLTHLNNEVLHDNYTRNIEKLKELDNLVAVSFERDDIVEPPYSTGFGWINETTGENIEMEDFVLYESLGLKDLVNQGKLETISFPGRHLQ\n\n### Mutant:\nMLSCSSFLIFFLFSWVLLPMKSFAIPIISLDKVRLAINDGASEQLPVVIWHGLGDTPTSFTLTEVSQRVQKLTKGAVYAIRVGDNEFEDIKAGYLGKLEDQLDEVCDLIGNEDSLSNGFYALGLSQGGLFLRALAQTCDAAKIRSLITLGSPHSGINTIPGCSPTNLICKAVVHSILGLGIWHSWIQNHVVQAQYYRTEKQYDKYLENNKFLTHLNNEVLHDNYTRNIEKLKELDNLVAVSFERDDIVEPPYSTGFGWINETTGENIEMEDFVLYESLGLKDLVNQGKLETISFPGRHLQ\n\n### Answer:", "output": "Enabled cleavage of the linker domain. Activating mutations in the Dolpp1p domain do not cause lethality, whereas mutations in the Ppt1p domain result in cells that are non-viable but normally sensitive to sodium orthovanadate and decreased extracellular pH.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00525", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD3D, CD45RO, IL7R, S100A4, CD4", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD45RO, IL7R, S100A4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "brain", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD3D", "CD45RO", "IL7R", "S100A4", "CD4"]}}
{"id": "eval_00526", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYNFB_SALAR\n\n### Answer:", "output": "UPF0482 protein YnfB.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00527", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL1_CLOK1\n\n### Answer:", "output": "Protein L1 is also a translational repressor protein, it controls the translation of the L11 operon by binding to its mRNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00528", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y2U9-K241R on protein Kelch domain-containing protein 2.", "input": "### Wild-type:\nMADGNEDLRADDLPGPAFESYESMELACPAERSGHVAVSDGRHMFVWGGYKSNQVRGLYDFYLPREELWIYNMETGRWKKINTEGDVPPSMSGSCAVCVDRVLYLFGGHHSRGNTNKFYMLDSRSTDRVLQWERIDCQGIPPSSKDKLGVWVYKNKLIFFGGYGYLPEDKVLGTFEFDETSFWNSSHPRGWNDHVHILDTETFTWSQPITTGKAPSPRAAHACATVGNRGFVFGGRYRDAKMNDLHYLNLDTWEWNELIPQGICPVGRSWHSLTPVSSDHLFLFGGFTTDKQPLSDAWTY\n\n### Mutant:\nMADGNEDLRADDLPGPAFESYESMELACPAERSGHVAVSDGRHMFVWGGYKSNQVRGLYDFYLPREELWIYNMETGRWKKINTEGDVPPSMSGSCAVCVDRVLYLFGGHHSRGNTNKFYMLDSRSTDRVLQWERIDCQGIPPSSKDKLGVWVYKNKLIFFGGYGYLPEDKVLGTFEFDETSFWNSSHPRGWNDHVHILDTETFTWSQPITTGKAPSPRAAHACATVGNRGFVFGGRYRDARMNDLHYLNLDTWEWNELIPQGICPVGRSWHSLTPVSSDHLFLFGGFTTDKQPLSDAWTY\n\n### Answer:", "output": "Affects ability to recognize truncated SELENOK with a diglycine (Gly-Gly) at the C-terminus. Enables ability to recognize cleaved USP1 with a diglycine (Gly-Gly) at the C-terminus. The crystal structures of KLHDC2 in complex with the diglycine-ending C-end degrons of two early-terminated selenoproteins and the N-terminal proteolytic fragment of USP1 were reported. The E3 recognizes the degron peptides in a dissimilarly coiled conformation and discards their C-terminal diglycine with a shallow surface pocket. By disrupting hydrogen bonding with multiple backbone carbonyls of the peptides, KLHDC2 further releases the otherwise degenerate degrons with a loose interface and expected low affinities.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00529", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CN1CCN(CC(=O)Nc2ccc(-c3cccn4c(=O)cc(N5CCOCC5)nc34)c3sc4ccccc4c23)CC1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, ether, pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "CN1CCN(CC(=O)Nc2ccc(-c3cccn4c(=O)cc(N5CCOCC5)nc34)c3sc4ccccc4c23)CC1", "compound_name": null, "pubchem_cid": null, "MW": 568.7, "LogP": 3.75, "TPSA": 82.42, "QED": 0.35, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "pyridine"], "moleculenet_labels": {"label": 2.6}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "ether", "pyridine"]}}
{"id": "eval_00530", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CC(C)(Cl)[N+](=O)[O-]", "output": "The molecule contains the following functional groups: amine, nitro, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "CC(C)(Cl)[N+](=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 123.54, "LogP": 1.24, "TPSA": 43.14, "QED": 0.23, "functional_groups": ["amine", "nitro", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amine", "nitro", "halide"]}}
{"id": "eval_00531", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGFRA4_MOUSE\n\n### Answer:", "output": "GDNF family receptor alpha-4, GDNF receptor alpha-4, GDNFR-alpha-4, GFR-alpha-4, Persephin receptor.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00532", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: CC(C)Cl\nDescriptors: MW=78.54, LogP=1.63, TPSA=0.0, HBD=0, HBA=0", "output": "Descriptors: MW=78.54, LogP=1.63, TPSA=0.0, HBD=0, HBA=0, QED=0.39, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "ESOL", "canonical_smiles": "CC(C)Cl", "compound_name": null, "pubchem_cid": null, "MW": 78.54, "LogP": 1.63, "TPSA": 0.0, "QED": 0.39, "functional_groups": ["halide"], "moleculenet_labels": {"label": -1.41}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00533", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMCPRPRRAVLLGLGVAMSAIAGCRETAPSTTQTSDGGPEPTGRSGDTQTDGGGETTRQATVAEGGESATRADETETSELDLREANVVDVTLQSAGGRAVEFSVTLYHDDDGEDGYADWWQVETLAGDRLGRRELLHAHSTAPFTRSETIEVPEGTTCVVVRGHDQTHGYGGQAMVVDTESGATRAVQQGPEPAEFDDGDCP\n\n### Sequence 2:\nMPRPENPIQLAVIGAAHGTRGEVRVKTFTGDPLAIADYGLLYDEQGKAYEILEARVAKTVVIVRFKGVNDRNAAEALNGTELFIDRSQLPDEELDEDEFFQTDLIGLEAVDGDGKSYGVVSAIFDFGGGDLIELSEKGKRPMLIPFTEAAVPEIDFDKGIIKVEPHAAGLIADEHDNPPHESGKKPKKP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00534", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTCPD_PONAB\n\n### Answer:", "output": "Belongs to the TCP-1 chaperonin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00535", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNYQVLLYYKYTTIDDPEQFAQDHLAFCKAHHLKGRILVSTEGINGTLSGTKEETEQYMAHMHADERFKDMVFKIDEAEGHAFKKMHVRPRKEIVALDLEDDVDPRHTTGQYLSPVEFRKALEDDDTVIIDARNDYEFDLGHFRGAIRPNITRFRDLPDWIKENKALFADKKVVTYCTGGIRCEKFSGWLLKEGFEDVAQLHGGIATYGKDPETKGEYWDGKMYVFDDRISVNINQVEKTIIGKDWFDGKPCERYINCANPECNKQILVSEENETKYLGACSYECAKHERNRYVQANNISDNEWQQRLTNFDDLHQHA\n\n### Sequence 2:\nQYRVLLYYKYVHIDNPEQFAEDHLKFCKDLGLKGRILVAGEGINGTVSGTVEQTDRYMSEMKSDPRFEDMVFKIDESEGHAFKKMHVRHRDELVTLRLEDDIDPNELTGKYLEPKEFYEAMQQEDTIVVDARNDYEYDLGHFRGAIRPDIKAFRELPEWIRDNKEKLEGKKILTYCTGGIRCEKFSGWLKKEGFEDVSQLHGGIVTYGKDPEVQGELWDGKCYVFDERISVPVNQKEHVIVGKDYFTGEPCERYVNCANPECNKQIICSEENEHRYLRGCTHECRVHPRNLYVKEHGLSEEEVQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00536", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: CDK2\nLigand SMILES: Nc1ncc(-c2ccc(F)cc2)[nH]1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and cyclin-dependent kinase 2 (CDK2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "Nc1ncc(-c2ccc(F)cc2)[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 177.18, "LogP": 1.8, "TPSA": 54.7, "QED": 0.7, "functional_groups": ["aromatic ring", "halide", "imidazole"], "moleculenet_labels": {"label": 1.39}, "split": "train", "generation_method": "template_protein_ligand", "target": "CDK2", "has_evidence": false}}
{"id": "eval_00537", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nEBO_DROME\n\n### Answer:", "output": "Belongs to the NRP synthetase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00538", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTARPLSELVEQGWAAALEPVVDQVAEMGRFLRAEIAAGRRYLPAGHSVLRAFTYPFDNVRVLIVGQDPYPTPGHAVGLSFSVAPDVRPLPRSLANVFDEYTADLGYPLPVCGDLTPWAQRGVLLLNRVLTVRPSNPASHRGKGWEVITECAIRALAARSEPMVAILWGRDAATLKPLLTVDNCVVIESPHPSPLSASRGFFGSRPFSRTNEILVGMGAGPINWRLP\n\n### Sequence 2:\nMKRKNIALIPAAGIGARFGADKPKQYVEIGSKTVLEHTIGIFERHEAVDLTVVVVSPEDTFADKVQTAFPQVRVWKNGGQTRAETVRNGVAKLLEIGLASADDNILVHDAARCCLPSEALTRLIEQAGNAAEGGILAISIADTLKRAEGGQISATVERTSLWQAQTPQLFRAGLLHRALAAENLDGITDEASAVEKLGIRPLLVQGDARNLKLTQPQDAYIVRLLLNAV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00539", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSRYLGSITKKSRRYGFSLLETEREFIKGKKRTYAPGQHGNKRVKLSGYGEQLQEKQKMMYLYGLNDRQFRRTFVIAKHMKGALTLNTFIALESRLDNLVYRMGFAPTRRAARQLVNHGHILLDGKKVTIPSCMVKLEQTIEVAPKSKDLPIVAAGASNTPCKFVDSDLKTKKGKYVRFPERDELPEGINEAYVVEWFNRLV\n\n### Sequence 2:\nMSRYLGSITKKSRRYGFSLLETEREFIKGKKRTYAPGQHGNKRVKLSGYGEQLQEKQKMMYLYGLNDRQFRRTFVIAKHMKGALTLNTFIALESRLDNLVYRMGFAPTRRAARQLVNHGHILLDGKKVTIPSCMVKLEQTIEVAPKSKDLPIVAAGASNTPCKFVDSDLKTKKGKYVRFPERDELPEGINEAYVVEWFNRLV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00540", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLAEALPWLQHFRDKIVVVKYGGNAMVDDSLKAAFAADMVFLRTVGAKPVVVHGGGPQISDMLRRVGLEGEFKGGFRVTTPEVMEIVRMVLFGQVGRDLVGLINSHGPYAVGTSGEDAGLFTAEKRLVDIDGTPTDIGLVGNIVNVDATSLMDLIDAGRIPVVSTIAPGADGQVYNINADTAAGALASAIGAERLLVLTNVEGLYTDWPNKSSLVSKIVASELDAILPGLDAGMIPKMESCLNAVRGGVNAAHVIDGRIAHSVLLELLTMGGIGTMVLPDNYAREDYPEGTVFRKNYQDGHA\n\n### Sequence 2:\nMQHFRDKIVVVKYGGNAMVDDDLKAAFAADMVFLRTVGAKPVVVHGGGPQISEMLNRVGLQGEFKGGFRVTTPEVMDIVRMVLFGQVGRDLVGLINSHGPYAVGTSGEDAGLFTAQKRMVNIDGVPTDIGLVGDIINVDASSLMDIIEAGRIPVVSTIAPGEDGQIYNINADTAAGALAAAIGAERLLVLTNVEGLYTDWPDKSSLVSKIKATELEAILPGLDSGMIPKMESCLNAVRGGVSAAHVIDGRIAHSVLLELLTMGGIGTMVLPDVFDRENYPEGTVFRKDDKDGE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00541", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD45RO, CD3D, S100A4, CD4", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD45RO, CD3D, S100A4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "brain", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD45RO", "CD3D", "S100A4", "CD4"]}}
{"id": "eval_00542", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTKYEILYIIRPNIDEEAKAALVERFDGILTDNGAANLESKDWEKRKLAYEINDFREGIYHIATFEAETTSEALSEFDRLAKINLDILRHMIVKVEA\n\n### Sequence 2:\nMTDKTITRAQLSEAVYQEVGLSRNESADLLEAVLDEISGALAKGDAVKISSFGSFSVRSKGQRIGRNPKTGEEVPITPRRVLVFRPSQLLKKKINDGMAAKRGGAK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00543", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nMYH6 TNNI3 MYH7 TNNT2\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as MYH6, TNNI3, MYH7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "heart", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MYH6", "TNNI3", "MYH7", "TNNT2"]}}
{"id": "eval_00544", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNADO1_ORYSJ\n\n### Answer:", "output": "Probable NAD(P)H-dependent oxidoreductase 1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00545", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: HIV protease\nLigand SMILES: O=[N+]([O-])c1ccc2ccc3cccc4ccc1c2c34\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "O=[N+]([O-])c1ccc2ccc3cccc4ccc1c2c34", "compound_name": null, "pubchem_cid": null, "MW": 247.25, "LogP": 4.49, "TPSA": 43.14, "QED": 0.28, "functional_groups": ["aromatic ring", "nitro"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": null, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 1.0, "NR-PPAR-gamma": null, "SR-ARE": 1.0, "SR-ATAD5": null, "SR-HSE": null, "SR-MMP": 1.0, "SR-p53": null}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00546", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMERIRLRLKAYDHRVLDRSVASIVEAVKRTGSEIRGPVPLPTKKKRYTVLRSPHINKDSREQFEIRVHHRIIDIMSATPDTVDSLMKLDLAPEVDVEVMSMSK\n\n### Sequence 2:\nMFTINAEVRKEQGKGASRRLRAANKFPAIIYGGKEAPLAIELDHDKVMNMQAKAEFYSEVLTIVVDGKEIKVKAQDVQRHPYKPKLQHIDFVRA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00547", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "CXCR2, S100A9, CSF3R", "output": "The expression pattern is consistent with a healthy neutrophil without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "tumor", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CXCR2", "S100A9", "CSF3R"]}}
{"id": "eval_00548", "category": "Mutation", "instruction": "Describe the effect of mutation Q9XW26-F344I on protein Specificity protein transcription factor 3.", "input": "### Wild-type:\nMSKSGQQQQPHQQMQQQQQQQGKMMVVHHQRTTTPDGSQVYLLQPSQSHPQQRQVQFIPIQVAQGEKKPQTVMMNQMMGQQQQQQPSSSNQQSQQQVQQQQQQVHQVQQVQQHQMQQQNQTVQQQHLMMETIDQVGSGIIGNSNGNNQQQQQQQQQQQQRRDPIPIAPAGMVSGQRMQQQQQQPQQQQQHNNNNTGASTSNGGVNAAPVMPVMGGQQRITLGNLHFQQDPNDPQKWIITNEGPPVAGPSAPARKPNQIQHRQMIPQMNEDGMMNMGNDYGDINGQGLGEAPKRCACTCPN\n\n### Mutant:\nMSKSGQQQQPHQQMQQQQQQQGKMMVVHHQRTTTPDGSQVYLLQPSQSHPQQRQVQFIPIQVAQGEKKPQTVMMNQMMGQQQQQQPSSSNQQSQQQVQQQQQQVHQVQQVQQHQMQQQNQTVQQQHLMMETIDQVGSGIIGNSNGNNQQQQQQQQQQQQRRDPIPIAPAGMVSGQRMQQQQQQPQQQQQHNNNNTGASTSNGGVNAAPVMPVMGGQQRITLGNLHFQQDPNDPQKWIITNEGPPVAGPSAPARKPNQIQHRQMIPQMNEDGMMNMGNDYGDINGQGLGEAPKRCACTCPN\n\n### Answer:", "output": "In gu85; fails to bind DNA. Defects in the morphology of vulval structures, resulting in protruding vulva and occasional bivulva phenotypes in adults. Reduced brood size. Embryonic and L1 larval arrest and morphological defects.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00549", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: COc1ccc2cc([C@H](C)C(=O)OCCCCO[N+](=O)[O-])ccc2c1\nDescriptors: MW=347.37, LogP=3.48, TPSA=87.9, HBD=0, HBA=6", "output": "Descriptors: MW=347.37, LogP=3.48, TPSA=87.9, HBD=0, HBA=6, QED=0.3, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "ClinTox", "canonical_smiles": "COc1ccc2cc([C@H](C)C(=O)OCCCCO[N+](=O)[O-])ccc2c1", "compound_name": null, "pubchem_cid": null, "MW": 347.37, "LogP": 3.48, "TPSA": 87.9, "QED": 0.3, "functional_groups": ["aromatic ring", "ester", "amine", "ether", "nitro"], "moleculenet_labels": {"FDA_APPROVED": 0, "CT_TOX": 1}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00550", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nPDCD1:medium CD8A:low LAG3:high\n</cell>", "output": "This cell is most consistent with exhausted CD8 T cell, an exhausted CD8+ T cell. Key markers such as CD8A, LAG3, PDCD1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "colon", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["PDCD1", "CD8A", "LAG3"]}}
{"id": "eval_00551", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY1289_PSEPK\n\n### Answer:", "output": "UPF0246 protein PP_1289.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00552", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSYS_XYLFT\n\n### Answer:", "output": "Catalyzes the attachment of serine to tRNA(Ser).  Is also able to aminoacylate tRNA(Sec) with serine, to form the misacylated tRNA L-seryl-tRNA(Sec), which will be further converted into selenocysteinyl-tRNA(Sec).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00553", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL20_MYCGI\n\n### Answer:", "output": "50S ribosomal protein L20.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00554", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPHCA_APHFL\n\n### Answer:", "output": "C-phycocyanin alpha subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00555", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMEFLERTYLVNDQATKMYAFPLDRELRQKPVNKDQCPGDRPEHPEAGGIYHCHNSAKATGNRSSKQAHAKWRLCAASAICFIFMVAEVVGGHVAGSLAILTDAAHLLIDLTSFLLSLFSLWLSSRPPSKRLTFGWYRAEILGALLSVLCIWVVTGVLLYLACERLLYPDYQIQAGIMITVSGCAVAANIVLTMILHQRNFGYNHKDVQANASVRAAFVHALGDVFQSISVLISALIIYFKPDYKIADPVCTFIFSILVLASTVMILKDFSILLMEGVPKGLSYNSVKEIILAVDGVISVHSLHIWSLTVNQVILSVHVATAASQDSQSVRTGIAQALSSFDLHSLTIQIESAADQDPSCLLCEDPQD\n\n### Sequence 2:\nMDSLQPLFEQLFGSEWGGPLFLLIKNLLLILAIVIPLLLAVAYLTFAERKIIAYMQVRVGPNRVTFFDIPWLRGWGQPIADAVKAIMKEIIIPTGANKFLFLLAPVLAIGPALAAWAVVPFSPELVLADINAGLLYILAMTSLGVYGVIIAGWASNSKYAFLGAMRSAAQVVSYELAMGFALVCVLMMSSSLNLGDIVAGQQGGSFLNWYLIPLFPMFLVYFISGVAETNRAPFDVAEGESEIVAGFHVDYSGMAFTVFFLAEYANMILVATLASIMFLGGWLPPVDIAPFNLIPGMVWLLLKIAFMLFFFLWFRATFPRYRYDQIMRLGWKVFIPLTLVWIVVLGMVMQLPEVVRQSFPLNLWFN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00556", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGFKQDIATIRGDLRTYAQDIFLAFLNKYPDERRYFKNYVGKSDQELKSMAKFGDHTEKVFNLMMEVADRATDCVPLASDANTLVQMKQHSSLTTGNFEKLFVALVEYMRASGQSFDSQSWDRFGKNLVSALSSAGMK\n\n### Answer:", "output": "<SEQ_2D>CHHHHHHHHHHTHHHHHHHHHHHHHHHCGGGGGGCCCCTTCCHHHHHTSSSHHHHHHHHHHHHHHHHHHCBTTBCCHHHHHHHHHCGGGTTCCHHHHHHHHHHHHHHHHHSSSCCCHHHHHHHHHHHHHHHHHTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00557", "category": "Mutation", "instruction": "Describe the effect of mutation Q9VBC7-L557S on protein Neuroendocrine convertase 2.", "input": "### Wild-type:\nMAAATWSWLLAPFLLLHWASAGAGGGAGGSGAGLSGPAVFTSSFLVRFRRGVDNSFAHDVADKYGFDNLGPLVGADGHEYHFKHRTLPHARSRRSLTHTRALKSHPAVHTAVQQPGFKRVKRGLRPAVPAIHGMKFDLKVGEGNRIDEEPTDPYFPMQWYLKNTGQNGGKVRLDLNVQAAWAQGITGKNVTTAIMDDGVDYMHPDLKFNYNAEASYDFSSNDPFPYPRYTDDWFNSHGTRCAGEVAAARDNGICGVGVAYDSKIAGIRMLDQPYMTDLIEANSMGHEPHKIHIYSASWGP\n\n### Mutant:\nMAAATWSWLLAPFLLLHWASAGAGGGAGGSGAGLSGPAVFTSSFLVRFRRGVDNSFAHDVADKYGFDNLGPLVGADGHEYHFKHRTLPHARSRRSLTHTRALKSHPAVHTAVQQPGFKRVKRGLRPAVPAIHGMKFDLKVGEGNRIDEEPTDPYFPMQWYLKNTGQNGGKVRLDLNVQAAWAQGITGKNVTTAIMDDGVDYMHPDLKFNYNAEASYDFSSNDPFPYPRYTDDWFNSHGTRCAGEVAAARDNGICGVGVAYDSKIAGIRMLDQPYMTDLIEANSMGHEPHKIHIYSASWGP\n\n### Answer:", "output": "In amonS557L1 and amonS557L2: Partial embryonic viability and enhanced larval growth with pupariation and survival to adulthood.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00558", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMEMEKEFEQIDKSGSWAAIYQDIRHEASDFPSRVAKLPKNKNRNRYRDVSPFDHSRIKLHQEDNDYINASLIKMEEAQRSYILTQGPLPNTVGHFWEMVWEQKSRGVVMLNRVMEKGSLKCAQYWPQKEEKEMIFEDTNLKLTLISEDIKSYYTVRQLELENLTTQETREILHFHYTTWPDFGVPESPASFLNFLFKVRESGSLSPEHGPVVVHCSAGIGRSGTFCLADTCLLLMDKRKDPSSVDIKKVLLEMRKFRMGLIQTADQLRFSYLAVIEGAKFIMGDSSVQDQWKELSHEDLEPPPEHIPPPPRPPKRILEPHN\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHHHHHHHTCHHHHHHHHHHHSCCCCCTTTTSGGGGGGCSCTTCCCCGGGEEECSCSSCCEEEEEEEEETTTTEEEEEECCCCTTTHHHHHHHHHHTTBCEEEECSCSEETTEECSCCCSCSSTTSCEEETTTTEEEEEEEEEECSSEEEEEEEEEETTTTEEEEEEEEEECCCCSSSCCSCSHHHHHHHHHHHHTTTTSTTSBCEEEESSSSSSHHHHHHHHHHHHHHHHHSSCGGGCCHHHHHHHHTTTSTTCSCSHHHHHHHHHHHHHHHHHHCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00559", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: VEGFR2\nLigand SMILES: CC(=O)Nc1ccc(C=C2SC(=S)N(C=C(c3ccccc3)S(=O)Cc3ccccc3)C2=O)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "CC(=O)Nc1ccc(C=C2SC(=S)N(C=C(c3ccccc3)S(=O)Cc3ccccc3)C2=O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 518.69, "LogP": 5.79, "TPSA": 66.48, "QED": 0.32, "functional_groups": ["aromatic ring", "amide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_00560", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMHIALIAHDEKKDLMVGFATAYKHLLEPHQLYATGTTGLRIIEATGLTVHRFKSGPLGGDQQIGARISENKMDLVIFLRDPLTAQPHEPDVTALIRLCDVYEIPLATNIGTAEILIRGLGAGFLDWRDLRRNDE\n\n### Sequence 2:\nMKIALIAHDRKKTLMIKLATAYKHILEKHELYATGTTGMKVMEATGLPVHCFKSGPLGGDQQIGAMISEDNIDLVIFLRDPLSAQPHEPDVTALIRLSDVYEIPLATNIGSAEILLRGVEAGFADFREV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00561", "category": "Mutation", "instruction": "Describe the effect of mutation O15399-V845M on protein Glutamate receptor ionotropic, NMDA 2D.", "input": "### Wild-type:\nMRGAGGPRGPRGPAKMLLLLALACASPFPEEAPGPGGAGGPGGGLGGARPLNVALVFSGPAYAAEAARLGPAVAAAVRSPGLDVRPVALVLNGSDPRSLVLQLCDLLSGLRVHGVVFEDDSRAPAVAPILDFLSAQTSLPIVAVHGGAALVLTPKEKGSTFLQLGSSTEQQLQVIFEVLEEYDWTSFVAVTTRAPGHRAFLSYIEVLTDGSLVGWEHRGALTLDPGAGEAVLSAQLRSVSAQIRLLFCAREEAEPVFRAAEEAGLTGSGYVWFMVGPQLAGGGGSGAPGEPPLLPGGAPL\n\n### Mutant:\nMRGAGGPRGPRGPAKMLLLLALACASPFPEEAPGPGGAGGPGGGLGGARPLNVALVFSGPAYAAEAARLGPAVAAAVRSPGLDVRPVALVLNGSDPRSLVLQLCDLLSGLRVHGVVFEDDSRAPAVAPILDFLSAQTSLPIVAVHGGAALVLTPKEKGSTFLQLGSSTEQQLQVIFEVLEEYDWTSFVAVTTRAPGHRAFLSYIEVLTDGSLVGWEHRGALTLDPGAGEAVLSAQLRSVSAQIRLLFCAREEAEPVFRAAEEAGLTGSGYVWFMVGPQLAGGGGSGAPGEPPLLPGGAPL\n\n### Answer:", "output": "Decreased glutamate and glycine agonist potency.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00562", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: COc1ccccc1CNCCCCCCNCCSSCCNCCCCCCNCc1ccccc1OC\nDescriptors: MW=590.94, LogP=6.26, TPSA=66.58, HBD=4, HBA=8", "output": "Descriptors: MW=590.94, LogP=6.26, TPSA=66.58, HBD=4, HBA=8, QED=0.07, Lipinski violations=2. Overall the molecule is somewhat outside the typical oral drug-like ranges, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "COc1ccccc1CNCCCCCCNCCSSCCNCCCCCCNCc1ccccc1OC", "compound_name": null, "pubchem_cid": null, "MW": 590.94, "LogP": 6.26, "TPSA": 66.58, "QED": 0.07, "functional_groups": ["aromatic ring", "amine", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00563", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRLNTIKPGEGSKKTAKRVGRGIGSGLGKTCGRGHKGQKSRSGGFHKVGFEGGQMPLQRRLPKRGFNSLTRARNYEVRLTDLDRLPVDEIDLLALQVAGIVPGDALSAKVILSGAISRKVVLKGVGATKGAKAAIEAVGGSIAE\n\n### Sequence 2:\nMRLNTIKPGEGSKKTAKRVGRGIGSGLGKTCGRGHKGQKSRSGGFHKVGFEGGQMPLQRRLPKRGFNSLTRARNYEVRLTDLDRLPVDEIDLLALQVAGIVPGDALSAKVILSGAISRKVVLKGVGATKGAKAAIEAVGGSIAE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00564", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPUP_BIFA0\n\n### Answer:", "output": "Belongs to the prokaryotic ubiquitin-like protein family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00565", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "ACTA2, PDGFRB, COL1A1", "output": "The expression pattern is consistent with cancer-associated fibroblast in a context associated with COVID-19, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "intestine", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["ACTA2", "PDGFRB", "COL1A1"]}}
{"id": "eval_00566", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CCC[C@@H]1C[C@@H](C(=O)NC(C(C)Cl)[C@H]2O[C@H](SC)[C@H](O)[C@@H](O)[C@H]2O)N(C)C1", "output": "The molecule contains the following functional groups: amide, amine, hydroxyl, ether, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BBBP", "canonical_smiles": "CCC[C@@H]1C[C@@H](C(=O)NC(C(C)Cl)[C@H]2O[C@H](SC)[C@H](O)[C@@H](O)[C@H]2O)N(C)C1", "compound_name": null, "pubchem_cid": null, "MW": 424.99, "LogP": 0.39, "TPSA": 102.26, "QED": 0.44, "functional_groups": ["amide", "amine", "hydroxyl", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amide", "amine", "hydroxyl", "ether", "halide"]}}
{"id": "eval_00567", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nBH006_ORYSJ\n\n### Answer:", "output": "Transcription factor BHLH6, Basic helix-loop-helix protein 6, OsbHLH6, osbHLH006, bHLH transcription factor 6.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00568", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "COc1cc2c(ccc(=O)n2CCN2CCC(NCc3cc4c(cn3)OCCO4)CC2)cn1", "output": "The SMILES represents an unnamed molecule with MW=451.53, LogP=1.83, TPSA=90.74. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Lipophilicity", "canonical_smiles": "COc1cc2c(ccc(=O)n2CCN2CCC(NCc3cc4c(cn3)OCCO4)CC2)cn1", "compound_name": null, "pubchem_cid": null, "MW": 451.53, "LogP": 1.83, "TPSA": 90.74, "QED": 0.58, "functional_groups": ["amine", "ether", "pyridine"], "moleculenet_labels": {"label": 0.68}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00569", "category": "Mutation", "instruction": "Describe the effect of mutation Q99856-Y530F on protein AT-rich interactive domain-containing protein 3A.", "input": "### Wild-type:\nMKLQAVMETLLQRQQRARQELEARQQLPPDPPAAPPGRARAAPDEDREPESARMQRAQMAALAAMRAAAAGLGHPASPGGSEDGPPGSEEEDAAREGTPGSPGRGREGPGEEHFEDMASDEDMKPKWEEEEMEEDLGEDEEEEEEDYEDEEEEEDEEGLGPPGPASLGTTALFPRKAQPPQAFRGDGVPRVLGGQERPGPGPAHPGGAAHVAPQLQPPDHGDWTYEEQFKQLYELDGDPKRKEFLDDLFSFMQKRGTPVNRIPIMAKQVLDLFMLYVLVTEKGGLVEVINKKLWREITKG\n\n### Mutant:\nMKLQAVMETLLQRQQRARQELEARQQLPPDPPAAPPGRARAAPDEDREPESARMQRAQMAALAAMRAAAAGLGHPASPGGSEDGPPGSEEEDAAREGTPGSPGRGREGPGEEHFEDMASDEDMKPKWEEEEMEEDLGEDEEEEEEDYEDEEEEEDEEGLGPPGPASLGTTALFPRKAQPPQAFRGDGVPRVLGGQERPGPGPAHPGGAAHVAPQLQPPDHGDWTYEEQFKQLYELDGDPKRKEFLDDLFSFMQKRGTPVNRIPIMAKQVLDLFMLYVLVTEKGGLVEVINKKLWREITKG\n\n### Answer:", "output": "No effect on DNA-binding. The mutation in the Bright/Dril1/ARID3a protein causes a functional change. However, the specific functional information related to this mutation is not mentioned in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00570", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMENTDVDQPHSMGTTIIGVTYNGGVVLGADSRTSTGMYVANRASDKITQLTDNVYVCRSGSAADSQIVSDYVRYFLHQHTIQLGQPATVKVAANLTRLLSYNNKDRLQTGMIIGGWDKYEGGKIYGIPPGGTVLEQPFAIGGSGSSYLYGFFDQAWKEGMTQEEAEKLVVTAVSLAIARDGASGGVVRTVTINKDGATRKFYSGDSLQLWHEELEPVNSLLDVVFASSPVPMVS\n\n### Sequence 2:\nMENTDVDQPHSMGTTIIGVTYNGGVVLGADSRTSTGMYVANRASDKITQLTDNVYVCRSGSAADSQIVSDYVRYFLHQHTIQLGQPATVKVAANLTRLLSYNNKDRLQTGMIIGGWDKYEGGKIYGIPPGGTVLEQPFAIGGSGSSYLYGFFDQAWKEGMTQEEAEKLVVTAVSLAIARDGASGGVVRTVTINKDGATRKFYSGDSLQLWHEELEPVNSLLDVVFASSPVPMVS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00571", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UM01-R334L on protein Y+L amino acid transporter 1.", "input": "### Wild-type:\nMVDSTEYEVASQPEVETSPLGDGASPGPEQVKLKKEISLLNGVCLIVGNMIGSGIFVSPKGVLIYSASFGLSLVIWAVGGLFSVFGALCYAELGTTIKKSGASYAYILEAFGGFLAFIRLWTSLLIIEPTSQAIIAITFANYMVQPLFPSCFAPYAASRLLAAACICLLTFINCAYVKWGTLVQDIFTYAKVLALIAVIVAGIVRLGQGASTHFENSFEGSSFAVGDIALALYSALFSYSGWDTLNYVTEEIKNPERNLPLSIGISMPIVTIIYILTNVAYYTVLDMRDILASDAVAVTF\n\n### Mutant:\nMVDSTEYEVASQPEVETSPLGDGASPGPEQVKLKKEISLLNGVCLIVGNMIGSGIFVSPKGVLIYSASFGLSLVIWAVGGLFSVFGALCYAELGTTIKKSGASYAYILEAFGGFLAFIRLWTSLLIIEPTSQAIIAITFANYMVQPLFPSCFAPYAASRLLAAACICLLTFINCAYVKWGTLVQDIFTYAKVLALIAVIVAGIVRLGQGASTHFENSFEGSSFAVGDIALALYSALFSYSGWDTLNYVTEEIKNPERNLPLSIGISMPIVTIIYILTNVAYYTVLDMRDILASDAVAVTF\n\n### Answer:", "output": "Highly induces sodium-independent cationic amino acid and sodium-dependent neutral amino acid transport activity. The missense mutation enables y+LAT-1 amino-acid transport activity when co-expressed with the heavy chain of the cell-surface antigen 4F2 (4F2hc, also known as CD98) in Xenopus laevis oocytes. The missense mutant failed to reach the oocyte plasma membrane when co-expressed with 4F2hc, demonstrating that it is a transport-activating mutation. This finding indicates that residue L334 plays a negligible role in the function of the y(+)LAT-1 transporter. The mutation in the SLC7A7 gene alleviates Kinuric protein intolerance (LPI), an inherited aminoaciduria. This mutation leads to enhanced cationic amino acid (CAA) transport at the basolateral membrane of epithelial cells in the intestine and kidney.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00572", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nGAD2 GAD1 SLC32A1\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as SLC32A1, GAD1, GAD2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["GAD2", "GAD1", "SLC32A1"]}}
{"id": "eval_00573", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nCNN1:medium TAGLN:low ACTA2:high MYH11:medium\n</cell>", "output": "This cell is most consistent with smooth muscle cell, a smooth muscle cell. Key markers such as CNN1, TAGLN, MYH11 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "smooth muscle cell", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CNN1", "TAGLN", "ACTA2", "MYH11"]}}
{"id": "eval_00574", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNIKPEEITSIIRQQIENFNTNIETIDSGTIIQIGDGIARVYGLEDCMEGELIEFPNDVYGMALNLEQDNVGCVLLGSEEGIKEGNVVKRTKKVVEVPVGEALVGRVVNSLGMPIDGKGPVLTTETRDVEVPAPGVIDRQSVKEPLQTGIKAIDSMIPIGKGQRELIIGDRQTGKTAIAMDTILNQKGKDVICIYVAIGQKQSTVAHIVNDLTKMGAMDYTIVVSSTASDSAPLQYLAPYAGCSMGEYFMHKGKDVLIVYDDLSKHAVAYRTMSLLLRRPPGREAYPGDVFYLHSRLLERSARLSEKLGGGSLTALPIVETLAGDVTAYIPTNVISITDGQIFLESELFNAGQRPAVNAGISVSRVGGNAQIKAMKQVAGTLRLELAQYRELAAFSQFGSDLDKESVKRLEKGKRLVEILKQPQYSPMPVEKEIIILYAAVSNHLIDIPVNKIKEFEKELFNYIDTHYRDIGKDILEHKQLTDELKSKLDKAINDFKNVFLSEI\n\n### Sequence 2:\nMTQDYETVLKGKYPAKEHALRVADYVKSKVPDATGILYVEGRMTKMLEDNDEPEPFRQRRYFYYLTGCPLADCHYIFDLATSKSTLFIPPIDPDSVIWSGLPVSAAEAKELYDVDEVKYTTDVNAELARLGKGPKKTVFAIQNQVLDSITFLEFDEKNFSILKDAIERCRVVKDDYEIALTRKANAVSTVAHHAVVEYVKKAKNERELEALFLQRSVANGAKNQAYHGIFAGGRAAATLHYVANDAPLEGKLNLLLDAGTEWNCYASDITRTFPISGKFSKESRQIYDIVLKMQLETTAALKEGVIWDEIHLLAHKIAIDGLHLIGILKGDKDEILKNRTSVAFFPHGLGHYLGMDTHDVGGNANYADRDPMFRYLRVRGALPAGSIVTVEPGIYFCSFIIEPYLKDPVHSKFIDSAVLEKYWDVGGVRIEDNILITKDGSENLTPTIKDPDELEKIIQAS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00575", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: O=C(NCC1CCCCN1)c1cc(OCC(F)(F)F)ccc1OCC(F)(F)F\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "O=C(NCC1CCCCN1)c1cc(OCC(F)(F)F)ccc1OCC(F)(F)F", "compound_name": null, "pubchem_cid": null, "MW": 414.35, "LogP": 3.44, "TPSA": 59.59, "QED": 0.67, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_00576", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nMS4A1 IGHM CD79A TCL1A CD79B\n</cell>", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as TCL1A, CD79B, CD79A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "heart", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MS4A1", "IGHM", "CD79A", "TCL1A", "CD79B"]}}
{"id": "eval_00577", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nLYZ DEFA5 DEFA6\n</cell>", "output": "This cell is most consistent with paneth cell, a paneth cell. Key markers such as LYZ, DEFA5, DEFA6 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "paneth cell", "tissue": "prostate", "disease": "normal", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["LYZ", "DEFA5", "DEFA6"]}}
{"id": "eval_00578", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nNIVLTQSPASLAVSLGQRATISCRASESVDSYGNSFMHWYQQKPGQPPKLLIYLASNLESGVPARFSGSGSRTDFTLTIDPVEADDAATYYCQQNNEDPLTFGAGTKLELK\n\n### Sequence 2:\nNIVLTQSPASLAVSLGQRATISCRASESVDSYGNSFMHWYQQKPGQPPKLLIYLASNLESGVPARFSGSGSRTDFTLTIDPVEADDAATYYCQQNNEDPLTFGAGTKLELK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00579", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRHYEIVFMVHPDQSEQVPGMIERYTAAITGAEGKIHRLEDWGRRQLAYPINKLHKAHYVLMNVEAPQEVIDELETTFRFNDAVIRSMVMRTKHAVTEASPMVKAKDERRERRDDFANETADDAEAGDSEE\n\n### Sequence 2:\nMRAVIQRAKSGSVTVDSKIVSQISHGLVVLIGVGHEDTAEDVEKVANKIIKTKLWPSVDGAQQWKQSVLDVGGEVLCVSQFTLFAKVKKGQKPDFHNAAKGPQAKELYDQVLAKIQAALPEGRTVKDGVFGAMMDVALVNDGPVTIQYDTKNDK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00580", "category": "Mutation", "instruction": "Describe the effect of mutation Q3T906-D1018G on protein N-acetylglucosamine-1-phosphotransferase subunits alpha/beta.", "input": "### Wild-type:\nMLFKLLQRQTYTCLSHRYGLYVCFLGVVVTIVSAFQFGEVVLEWSRDQYHVLFDSYRDNIAGKSFQNRLCLPMPIDVVYTWVNGTDLELLKELQQVREQMEEEQKAMREILGKNTTEPTKKSEKQLECLLTHCIKVPMLVLDPALPANITLKDLPSLYPSFHSASDIFNVAKPKNPSTNVSVVVFDSTKDVEDAHSGLLKGNSRQTVWRGYLTTDKEVPGLVLMQDLAFLSGFPPTFKETNQLKTKLPENLSSKVKLLQLYSEASVALLKLNNPKDFQELNKQTKKNMTIDGKELTISPA\n\n### Mutant:\nMLFKLLQRQTYTCLSHRYGLYVCFLGVVVTIVSAFQFGEVVLEWSRDQYHVLFDSYRDNIAGKSFQNRLCLPMPIDVVYTWVNGTDLELLKELQQVREQMEEEQKAMREILGKNTTEPTKKSEKQLECLLTHCIKVPMLVLDPALPANITLKDLPSLYPSFHSASDIFNVAKPKNPSTNVSVVVFDSTKDVEDAHSGLLKGNSRQTVWRGYLTTDKEVPGLVLMQDLAFLSGFPPTFKETNQLKTKLPENLSSKVKLLQLYSEASVALLKLNNPKDFQELNKQTKKNMTIDGKELTISPA\n\n### Answer:", "output": "In MLIIIA; patients with intermediate phenotype between MLII and MLIIIA; unknown pathological significance; no effect on protein abundance; decreased localization to the Golgi; loss of UDP-N-acetylglucosamine-lysosomal-enzyme N-acetylglucosaminephosphotransferase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00581", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMED11_ARATH\n\n### Answer:", "output": "Component of the Mediator complex, a coactivator involved in the regulated transcription of nearly all RNA polymerase II-dependent genes.  Mediator functions as a bridge to convey information from gene-specific regulatory proteins to the basal RNA polymerase II transcription machinery.  The Mediator complex, having a compact conformation in its free form, is recruited to promoters by direct interactions with regulatory proteins and serves for the assembly of a functional preinitiation complex with RNA polymerase II and the general transcription factors.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00582", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSDTLTRLAAVLEERRNAAPDSSYVASLYHKGLNKILEKVGEESVETIIAAKDAAASGDCQELIYETADLWFHSLVMLSALGQHPQAVLDELERRFGLSGHAEKAARQPSA\n\n### Sequence 2:\nMKTFVPKNNEQNWVVVDAAGVPLGRLATLVASRIRGKHRPDFTPNIIQGDFVVVVNAEKVVLTGNKLDGKVYTRYTGYQGGLKTETARQALAKHPERVIEHAVFGMLPKGRQGRALHSRLKVYAGTTHPHAAQKPQQLEVRTALEVK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00583", "category": "Mutation", "instruction": "Describe the effect of mutation P62314-I60R on protein Small nuclear ribonucleoprotein Sm D1.", "input": "### Wild-type:\nMKLVRFLMKLSHETVTIELKNGTQVHGTITGVDVSMNTHLKAVKMTLKNREPVQLETLSIRGNNIRYFILPDSLPLDTLLVDVEPKVKSKKREAVAGRGRGRGRGRGRGRGRGRGGPRR\n\n### Mutant:\nMKLVRFLMKLSHETVTIELKNGTQVHGTITGVDVSMNTHLKAVKMTLKNREPVQLETLSRRGNNIRYFILPDSLPLDTLLVDVEPKVKSKKREAVAGRGRGRGRGRGRGRGRGRGGPRR\n\n### Answer:", "output": "Loss of interaction with CLNS1A. The mutation in the Sm proteins may cause functional changes in the assembly of the Sm core of spliceosomal snRNPs.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00584", "category": "Mutation", "instruction": "Describe the effect of mutation Q92974-A394Y on protein Rho guanine nucleotide exchange factor 2.", "input": "### Wild-type:\nMSRIESLTRARIDRSRELASKTREKEKMKEAKDARYTNGHLFTTISVSGMTMCYACNKSITAKEALICPTCNVTIHNRCKDTLANCTKVKQKQQKAALLKNNTALQSVSLRSKTTIRERPSSAIYPSDSFRQSLLGSRRGRSSLSLAKSVSTTNIAGHFNDESPLGLRRILSQSTDSLNMRNRTLSVESLIDEAEVIYSELMSDFEMDEKDFAADSWSLAVDSSFLQQHKKEVMKQQDVIYELIQTELHHVRTLKIMTRLFRTGMLEELHLEPGVVQGLFPCVDELSDIHTRFLSQLLER\n\n### Mutant:\nMSRIESLTRARIDRSRELASKTREKEKMKEAKDARYTNGHLFTTISVSGMTMCYACNKSITAKEALICPTCNVTIHNRCKDTLANCTKVKQKQQKAALLKNNTALQSVSLRSKTTIRERPSSAIYPSDSFRQSLLGSRRGRSSLSLAKSVSTTNIAGHFNDESPLGLRRILSQSTDSLNMRNRTLSVESLIDEAEVIYSELMSDFEMDEKDFAADSWSLAVDSSFLQQHKKEVMKQQDVIYELIQTELHHVRTLKIMTRLFRTGMLEELHLEPGVVQGLFPCVDELSDIHTRFLSQLLER\n\n### Answer:", "output": "Increases phosphorylation level, abnormal microtubule localization and activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00585", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMEKNNNKKRATVRIQGQTYKVVSSEEPAHVKEVASYMNKKMEELKKRNPYLDSTKLAVLTALNIADEYLKLKRQYEGE\n\n### Sequence 2:\nKTRVTVDIYGQQYTIVGTESSSHIRLVASIVDDKMREISEKNPTLDISKLAVLTAINIVHDYIKLKEEYD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00586", "category": "Mutation", "instruction": "Describe the effect of mutation O53666-R301A on protein Broad-specificity linear acyl-CoA dehydrogenase FadE5.", "input": "### Wild-type:\nMSHYRSNVRDQVFNLFEVLGVDKALGHGEFSDVDVDTARDMLAEVSRLAEGPVAESFVEGDRNPPVFDPKTHSVMLPESFKKSVNAMLEAGWDKVGIDEALGGMPMPKAVVWALHEHILGANPAVWMYAGGAGFAQILYHLGTEEQKKWAVLAAERGWGSTMVLTEPDAGSDVGAARTKAVQQADGSWHIDGVKRFITSGDSGDLFENIFHLVLARPEGAGPGTKGLSLYFVPKFLFDVETGEPGERNGVFVTNVEHKMGLKVSATCELAFGQHGVPAKGWLVGEVHNGIAQMFEVIEQA\n\n### Mutant:\nMSHYRSNVRDQVFNLFEVLGVDKALGHGEFSDVDVDTARDMLAEVSRLAEGPVAESFVEGDRNPPVFDPKTHSVMLPESFKKSVNAMLEAGWDKVGIDEALGGMPMPKAVVWALHEHILGANPAVWMYAGGAGFAQILYHLGTEEQKKWAVLAAERGWGSTMVLTEPDAGSDVGAARTKAVQQADGSWHIDGVKRFITSGDSGDLFENIFHLVLARPEGAGPGTKGLSLYFVPKFLFDVETGEPGERNGVFVTNVEHKMGLKVSATCELAFGQHGVPAKGWLVGEVHNGIAQMFEVIEQA\n\n### Answer:", "output": "Increases affinity for eicosanoyl-CoA; when associated with A-447.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00587", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "FCGR3B, CSF3R, CXCR2, S100A9, S100A8", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as FCGR3B, S100A9, S100A8 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "pancreas", "disease": "normal", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["FCGR3B", "CSF3R", "CXCR2", "S100A9", "S100A8"]}}
{"id": "eval_00588", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVKVKSKNSVIKLLSTAASGYSRYISIKKGAPLVTQVRYDPVVKRHVLFKEAKKRKVAERKPLDFLRTAK\n\n### Sequence 2:\nMVKVKSKNSVIKLLSTAASGYSRYISIKKGAPLVTQVRYDPVVKRHVLFKEAKKRKVAERKPLDFLRTAK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00589", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDEGLY_PYRFU\n\n### Answer:", "output": "Belongs to the peptidase C56 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00590", "category": "Mutation", "instruction": "Describe the effect of mutation Q2LKW6-L45Q on protein NACHT, LRR and PYD domains-containing protein 1b allele 1.", "input": "### Wild-type:\nMEESPPKQKSNTKVAQHEGQQDLNTTRHMNVELKHRPKLERHLKLGMIPVVYMKQGEEILYPAQSLREENLIQNFTSLLLLQKLCPKDPENMIRKSWASCVPEEGGHMINIQDLFGPNIGTQKEPQLVIIEGAAGIGKSTLARLVKRAWKEGQLYRDHFQHVFFFSCRELAQCKKLSLAELIAQGQEVPTAPINQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQRQPVHTLLGSLLGKSILPEAFFLLTTRTTALQKFIPSLPMPCQVEVLGFSGIERENYFYKYFANQRHAIT\n\n### Mutant:\nMEESPPKQKSNTKVAQHEGQQDLNTTRHMNVELKHRPKLERHLKQGMIPVVYMKQGEEILYPAQSLREENLIQNFTSLLLLQKLCPKDPENMIRKSWASCVPEEGGHMINIQDLFGPNIGTQKEPQLVIIEGAAGIGKSTLARLVKRAWKEGQLYRDHFQHVFFFSCRELAQCKKLSLAELIAQGQEVPTAPINQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQRQPVHTLLGSLLGKSILPEAFFLLTTRTTALQKFIPSLPMPCQVEVLGFSGIERENYFYKYFANQRHAIT\n\n### Answer:", "output": "Loss of cleavage by LT and of LT-induced IL1B processing. Loss of cleavage by LT and of LT-induced IL1B processing; when associated with Q-43 and A-44. No effect on IL1B release in response to metabolic inhibitors; when associated with Q-43 and A-44. The mutation in NLRP1B has been shown to induce inflammasome activation in the absence of the lethal factor (LF) protease.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00591", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nXRE_PSEPK\n\n### Answer:", "output": "Probable antitoxin component of a type II toxin-antitoxin (TA) system.  In vivo probably neutralizes the toxic effect of cognate toxin Res.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00592", "category": "Mutation", "instruction": "Describe the effect of mutation P0AG11-G92D on protein Protein UmuD.", "input": "### Wild-type:\nMLFIKPADLREIVTFPLFSDLVQCGFPSPAADYVEQRIDLNQLLIQHPSATYFVKASGDSMIDGGISDGDLLIVDSAITASHGDIVIAAVDGEFTVKKLQLRPTVQLIPMNSAYSPITISSEDTLDVFGVVIHVVKAMR\n\n### Mutant:\nMLFIKPADLREIVTFPLFSDLVQCGFPSPAADYVEQRIDLNQLLIQHPSATYFVKASGDSMIDGGISDGDLLIVDSAITASHGDIVIAAVDDEFTVKKLQLRPTVQLIPMNSAYSPITISSEDTLDVFGVVIHVVKAMR\n\n### Answer:", "output": "In umuD77; non-cleavable. The umuD1 mutation resulted in a non-cleavable UmuD protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00593", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nAMP2_GEOSE\n\n### Answer:", "output": "Belongs to the peptidase M29 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00594", "category": "Mutation", "instruction": "Describe the effect of mutation P0AB87-A131F on protein L-fuculose phosphate aldolase.", "input": "### Wild-type:\nMERNKLARQIIDTCLEMTRLGLNQGTAGNVSVRYQDGMLITPTGIPYEKLTESHIVFIDGNGKHEEGKLPSSEWRFHMAAYQSRPDANAVVHNHAVHCTAVSILNRSIPAIHYMIAAAGGNSIPCAPYATAGTRELSEHVALALKNRKATLLQHHGLIACEVNLEKALWLAHEVEVLAQLYLTTLAITDPVPVLSDEEIAVVLEKFKTYGLRIEE\n\n### Mutant:\nMERNKLARQIIDTCLEMTRLGLNQGTAGNVSVRYQDGMLITPTGIPYEKLTESHIVFIDGNGKHEEGKLPSSEWRFHMAAYQSRPDANAVVHNHAVHCTAVSILNRSIPAIHYMIAAAGGNSIPCAPYATFGTRELSEHVALALKNRKATLLQHHGLIACEVNLEKALWLAHEVEVLAQLYLTTLAITDPVPVLSDEEIAVVLEKFKTYGLRIEE\n\n### Answer:", "output": "Has a slight preference for the L-aldehyde and shows a retention of the diastereoselectivity. Gain of aldolase activity; when associated with W-206.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00595", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nCOL1A2:medium LUM:high DCN:low COL1A1:high PDGFRA:medium\n</cell>", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as DCN, PDGFRA, COL1A2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "spleen", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["COL1A2", "LUM", "DCN", "COL1A1", "PDGFRA"]}}
{"id": "eval_00596", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nCSF1R TMEM119 P2RY12 CX3CR1\n</cell>", "output": "This cell is most consistent with microglia, a brain microglia cell. Key markers such as P2RY12, CX3CR1, CSF1R support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "brain", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CSF1R", "TMEM119", "P2RY12", "CX3CR1"]}}
{"id": "eval_00597", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGVRAQQKEKTRRSLVEAAFSQLSAERSFASLSLREVAREAGIAPTSFYRHFRDVDELGLTMVDESGLMLRQLMRQARQRIAKGGSVIRTSVSTFMEFIGNNPNAFRLLLRERSGTSAAFRAAVAREIQHFIAELADYLELENHMPRAFTEAQAEAMVTIVFSAGAEALDVGVEQRRQLEERLVLQLRMISKGAYYWYRREQEKTTIIPGNVKDE\n\n### Sequence 2:\nMGVRAQQKEKTRRSLVEAAFSQLSAERSFASLSLREVAREAGIAPTSFYRHFRDVDELGLTMVDESGLMLRQLMRQARQRIAKGGSVIRTSVSTFMEFIGNNPNAFRLLLRERSGTSAAFRAAVAREIQHFIAELADYLELENHMPRAFTEAQAEAMVTIVFSAGAEALDVGVEQRRQLEERLVLQLRMISKGAYYWYRREQEKTTIIPGNVKDE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00598", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSGGSGMQIFVDTVQWKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKESALILLLTLR\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCEEEEEEEEEEEEEEECCTTCBHHHHHHHHHHHHCCCGGGEEEEETTEECCTTSBTGGGTCCTTCEEEEEEBCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00599", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSMKRLKDLREYLAVLEAHQDVREIDEPVDPHLEAGAAARWTYENRGPALMLNDLTGTGRFCRILAAPAGLSTIPGSPLARVALSLGLDVSATAHEIVDSLAAARTREPVAPVVVDSAPCQDNVLLGDDANLDRFPAPLLHEGDGGPYLNTWGTIIVSTPDGSFTNWAIARVMKIDGKRMTGTFIPTQHLGQIRKLWDNLGQPMPFAIVQGTEPGIPFVASMPLPDGIEEVGFLGAYFGEPLELVRAKTVDLLVPASAEIVIEGHVMPPEYVVDAITYRDDPIWPISVAGEPVDETHTAWGLVTAAEALALLRAAKLPVATAWMPFEAAAHWLIVCLTEDWRERMPGLSRDGICLRISQVLAATRIEAMMTRVFVLDDDVDPSDQTELAWAIATRVSPAHGRLVRHGMINPLAGCYSAEERRLGYGPKAVLNGLLPPMAERSRRSSFRHTYPEPVRQRVIELLA\n\n### Answer:", "output": "<SEQ_3Di>DFQFQFFLVSLVVVLVVQVQEDEDQPEAELAARVLVVLVVCQVVVGGWYWDQAYPQQGRQEIETHQQQHQTPRPQQRRQSVQRLLRHDSRDDLFNSLVLLLVLVVDWFDAAAEDAAFLLFPFKQFDPRQWCSNDRQHPNDPPDPAGWQFAQKWKWFAAPVNPAIATAGFTWGDPGRFKIDGDLDCLDPSVVSLVRCVVVQHWTKMKIFFRGRSNRSQLNLAPDDDVDHSQRSSCSSVVHHFYWYADPPGGTTHGSNTQKMFIATQHVPMTGTGMITGGGSGYGYHWNDDAPDTCLFSRNLSSHQSVLVNQCVVVVQQWPGKHQARNQSSQEIETAGELCRCVSDPPADLLRVLVVSLVSLVVDPPNQSNLEYEYEHVVADSNPVVSVVVLCVPQFDPPDQKDKDKDAHDLPRDPDDPVCSVVRIDIHIYTRRYDDHPVPDDFDDDLVTPDDVVVSVVVVVVPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00600", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGIDPFTSTQQHTEPAEEETLHNIITDTENVQGSFSKHEFQAETKKLLDIVARSLYSEKEVFIRELISNGSDALEKLRHRMITAGGDTAPMEIHLQTDSVKGTFTIQDTGVGMNKEDLVSNLGTIARSGSKAFLDALQNQAEASSSIIGQFGVGFYSAFMVADKVEVYSQSAEADAPGYKWSSDGSGVFEVAEASGVRQGTKIVLHLKDDCKEFSSEDRVKEVVTKYSNFVSFPIFLNGRRLNTLQALWMMEPKDISEWQHEEFYRYVAQAYDKPRYTLHYRADAPLNIRSIFYVPEMKPSMFDVSREMGSSVALYSRKILIQTKATDILPKWLRFLRGVVDSEDIPLNLSRELLQESALIRKLRDVLQQRVIRFLLDQSKKDPEKYARFFEDYGLFMREGIVTTGEQSVKEDIAKLLRFESSALPAGQQTSLMEYSSRMKAGTRNIYYLCAPNRHLAEHSPYFEAMKQKDMEVLFCFEQFDELTLLHLREFDRKKLISAETDIVVDHYKEEKFQDSKPASERLSSEQAEDLLAWMRNALVQRVTNIKVTPRLDTHPAMITVLEMGAARHFLRTQQLARSSEERAQILQPTLEINTGHDLIKKLHALKDSNPELAQLLLEQIYDNAMIAAGLNEDPRPMISRLNQLLTRALEKH\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCSCCEEEEEEEEEECEECHHHHHHHHHHHTCSSTTHHHHHHHHHHHHHHHHHHTTSCCCCCCCCCCCEEEEEETTTTEEEEEECSSCCCHHHHHHTTSEETCCHHHHTCCCCCCCCCCCTTCCCCSSCGGGGGGGTEEEEEEEEECSSSSCCEEEEEECSSSEEEEEEEESCCSEEEEEEEECGGGGGGGCHHHHHHHHHHHSSSCSSCEEETTEECCCCCCGGGSCGGGCCHHHHHHHHHHHSCCSSCEEEEECCEECSSSCEEEEEEEESSCCCHHHHHCCCCCCEEEEETTEEECCCCCCCSCGGGTTCEEEEECTTSCBCSSSCCBSCHHHHHHHHHHHHHHHHHHHHHHHHHCHHHHHHHHHHHHHHHHHHHHTCCCHHHHHHHHTTCCEEETTSCTTCEECHHHHHHHSCSSCCEEEEEECSSHHHHHTCHHHHHHHTSCCCEEEECSTTHHHHHHHHCEETTEEEEEHHHHHHHHHTSCCCCCCSSCGGGSCCHHHHHHHHHHHHHHSTTTCSEEEECSCCSSCSEEECCTTHHHHHHHTCCCCCCCCCCCCCCCCCCEEEECTTSHHHHHHHHHHHHCHHHHHHHHHHHHHHHHHTTTCCCCSSHHHHHHHHHHHHHTCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00601", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSKNQLNLQDAFLNQVRKENVGVTIFLINGFQLKGFVKGFDNFTVILESEGKQHMIYKHAISTIIPQRPVNTYLAKGGNEENTPS\n\n### Sequence 2:\nKNNINLQDVFLNQVRKENIGITIFLVNGFQLKGFVRGFDNYTIVLDSDGKQQMIYKHAVSTISPNSPVN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00602", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGLPK_COREF\n\n### Answer:", "output": "Key enzyme in the regulation of glycerol uptake and metabolism.  Catalyzes the phosphorylation of glycerol to yield sn-glycerol 3-phosphate.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00603", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCPA3 HPGDS TPSB2 KIT\n</cell>", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as CPA3, HPGDS, KIT support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "bone marrow", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CPA3", "HPGDS", "TPSB2", "KIT"]}}
{"id": "eval_00604", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nNHAA_CAMC5\n\n### Answer:", "output": "Na(+)/H(+) antiporter that extrudes sodium in exchange for external protons.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00605", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRPC5_MOUSE\n\n### Answer:", "output": "DNA-dependent RNA polymerase catalyzes the transcription of DNA into RNA using the four ribonucleoside triphosphates as substrates.  Specific periphjeric component of RNA polymerase III which synthesizes small RNAs, such as 5S rRNA and tRNAs.  Essential for efficient transcription from both the type 2 VAI and type 3 U6 RNA polymerase III promoters.  Plays a key role in sensing and limiting infection by intracellular bacteria and DNA viruses.  Acts as nuclear and cytosolic DNA sensor involved in innate immune response.  Can sense non-self dsDNA that serves as template for transcription into dsRNA.  The non-self RNA polymerase III transcripts induce type I interferon and NF- Kappa-B through the RIG-I pathway (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00606", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGLEVLFQGPHMESLQQQVAQLLEQQPTLLPAAMAEQLNVTEFDIVHALPEEMVAVVDGSHAQTILESLPEWGPVTTIMTIAGSIFEVKAPFPKGKVARGYYNLMGRDGELHGHLKLENISHVALVSKPFMGRESHYFGFFTAQGENAFKIYLGRDEKRELIPEQVARFKAMQQQHKQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHCTTCCHHHHHHHHTSCHHHHHHHSCGGGEEEEEGGGHHHHHHHHHHHCEEEEEEEETTEEEEEEEECCEEEEETTEEEEECCTTSCEEEECGGGCCEEEEECCCBTTBCCCEEEEECTTCCEEEEEEECBCTTSCBCHHHHHHHHHHHHHHTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00607", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD8A PDCD1 TOX LAG3\n</cell>", "output": "This cell is most consistent with exhausted CD8 T cell, an exhausted CD8+ T cell. Key markers such as LAG3, TOX, PDCD1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "spleen", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD8A", "PDCD1", "TOX", "LAG3"]}}
{"id": "eval_00608", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFEN1_ZYGRC\n\n### Answer:", "output": "Flap endonuclease 1, FEN-1, Flap structure-specific endonuclease 1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00609", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMTLD_ECOSE\n\n### Answer:", "output": "Mannitol-1-phosphate 5-dehydrogenase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00610", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDIQTILEKTLPGLGYELVDFELAAQGTLRVFIDKEGGITVEDCATVSNHLSRVFMVEDIGYKNLEISSPGLDRPLKKAADFVRFAGQNAKIKTRLPIGGQKNFIGKIEGCENDTVTVSFDGKTVQIELGNIDKARLRPEFKF\n\n### Sequence 2:\nMAAAVAMETDDAGNRLRFQLELEFVQCLANPNYLNFLAQRGYFKDKAFVNYLKYLLYWKEPEYAKYLKYPQCLHMLELLQYEHFRKELVNAQCAKFIDEQQILHWQHYSRKRMRLQQALAEQQQQNNASGK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00611", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRS14_YERPN\n\n### Answer:", "output": "30S ribosomal protein S14.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00612", "category": "Mutation", "instruction": "Describe the effect of mutation Q62720-H43A on protein Proton-coupled zinc antiporter SLC30A1.", "input": "### Wild-type:\nMGCWGRNRGRLLCMLLLTFMFMVLEVVVSRVTASLAMLSDSFHMLSDVLALVVALVAERFARRTHATQKNTFGWIRAEVMGALVNAIFLTGLCFAILLEAVERFIEPHEMQQPLVVLSVGVAGLLVNVLGLCLFHHHSGEGQGAGHGHSHGHGHGHLAKGARKAGRAGGEAGAPPGRAPDQEPDQEETNTLVANTSNSNGLKADQAEPEKLRSDDPVDVQVNGNLIQESDSLESEDNRAGQLNMRGVFLHVLGDALGSVIVVVNALVFYFSWKGCTEDDFCVNPCFPDPCKSSVELMNST\n\n### Mutant:\nMGCWGRNRGRLLCMLLLTFMFMVLEVVVSRVTASLAMLSDSFAMLSDVLALVVALVAERFARRTHATQKNTFGWIRAEVMGALVNAIFLTGLCFAILLEAVERFIEPHEMQQPLVVLSVGVAGLLVNVLGLCLFHHHSGEGQGAGHGHSHGHGHGHLAKGARKAGRAGGEAGAPPGRAPDQEPDQEETNTLVANTSNSNGLKADQAEPEKLRSDDPVDVQVNGNLIQESDSLESEDNRAGQLNMRGVFLHVLGDALGSVIVVVNALVFYFSWKGCTEDDFCVNPCFPDPCKSSVELMNST\n\n### Answer:", "output": "Loss of zinc export across plasma membrane and no effect on localization to the plasma membrane; when associated with A-254.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00613", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CCCOP(=S)(OCCC)SCC(=O)N1CCCCC1C", "output": "This molecule is medium-sized and relatively nonpolar, with lipophilic character. It contains amide. Its descriptor profile: MW=353.49, LogP=4.2, TPSA=38.77, HBD=0, HBA=5, QED=0.58.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "ESOL", "canonical_smiles": "CCCOP(=S)(OCCC)SCC(=O)N1CCCCC1C", "compound_name": null, "pubchem_cid": null, "MW": 353.49, "LogP": 4.2, "TPSA": 38.77, "QED": 0.58, "functional_groups": ["amide"], "moleculenet_labels": {"label": -4.15}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00614", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nAMP3_COCNU\n\n### Answer:", "output": "Secreted, extracellular space.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00615", "category": "Mutation", "instruction": "Describe the effect of mutation P59047-R143P on protein NACHT, LRR and PYD domains-containing protein 5.", "input": "### Wild-type:\nMKVAGGLELGAAALLSASPRALVTLSTGPTCSILPKNPLFPQNLSSQPCIKMEGDKSLTFSSYGLQWCLYELDKEEFQTFKELLKKKSSESTTCSIPQFEIENANVECLALLLHEYYGASLAWATSISIFENMNLRTLSEKARDDMKRHSPEDPEATMTDQGPSKEKVPGISQAVQQDSATAAETKEQEISQAMEQEGATAAETEEQEISQAMEQEGATAAETEEQGHGGDTWDYKSHVMTKFAEEEDVRRSFENTAADWPEMQTLAGAFDSDRWGFRPRTVVLHGKSGIGKSALARRIV\n\n### Mutant:\nMKVAGGLELGAAALLSASPRALVTLSTGPTCSILPKNPLFPQNLSSQPCIKMEGDKSLTFSSYGLQWCLYELDKEEFQTFKELLKKKSSESTTCSIPQFEIENANVECLALLLHEYYGASLAWATSISIFENMNLRTLSEKAPDDMKRHSPEDPEATMTDQGPSKEKVPGISQAVQQDSATAAETKEQEISQAMEQEGATAAETEEQEISQAMEQEGATAAETEEQGHGGDTWDYKSHVMTKFAEEEDVRRSFENTAADWPEMQTLAGAFDSDRWGFRPRTVVLHGKSGIGKSALARRIV\n\n### Answer:", "output": "Found in patients with female infertility; when associated in cis with C-462; unknown pathological significance.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00616", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "LAG3, TOX, CD8A, PDCD1", "output": "This cell is most consistent with exhausted CD8 T cell, an exhausted CD8+ T cell. Key markers such as CD8A, LAG3, TOX support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "intestine", "disease": "normal", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["LAG3", "TOX", "CD8A", "PDCD1"]}}
{"id": "eval_00617", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nVDSVYRTRSLGVAAEGIPDQYADGEAARVWQLYIGDTRSRTAEYKAWLLGLLRQHGCHRVLDVACGTGVDSIMLVEEGFSVTSVDASDKMLKYALKERWNRRKEPAFDKWVIEEANWLTLDKDVPAGDGFDAVICLGNSFAHLPDSKGDQSEHRLALKNIASMVRPGGLLVIDHRNYDYILSTGCAPPGKNIYYKSDLTKDITTSVLTVNNKAHMVTLDYTVQVPGAGRDGAPGFSKFRLSYYPHCLASFTELVQEAFGGRCQHSVLGDFKPYRPGQAYVPCYFIHVLKKTG\n\n### Answer:", "output": "<SEQ_2D>CCCEEESSCTTCCCTTSCCTTTTSHHHHHHHHHHHTTCSBCHHHHHHHHHHHHTTTCCEEEETTCTTSHHHHHHHHTTCEEEEEESCHHHHHHHHHHHHHTTTSHHHHTCEEEECCGGGHHHHSCCTTCEEEEEECTTCGGGSCCSSSSCHHHHHHHHHHHTTEEEEEEEEEEEECHHHHHHHSCCCTTCCSSBCCCSCCEEEEEEEEETTEEEEEEEEEEECCCCCCCCCCCCCEEEEEEECCCCHHHHHHHHHHTTTSCEECCEEETTEECCTTCSSCCSEEEEEEEECC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00618", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "COC(=O)c1ccccc1C(=O)OC", "output": "The molecule contains the following functional groups: aromatic ring, ester, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "COC(=O)c1ccccc1C(=O)OC", "compound_name": null, "pubchem_cid": null, "MW": 194.19, "LogP": 1.26, "TPSA": 52.6, "QED": 0.66, "functional_groups": ["aromatic ring", "ester", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "ester", "ether"]}}
{"id": "eval_00619", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "SLC1A3, AQP4, GFAP, S100B", "output": "This cell is most consistent with astrocyte, an astrocyte. Key markers such as GFAP, S100B, AQP4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "astrocyte", "tissue": "liver", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SLC1A3", "AQP4", "GFAP", "S100B"]}}
{"id": "eval_00620", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nTTR IRX2 GCG\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as IRX2, TTR, GCG support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "ovary", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TTR", "IRX2", "GCG"]}}
{"id": "eval_00621", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLIPYDQLQAETLTRLIEDFVTRDGTDNGDDTPLETRVLRVRQALAKGQAFILFDLESQQCQLLAKHDVPRELLE\n\n### Sequence 2:\nMLIPYEQLQAETLTRLIEDFVTRDGTDNGDDTPLETRVLRVRQALAKGQAFILFDPESQQCQLLAKHDVPRELLD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00622", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "EPCAM, KRT18, KRT8", "output": "This profile is most consistent with epithelial cell, which is commonly found in heart under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "epithelial cell", "tissue": "heart", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["EPCAM", "KRT18", "KRT8"]}}
{"id": "eval_00623", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "TCF7, CCR7, LTB, CD3E, IL7R, CD3D", "output": "This profile is most consistent with naive CD4+ T cell, which is commonly found in lung under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TCF7", "CCR7", "LTB", "CD3E", "IL7R", "CD3D"]}}
{"id": "eval_00624", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: COc1ccc(S(=O)(=O)OC)cc1C(=O)c1cc(S(=O)(=O)OC)ccc1OC\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "COc1ccc(S(=O)(=O)OC)cc1C(=O)c1cc(S(=O)(=O)OC)ccc1OC", "compound_name": null, "pubchem_cid": null, "MW": 430.46, "LogP": 1.6, "TPSA": 122.27, "QED": 0.45, "functional_groups": ["aromatic ring", "ether", "ketone"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_00625", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTSAC_ACTP7\n\n### Answer:", "output": "Belongs to the SUA5 family. TsaC subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00626", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGENLYFQGHMSLENVAFNVVNKGHFDGQQGEVPVSIINNTVYTKVDGVDVELFENKTTLPVNVAFELWAKRNIKPVPEVKILNNLGVDIAANTVIWDYKRDAPAHISTIGVCSMTDIAKKPTETICAPLTVFFDGRVDGQVDLFRNARNGVLITEGSVKGLQPSVGPKQASLNGVTLIGEAVKTQFNYYKKVDGVVQQLPETYFTQSRNLQEFKPRSQMEIDFLELAMDEFIERYKLEGYAFEHIVYGDFSHSQLGGLHLLIGLAKRFKESPFELEDFIPMDSTVKNYFITDAQTGSSKCVCSVIDLLLDDFVEIIKSQDLSVVSKVVKVTIDYTEISFMLWCKDGHVETFYPKLQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHSSCCCCCCCCCEEEETTEEEEEETTEEEEEEECCSSSCHHHHHHHHHTBCCSCBCCHHHHHHTTCCEEBSSBCEETTTTEESSSCEESSCTTTEEESSTTSGGGTTSCEEEETTSTTHHHHHHHCSSEEEEESSCCTTCCCEECCSCEEETTEEECCSSSCCCCEEEEEETTEECCCCCCBCBCCCCSTTCCCCSHHHHHHHHSCHHHHHHHTTCTTSCHHHHTTCBCSSSBCBCCCBHHHHHHHHHHSCEEEECSSCSCCSSEEEEEEETTTCCEEEEEEEBCBCHHHHHHHHHTSCCCSSEEEEEEEETTEEEEEEEEEETTEEEEEEECCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00627", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMKKIILTIGCPGSGKSTWAREFIAKNPGFYNINRDDYRQSIMAHEERDEYKYTKKKEGIVTGMQFDTAKSILYGGDSVKGVIISDTNLNPERRLAWETFAKEYGWKVEHKVFDVPWTELVKRNSKRGTKAVPIDVLRSMYKSMREYLGLPVYNGTPGKPKAVIFDVDGTLAKMDVINPMVVELSKMYALMGYQIVVVSGRESGTKEDPTKYYRMTRKWVEDIAGVPLVMQCQREQGDTRKDDVVKEEIFWKHIAPHFDVKLAIDDRTQVVEMWRRIGVECWQVASGDF\n\n### Answer:", "output": "<SEQ_3Di>DAEEEEEAAFPLLCLVVVQVVVCVVDFQEEEQELVVQQCVVVVHDLQDDGDDDPVSVVVSVVVRLVVQCVSVPDDPRHRYYYYRDNPLDPVSVVVVQVSCVVVPYYYDYHYRDDDPVVSVVRLVVVPRVRDDVVVRLVSVLSVCVVVVHAADDFDPPAFEAEEEEEDPQQQDVNAGPPVNLVVLVVVVVVRHAYEYEYADECDEPVRFRVPVVVVVCCCCPHSNRPHDYYYYDGYPDPDDSLVRVVVCCVPPNRNRHGYAEYEEAAVSNCSSQVSRHHHYDHPYHHDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00628", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAG_TOBAC\n\n### Answer:", "output": "Floral homeotic protein AGAMOUS, NAG1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00629", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKAEFDKELIKKYDRPGPRYTSYPPATEFTEEVKEDEYVKRLIKSNERKTPLSLYFHIPFCEQRCLYCGCNVIISHRKGIEEPYLERVCREMDLVSQYLDKDRKVIQLHWGGGTPNYLSPEQIKWFMEEIRKRFEFGDNAEISIELDPRYLTDEQIKAIKDAGFNRISLGVQDLDPKVQQAVNRVQPYELIKEKMEKLREAGFESINLDLIYGLPYQTKESFEKTVEKVIELNPDRIATYSFAYIPQVKPHQQLLPKEALPSAEEKLRIFEMVINKFQEAGYVYIGMDHFAKPEDELAVAQRKGELWRNFQGYTTKKGVELLGFGATSIGMLYDSYFQNWKTLRDYNKTVDEGKIPVFRGYVLNEDDFIRREVIMDIMCNLGVEFSKIENMFGINFREYFAKELEELKEMEEDGLIKVEEDRIKIMPVGRLLIRNVAMVFDAHLRRKKELNFSRTI\n\n### Sequence 2:\nMKAEFDKELIKKYDRPGPRYTSYPPATEFTEEVKEDEYVKRLIKSNERKTPLSLYFHIPFCEQRCLYCGCNVIISHRKGIEEPYLERVCREMDLVSQYLDKDRKVIQLHWGGGTPNYLSPEQIKWFMEEIRKRFEFGDNAEISIELDPRYLTDEQIKAIKDAGFNRISLGVQDLDPKVQQAVNRVQPYELIKEKMEKLREAGFESINLDLIYGLPYQTKESFEKTVEKVIELNPDRIATYSFAYIPQVKPHQQLLPKEALPSAEEKLRIFEMVINKFQEAGYVYIGMDHFAKPEDELAVAQRKGELWRNFQGYTTKKGVELLGFGATSIGMLYDSYFQNWKTLRDYNKTVDEGKIPVFRGYVLNEDDFIRREVIMDIMCNLGVEFSKIENMFGINFREYFAKELEELKEMEEDGLIKVEEDRIKIMPVGRLLIRNVAMVFDAHLRRKKELNFSRTI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00630", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLFACALLALLGLATSCSFIVPRSEWRALPSECSSRLGHPVRYVVISHTAGSFCNSPDSCEQQARNVQHYHKNELGWCDVAYNFLIGEDGHVYEGRGWNIKGDHTGPIWNPMSIGITFMGNFMDRVPAKRALRAALNLLECGVSRGFLRSNYEVKGHRDVQSTLSPGDQLYQVIQSWEHYRE\n\n### Sequence 2:\nMKIWVDADACPKVIRETIVRAAERTGVECTFVANHLVPVPKRNNIHSIQVPSGFDIADDEIVKRTEPGDLVITSDIPLADEVITKGGQALSSRGELYTKETIKARLNIRDFMDTMRSSGIQTGGPSALSQTDRREFANHLDRLLAKR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00631", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTMHTTMATLTLTSLIPPILTTLVNPNKKNSYPHYVKSIVASTFIISLFPTTMFMCLDQEVIISNWHWATTQTTQLSLSFKLDYFSMMFIPVALFVTWSIMEFSLWYMNSDPNINQFFKYLLIFLITMLILVTANNLFQLFIGWEGVGIMSFLLISWWYARADANTAAIQAILYNRIGDIGFILALAWFILHSNSWDPQQMALLNANPSLTPLLGLLLAAAGKSAQLGLHPWLPSAMEGPTPVSALLHSSTMVVAGIFLLIRFHPLAENNPLIQTLTLCLGAITTLFAAVCALTQNDIKKIVAFSTSSQLGLMMVTIGINQPHLAFLHICTHAFFKAMLFMCSGSIIHNLNNEQDIRKMGGLLKTMPLTSTSLTIGSLALAGMPFLTGFYSKDHIIETANMSYTNAWALSITLIATSLTSAYSTRMILLTLTGQPRFPTLTNINENNPTLLNPIKRLAAGSLFAGFLITNNISPASPFQTTIPLYLKLTALAVTFLGLLTALDLNYLTNKLKMKSPLCTFYFSNMLGFYPSIMHRTIPYLGLLTSQNLPLLLLDLTWLEKLLPKTISQHQISTSIITSTQKGMIKLYFLSFFFPLILTLLLIT\n\n### Answer:", "output": "<SEQ_2D>CHHHHHHHHHHHHTTSHHHHHHHHCCCCCTHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCEEECCCBSSCCSSCCCBCCEECCHHHHHHHHHHHHHHHHHHHHHHHHTTTCSCHHHHHHHHHHHHHHHHHHHHBCSHHHHHHHHHHHHHHHHHHHTSSTTSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHSCCSHHHHHHHCSSCCHHHHHHHHHHHHHHTTCGGGCTTTGGGTTSCHHHHHHHTTTSSTHHHHHHHHHTHHHHTTCHHHHHHHHHHHHHHHHHHHHHHTTCCBHHHHHHHHHHHHHHHHHHHHHTTCHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTSCCBGGGCCCCTTTSHHHHHHHHHHHHHHHTCTTTTTTTTHHHHHHHHHHSSBCSHHHHHHHHHHHHHHHHHHHHHHHHTTSCCCSCSSCCCCCCCSSTTHHHHHHHHHHHHHHHHHHHHSCCSSCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHTSCSCCCCCHHHHHHHTTTSHHHHHHHHHHHHHHHHHHIIIIIIIITTHHHHHTTHHHHHHHHHHHHHTCCSCCCHHHHHHHHHHHHHHHHTTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00632", "category": "Mutation", "instruction": "Describe the effect of mutation P0A729-Q154E on protein 7-methyl-GTP pyrophosphatase.", "input": "### Wild-type:\nMPKLILASTSPWRRALLEKLQISFECAAPEVDETPRSDESPRQLVLRLAQEKAQSLASRYPDHLIIGSDQVCVLDGEITGKPLTEENARLQLRKASGNIVTFYTGLALFNSANGHLQTEVEPFDVHFRHLSEAEIDNYVRKEHPLHCAGSFKSQGFGITLFERLEGRDPNTLVGLPLIALCQMLRREGKNPLMG\n\n### Mutant:\nMPKLILASTSPWRRALLEKLQISFECAAPEVDETPRSDESPRQLVLRLAQEKAQSLASRYPDHLIIGSDQVCVLDGEITGKPLTEENARLQLRKASGNIVTFYTGLALFNSANGHLQTEVEPFDVHFRHLSEAEIDNYVRKEHPLHCAGSFKSEGFGITLFERLEGRDPNTLVGLPLIALCQMLRREGKNPLMG\n\n### Answer:", "output": "Increase in activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00633", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nFAP:high COL1A1:medium PDGFRB:high S100A4:medium ACTA2:low\n</cell>", "output": "This cell is most consistent with cancer-associated fibroblast, a cancer-associated fibroblast (CAF). Key markers such as COL1A1, S100A4, PDGFRB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "ovary", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["FAP", "COL1A1", "PDGFRB", "S100A4", "ACTA2"]}}
{"id": "eval_00634", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMERAEKREFVTELNEVFKASGSVVVAHYAGVTVAQMNDFRSKMRAAGGTVKVAKNRLAKIALQGTESEGMTNLFKGQTLIAYSVDPMIAPKVVMDFAKTNDKVVVLGGSMGATTLNAEAVKSLATLPSLDELRAKLLGLLNAPATRVATVVAAPASQLARVFSAYAKKDEAA\n\n### Sequence 2:\nMERAEKREFVTELNEVFKASGSVVVAHYAGATVAQMNDFRSKMRAAGGTVKVAKNRLAKIALQGTEAEGMSNLFKGQTLIAYSTDPITAPKVVMDFAKTNDKIIVLGGAMGTTTLNADAVKSLATLPSLDELRAKLLGMIQTPATRIAGVVAAPASQLARVFAAYAKKDEAA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00635", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMYPAHLLVLLAVCVSLLGASDIPPQPLNLVQFSNMIQCANHGRRPTSNYMDYGCYCGKGGSGTPVDELDRCCKIHDDCYGEAEKSQKCAPYWTWYTWKCGSDGPQCDDSKTGCQRFVCDCDATAAKCFAKAPYNKENYNIKTRCQ\n\n### Sequence 2:\nMAHTLQTVVLLLGTSILHPILADVNVMKDVTLGFGQALEKCREESQLTEEKMEEFFHFWSEDFKFEHRELGCAILCMSRHFNLLTDSSRMHHENTDKFIKSFPNNEVLSKHMVNLIHSCEQQHDADLDHCWRILRVAECFKRSCQEAGVAPSMELLMAEFIMESEIN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00636", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMQFLDFLIALLPALFWGSVVLINVFVGGGPYNQIRGTTLGALIVGLGLLITGFAKFNNPTVIIVGLISGALWAFGQANQLKSISLIGVSNTMPVSTGMQLVGTTLFSVIFLGEWSSMTQIIFGLIAMILLVTGVALTSLKAKNERQSDNPEFKKAMGILIVSTVGYVGFVVLGDIFGVGGTDALFFQSVGMAIGGFILSMNHKTSLKSTALNLLPGVIWGIGNLFMFYSQPKVGVATSFSLSQLLVIVSTLGGIFILGERKDRRQMTGIWAGIIIIVIAAIILGNLK\n\n### Sequence 2:\nMDLLIALLPALFWGSVVLINVLVGGGPYNQIRGTTFGALIIGIILLLTGNAKFDDLTIIIVGLISGAFWALGQGYQLKSVSLIGVSKTMPISTGLQLVGTTLFSAIFLGEWSTGVQVTLGLVAMVLLVIGIALTSIKGKNEASESSKNFGKAMPILLISTVGYVVYVVVAQIFGVDGMNALFFQSIGMAIGGLILSAKHETSVKSTLWNLIPGIVWGIGNLFMFYSQPKVGVATSFSFSQLLVIVSTLGGIFLLGEKKDKRQMIGIWAGIVLIVIAPL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00637", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BRAF\nLigand SMILES: COc1cccc(OC)c1C(=O)OCc1ccccc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and B-Raf kinase (BRAF) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "COc1cccc(OC)c1C(=O)OCc1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 272.3, "LogP": 3.06, "TPSA": 44.76, "QED": 0.78, "functional_groups": ["aromatic ring", "ester", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BRAF", "has_evidence": false}}
{"id": "eval_00638", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAPAG_CROS8\n\n### Answer:", "output": "Protein ApaG.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00639", "category": "Mutation", "instruction": "Describe the effect of mutation P03466-A207W on protein Nucleoprotein.", "input": "### Wild-type:\nMASQGTKRSYEQMETDGERQNATEIRASVGKMIGGIGRFYIQMCTELKLSDYEGRLIQNSLTIERMVLSAFDERRNKYLEEHPSAGKDPKKTGGPIYRRVNGKWMRELILYDKEEIRRIWRQANNGDDATAGLTHMMIWHSNLNDATYQRTRALVRTGMDPRMCSLMQGSTLPRRSGAAGAAVKGVGTMVMELVRMIKRGINDRNFARGENGRKTRIAYERMCNILKGKFQTAAQKAMMDQVRESRNPGNAEFEDLTFLARSALILRGSVAHKSCLPACVYGPAVASGYDFEREGYSLVG\n\n### Mutant:\nMASQGTKRSYEQMETDGERQNATEIRASVGKMIGGIGRFYIQMCTELKLSDYEGRLIQNSLTIERMVLSAFDERRNKYLEEHPSAGKDPKKTGGPIYRRVNGKWMRELILYDKEEIRRIWRQANNGDDATAGLTHMMIWHSNLNDATYQRTRALVRTGMDPRMCSLMQGSTLPRRSGAAGAAVKGVGTMVMELVRMIKRGINDRNFWRGENGRKTRIAYERMCNILKGKFQTAAQKAMMDQVRESRNPGNAEFEDLTFLARSALILRGSVAHKSCLPACVYGPAVASGYDFEREGYSLVG\n\n### Answer:", "output": "Enhances RNA-binding activity. Mutation of W207 to alanine resulted in gain of high-affinity RNA binding at physiological temperature.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00640", "category": "Mutation", "instruction": "Describe the effect of mutation P04305-A509K on protein Uncoating factor OPG117.", "input": "### Wild-type:\nMDAAIRGNDVIFVLKTIGVPSACRQNEDPRFVEAFKCDELERYIENNPECTLFESLRDEEAYSIVRIFMDVDLDACLDEIDYLTAIQDFIIEVSNCVARFAFTECGAIHENVIKSMRSNFSLTKSTNRDKTSFHIIFLDTYTTMDTLIAMKRTLLELSRSSENPLTRSIDTAVYRRKTTLRVVGTRKNPNCDTIHVMQPPHDNIEDYLFTYVDMNNNSYYFSLQQRLEDLVPDKLWEPGFISFEDAIKRVSKIFINSIINFNDLDENNFTTVPLVIDYVTPCALCKKRSHKHPHQLSLEN\n\n### Mutant:\nMDAAIRGNDVIFVLKTIGVPSACRQNEDPRFVEAFKCDELERYIENNPECTLFESLRDEEAYSIVRIFMDVDLDACLDEIDYLTAIQDFIIEVSNCVARFAFTECGAIHENVIKSMRSNFSLTKSTNRDKTSFHIIFLDTYTTMDTLIAMKRTLLELSRSSENPLTRSIDTAVYRRKTTLRVVGTRKNPNCDTIHVMQPPHDNIEDYLFTYVDMNNNSYYFSLQQRLEDLVPDKLWEPGFISFEDAIKRVSKIFINSIINFNDLDENNFTTVPLVIDYVTPCALCKKRSHKHPHQLSLEN\n\n### Answer:", "output": "About 30% gain of ATP binding, and less than 90% gain of ATPase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00641", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nVATE_METTH\n\n### Answer:", "output": "Produces ATP from ADP in the presence of a proton gradient across the membrane.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00642", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMISSNDFRTGTSIELDGSVWRVVEFLHVKPGKGSAFVRTKLKAVQSGSVVEKTFRAGEMLPQALLEKSTLQHTYMESGDYVFMDMSSYEETRLTAQQIGDSRKYLKEGMEVNVVSWNGNPLEVELPNSVVLEITETDPGVKGDTATGGTKPAILETGAQVMVPLFLSVGEKIKVDTRNDSYLGREN\n\n### Sequence 2:\nMPRSQINGNFIDKTFSIVANILLRIIPTTSGEREAFTYYRDGMSAQSEGNYAEALQNYYEAMRLEIDPYDRSYILYNIGLIHTSNGEHTKALEYYFRALERNPFLPQAFNNMAVICHYRGEQAIRQGDSEIAEAWFDQAAEYWKQAISLTPGNYIEAHNWLKITGRFG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00643", "category": "Mutation", "instruction": "Describe the effect of mutation P49626-A334F on protein Large ribosomal subunit protein uL4B.", "input": "### Wild-type:\nMSRPQVTVHSLTGEATANALPLPAVFSAPIRPDIVHTVFTSVNKNKRQAYAVSEKAGHQTSAESWGTGRAVARIPRVGGGGTGRSGQGAFGNMCRGGRMFAPTKTWRKWNVKVNHNEKRYATASAIAATAVASLVLARGHRVEKIPEIPLVVSTDLESIQKTKEAVAALKAVGAHSDLLKVLKSKKLRAGKGKYRNRRWTQRRGPLVVYAEDNGIVKALRNVPGVETANVASLNLLQLAPGAHLGRFVIWTEAAFTKLDQVWGSETVASSKVGYTLPSHIISTSDVTRIINSSEIQSAIR\n\n### Mutant:\nMSRPQVTVHSLTGEATANALPLPAVFSAPIRPDIVHTVFTSVNKNKRQAYAVSEKAGHQTSAESWGTGRAVARIPRVGGGGTGRSGQGAFGNMCRGGRMFAPTKTWRKWNVKVNHNEKRYATASAIAATAVASLVLARGHRVEKIPEIPLVVSTDLESIQKTKEAVAALKAVGAHSDLLKVLKSKKLRAGKGKYRNRRWTQRRGPLVVYAEDNGIVKALRNVPGVETANVASLNLLQLAPGAHLGRFVIWTEAAFTKLDQVWGSETVASSKVGYTLPSHIISTSDVTRIINSSEIQSAIR\n\n### Answer:", "output": "Leads to an efficient release from ACL4 with an accelerated assembly into the 60S subunit; when associated with e-332.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00644", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMPNFFIDRPIFAWVIAIIIMLAGGLAILKLPVAQYPTIAPPAVTISASYPGADAKTVQDTVTQVIEQNMNGIDNLMYMSSNSDSTGTVQITLTFESGTDADIAQVQVQNKLQLAMPLLPQEVQQQGVSVEKSSSSFLMVVGVINTDGTMTQEDISDYVAANMKDAISRTSGVGDVQLFGSQYAMRIWMNPNELNKFQLTPVDVITAIKAQNAQVAAGQLGGTPPVKGQQLNASIIAQTRLTSTEEFGKILLKVNQDGSRVLLRDVAKIELGGENYDIIAEFNGQPASGLGIKLATGANALDTAAAIRAELAKMEPFFPSGLKIVYPYDTTPFVKISIHEVVKTLVEAIILVFLVMYLFLQNFRATLIPTIAVPVVLLGTFAVLAAFGFSINTLTMFGMVLAIGLLVDDAIVVVENVERVMAEEGLPPKEATRKSMGQIQGALVGIAMVLSAVFVPMAFFGGSTGAIYRQFSITIVSAMALSVLVALILTPALCATMLKPIAKGDHGEGKKGFFGWFNRMFEKSTHHYTDSVGGILRSTGRYLVLYLIIVVGMAYLFVRLPSSFLPDEDQGVFMTMVQLPAGATQERTQKVLNEVTHYYLTKEKNNVESVFAVNGAGAAGAGQNTGIAFVSLKDWADRPGEENKVEAITMRATRAFSQIKDAMVFAFNLPAIVELGTATGFDFELIDQAGLGHEKLTQARNQLLAEAAKHPDMLTSVRPNGLEDTPQFKIDIDQEKAQALGVSINDINTTLGAAWGGSYVNDFIDRGRVKKVYVMSEAKYRMLPDDIGDWYVRAADGQMVPFSAFSSSRWEYGSPRLERYNGLPSMEILGQAAPGKSTGEAMELMEQLASKLPTGVGYDWTGMSYQERLSGNQAPSLYAISLIVVFLCLAALYESWSIPFSVMLVVPLGVIGALLAATFRGLTNDVYFQVGLLTTIGLSAKNAILIVEFAKDLMDKEGKGLIEATLDAVRMRLRPILMTSLAFILGVMPLVISTGAGSGAQNAVGTGVMGGMVTATVLAIFFVPVFFVVVRRRFSRKNEDIEHSHTVDHH\n\n### Answer:", "output": "<SEQ_2D>CCGGGCCCHHHHHHHHHHHHHHHHHHHHTCCBCSCCSCSCCEEEEEEECTTCCHHHHHHHTHHHHHTTCSSCTTEEEEEEEEESSSEEEEEEEECTTCCHHHHHHHHHHHHHHHGGGSCHHHHHHCCEEEECCCCEEEEEEEEECSSCCCHHHHHHHHHHHTHHHHHHSTTEEEEEESSCCEEEEEEECSHHHHTTTCCHHHHHHHHHHHSCCCCCEEETCSSCCTTCCCCEEEECSCCCCCSSTTTTCEEEECTTSCEEEHHHHEEEEEEESCCSCEEEETTEEEEEEEEEECTTCCHHHHHHHHHHHHHHHGGGCCTTEEEEEEEESHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCHHHHHHHHTHHHHHHHHHHHHHHHHTCCBCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHTGGGGCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCCCCSSCTTCCCCTTTSHHHHHHHHHHHHHHHHHHHHTTCCHHHHHHHHHHHHHHHHHHHTSCBCSSCCCCCSEEEEEEECSTTCCHHHHHHHHHHHHHHHTTTTTTTEEEEEEEEEECSSCEEEEEEEEEEEECCGGGCCSSTTSHHHHHHHHHHHHHHTCSSEEEEECCCSSTTTCSSSSEEEEEEECSSCHHHHHHHHHHHHHHHHHTCTTTEESCCEESCCCEEEEEEEECHHHHHHTTCCHHHHHHHHHHHHTCEEEEEEEETTEEEEEEEEECGGGSSSGGGGGGCEEECSSSCEEEGGGTEEEEEEEECSEEEEETTEEEEEEEEEECSSCCHHHHHHHHHHHHTSSCTTEEEEECHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTCCSSHHHHHHTTHHHHHHHHHHHHHHTTCCBSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHSSCCCHHHHHHHHHHHTHHHHHHHHHHHHHHHHHHHSCCSTTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTTCCTTSCCSSCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00645", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMWPRSHRHLCLAFLLVCVLSAISFLIHFHQDSIRHGLGLSVLCPDRRLVTPPVAIFCLPGTPMSPNTSSPCPQHPASLSGTWTIYPDGRFGNQMGQYATLLALAQLNGRRAFILPAMHAALAPVFRITLPVLAPEVDSRTPWQELRLHDWMSEEYADLGDPFLKLSGFPCSWTFFHHLREQIRSEFTLHDHLREEAQSVLRRLRLGRSGARPRTFVGVHVRRGDYLQVMPQRWKGVVANSAYLREAMDWFRARHEAPVFVVTSNGMEWCRENIDASKGDVMFAGDGQEASPWKDFALLAQCNHTIMTIGTFGFWAAYLAGGDTVYLANFTLPDSEFLKIFKPEAAFLPEWVGINADLSPLWPLAEP\n\n### Sequence 2:\nMWPLSHRHLCLAFLLVCVLSAISFFLHIHQDSIRHGLGLSVLCPDRRLVTPPVAIFCLPGTPMSPNTSSPCPQNSASLSGTWTIYPDGRFGNQMGQYATLLALAQLNGRRAFILPAMHAALAPVFRITLPVLAPEVDSRTPWRELRLHDWMSEEYADLGDPFLKLSGFPCSWTFFHHLREQIRSEFTLHDHLREEAQSVLRRLRLGRSGDRPRTFVGVHVRRGDYLQVMPQRWKGVVGNGAYLREAMDWFRARHEAPVFVVTSNGMDWCRENIDASKGDVMFAGDGQEASPWKDFALLTQCNHTIMTIGTFGFWAAYLAGGDTVYLANFTLPDSEFLKIFKPEAAFLPEWVGINADLSPLWTLAEP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00646", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCD8B PRF1 GZMA GZMB CD8A CD3D NKG7\n</cell>", "output": "This cell is most consistent with cytotoxic T cell, an effector cytotoxic CD8+ T cell. Key markers such as GZMA, NKG7, GZMB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "skin", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD8B", "PRF1", "GZMA", "GZMB", "CD8A", "CD3D", "NKG7"]}}
{"id": "eval_00647", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSRVAKAPVSIPAGVEVTLNEQTLTVKGAKGSLTRVINNAVNVVIEDGVVKFLPVEGVVNAWAQAGTTRALVNNMVVGVSQGFERKLKLVGVGYRAKLVGADIDLTLGFSHPLVHKLPAGVTAECPSQTDIVLRGVDKQLIGQVAAEIRGYRPPEPYKGKGVRYDDEEVRRKEAKKK\n\n### Sequence 2:\nMIDPDGYRPNVGIVLMRQDGQVFWARRVRRDGWQFPQGGMNTDETPVEAMYRELQEETGLLPEHVEVLGATPGWLRYKLPARAIRRNERQVCIGQKQVWFLLRLTGDESHVKLDHTDSPEFDHWRWVDFWYPVEHVVMFKRGVYARALRHLAPLARGVAGQGVTAMPKSAAEAWMPGHTAGHDRPRKRPRTRGYWPKKATGDGPAS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00648", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRECO_PSESM\n\n### Answer:", "output": "Involved in DNA repair and RecF pathway recombination.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00649", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMARVSRGVQAHAKHKKILKKAKGYYGARSKVYRVAKQAVIKAGQYAYRDRRQRRRKFRRLWIVRINAEARNNGLSYSRMIDGISKADIEIDRKVLSDIAIFDKTAFAKIADQAKQALVV\n\n### Sequence 2:\nMRLNTLSPAEGAKHAPKRVGRGIGSGLGKTAGRGHKGQNSRSGGGVRRGFEGGQMPLYRRLPKFGFTSRKAMITAEVRLSELALVEGDVIDLNTLKAANVVGIQMEFVKVILSGEVNRAVTLRGLRVTKGARAAIEAAGGKIEE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00650", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL11_CYAP4\n\n### Answer:", "output": "50S ribosomal protein L11.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00651", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nYGFNKCTQYEFDIHHVLCIRKKITNLTEAISDIPRYTTHLNLTHNEIQVLPPWSFTNLSALVDLRLEWNSIWKIDEGAFRGLENLTLLNLVENKIQSVNNSFEGLSSLKTLLLSHNQITHIHKDAFTPLIKLKYLSLSRNNISDFSGILEAVQHLPCLERLDLTNNSIMYLDHSPRSLVSLTHLSFEGNKLRELNFSALSLPNLTNLSASRNGNKVIQNVYLKTLPQLKSLNLSGTVIKLENLSAKHLQNLRAMDLSNWELRHGHLDMKTVCHLLGNLPKLETLVFQKNVTNAEGIKQLAKCTRLLFLDLGQNSDLIYLNDSEFNALPSLQKLNLNKCQLSFINNRTWSSLQNLTSLDLSHNKFKSFPDFAFSPLKHLEFLSLSRNPITELNNLAFSGLFALKELNLAACWIVTIDRYSFTQFPNLEVLDLGDNNIRTLNHGTFRPLKKLQSLILSHNCLKILEPNSFSGLTNLRSLDLMYNSLSYFHEHLFSGLEKLLILKLGFNKITYETTRTLQYPPFIKLKSLKQLNLEGQRHGIQVVPSNFFQGLGSLQELLLGKNPSVFLDHHQFDPLINLTKLDISGTKDGDRSLYLNASLFQNLKRLKILRLENNNLESLVPDMFSSLQSLQVFSLRFNNLKVINQSHLKNLKSLMFFDVYGNKLQCTCDNLWFKNWSMNTEEVHIPFLRSYPCQQPGSQSLLIDFDDAMC\n\n### Answer:", "output": "<SEQ_2D>CCCSSCEEETTEEEEEECCSSCCCCHHHHHTTSCTTCCEEECCSSCCCEECTTTTTTCTTCSEEECCSSCCCEECTTTTTTCTTCCEEECCSSCCCEECSCCTTCTTCCEEECCSSCCCEECTTSSTTCTTCCEEECCSSCCCCHHHHHHHHTTCTTCCEEECCSSCCCSCCCSCEECSSCCEEECTTSCCSCEECTTEECTTCCEEECCSCCSSCBCCEECTTCTTCCEEECTTCCBCGGGCCGGGCTTCCEEECCCSSGGGCCCCHHHHHHHHTTCTTCCEEECCSSSCCHHHHHHHHHHCCCSEEECTTCTTCCEECSSTTTTCTTCSEEECTTSCCCEECTTTSTTCTTCCEEECCSSCCSCCCTTTTTTCTTCCEEECTTCCCSSCCGGGGTTCTTCCEEECTTSCCCCCCTTSSTTCTTCCEEECTTSCCCEECTTTTTTCTTCCEEECTTSCCCEECTTTTTTCTTCCEEECCSSCCCCCCTTTTTTCTTCCEEECTTSCCCBSSCCCCSSCTTTTCTTCCEEECTTCTTCCEECCTTTTTTCTTCCEEECTTCSCEEECSCTTTTCTTCCEEECCCCCSHHHHEEECTTTTTTCTTCCEEECCSSCCCCCCTTTTTTCTTCCEEECCSSCCCCCCTTTTTTCTTCCEEECTTSCCCSSGGGHHHHHHHHHCSSCBCTTGGGCBCSSTTCCCBGGGCCCTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00652", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "SFTPC, SFTPB, ABCA3, SFTPA1", "output": "This profile is most consistent with alveolar type II cell, which is commonly found in ovary under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "ovary", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["SFTPC", "SFTPB", "ABCA3", "SFTPA1"]}}
{"id": "eval_00653", "category": "Mutation", "instruction": "Describe the effect of mutation P31460-Y195H on protein Galactonate operon transcriptional repressor.", "input": "### Wild-type:\nMTLNKTDRIVITLGKQIVHGKYVPGSPLPAEAELCEEFATSRNIIREVFRSLMAKRLIEMKRYRGAFVAPRNQWNYLDTDVLQWVLENDYDPRLISAMSEVRNLVEPAIARWAAERATSSDLAQIESALNEMIANNQDREAFNEADIRYHEAVLQSVHNPVLQQLSIAISSLQRAVFERTWMGDEANMPQTLQEYKALFDAIRHQDGDAAEQAALTMIASSTRRLKEIT\n\n### Mutant:\nMTLNKTDRIVITLGKQIVHGKYVPGSPLPAEAELCEEFATSRNIIREVFRSLMAKRLIEMKRYRGAFVAPRNQWNYLDTDVLQWVLENDYDPRLISAMSEVRNLVEPAIARWAAERATSSDLAQIESALNEMIANNQDREAFNEADIRYHEAVLQSVHNPVLQQLSIAISSLQRAVFERTWMGDEANMPQTLQEHKALFDAIRHQDGDAAEQAALTMIASSTRRLKEIT\n\n### Answer:", "output": "Completely responsive to D-galactonate.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00654", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BTV5-D313S on protein Fibronectin type III and SPRY domain-containing protein 1.", "input": "### Wild-type:\nMEEQREALRKIIKTLAVKNEEIQSFIYSLKQMLLNVEANSAKVQEDLEAEFQSLFSLLEELKEGMLMKIKQDRASRTYELQNQLAACTRALESSEELLETANQTLQAMDSEDFPQAAKQIKDGVTMAPAFRLSLKAKVSDNMSHLMVDFAQERQMLQALKFLPVPSAPVIDLAESLVADNCVTLVWRMPDEDSKIDHYVLEYRRTNFEGPPRLKEDQPWMVIEGIRQTEYTLTGLKFDMKYMNFRVKACNKAVAGEFSEPVTLETPAFMFRLDASTSHQNLRVDDLSVEWDAMGGKVQDI\n\n### Mutant:\nMEEQREALRKIIKTLAVKNEEIQSFIYSLKQMLLNVEANSAKVQEDLEAEFQSLFSLLEELKEGMLMKIKQDRASRTYELQNQLAACTRALESSEELLETANQTLQAMDSEDFPQAAKQIKDGVTMAPAFRLSLKAKVSDNMSHLMVDFAQERQMLQALKFLPVPSAPVIDLAESLVADNCVTLVWRMPDEDSKIDHYVLEYRRTNFEGPPRLKEDQPWMVIEGIRQTEYTLTGLKFDMKYMNFRVKACNKAVAGEFSEPVTLETPAFMFRLDASTSHQNLRVDDLSVEWDAMGGKVQDI\n\n### Answer:", "output": "Increased ability to associate with microtubules; when associated with D-317; E-322 and D-324. Known functional changes.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00655", "category": "Mutation", "instruction": "Describe the effect of mutation P08236-F176L on protein Beta-glucuronidase.", "input": "### Wild-type:\nMARGSAVAWAALGPLLWGCALGLQGGMLYPQESPSRECKELDGLWSFRADFSDNRRRGFEEQWYRRPLWESGPTVDMPVPSSFNDISQDWRLRHFVGWVWYEREVILPERWTQDLRTRVVLRIGSAHSYAIVWVNGVDTLEHEGGYLPFEADISNLVQVGPLPSRLRITIAINNTFTPTTLPPGTIQYLTDTSKYPKGYFVQNTYFDFFNYAGLQRSVLLYTTPTTYIDDITVTTSVEQDSGLVNYQISVKGSNLFKLEVRLLDAENKVVANGTGTQGQLKVPGVSLWWPYLMHERPAYL\n\n### Mutant:\nMARGSAVAWAALGPLLWGCALGLQGGMLYPQESPSRECKELDGLWSFRADFSDNRRRGFEEQWYRRPLWESGPTVDMPVPSSFNDISQDWRLRHFVGWVWYEREVILPERWTQDLRTRVVLRIGSAHSYAIVWVNGVDTLEHEGGYLPFEADISNLVQVGPLPSRLRITIAINNTLTPTTLPPGTIQYLTDTSKYPKGYFVQNTYFDFFNYAGLQRSVLLYTTPTTYIDDITVTTSVEQDSGLVNYQISVKGSNLFKLEVRLLDAENKVVANGTGTQGQLKVPGVSLWWPYLMHERPAYL\n\n### Answer:", "output": "Substitution of C4 with S is obligatory for Cd(II) detection.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00656", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSHMRIGILYICTGKYDIFWKDFYLSAERYFMQDQSFIIEYYVFTDSPKLYDEENNKHIHRIKQKNLGWPDNTLKRFHIFLRIKEQLERETDYLFFFNANLLFTSPIGKEILPPSDSNGLLGTMHPGFYNKPNSEFTYERRDASTAYIPEGEGRYYYAGGLSGGCTKAYLKLCTTICSWVDRDATNHIIPIWHDQSLINKYFLDNPPAITLSPAYLYPEGWLLPFEPIILIRDKNKPQYGGHELLRRKN\n\n### Answer:", "output": "<SEQ_2D>CCCEEEEEEECCSSGGGHHHHHHHHHHHTBCCSTTCEEEEEEEECCSSCTTTTTCSSEEEEECCCCCTTHHHHTTTTTTTTSHHHHHHHCSEEEEECSSEEECSCBCGGGSCCTTSCCEEEEECTTSSSCCGGGCCCCCCTTSTTCCCSSCCSCCEEEEEEEEEHHHHHHHHHHHHHHHHHHHTTTCCCTTHHHHHHHHHHHHSCCSEEECGGGGEETTCCCSSCCSEEECCSCCTTTCSCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00657", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMFSRVIFCTFLILPLLAAATAIPRTNPPAPTCTTGSLQCCNSVQAASNPVVGLLAGLLGIVLGPITGQVGLTCSPITVIGVGGTSCSAQTVCCNGNSFNGLIVVGCSPVNISL\n\n### Sequence 2:\nSYMVTLKTPSGEQKVEVSPDSYILDAAEEAGVDLPYSCRAGSCSSCAGKVESGTVDQSDQSFLDDDQMDSGFVLTCVAYATSDCTIVTHQEENLY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00658", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD4, IL2RA, CD3D, FOXP3, IKZF2", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as CD3D, FOXP3, CD4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "ovary", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD4", "IL2RA", "CD3D", "FOXP3", "IKZF2"]}}
{"id": "eval_00659", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSGLRVYSTSVTGSREIKSQQSEVTRILDGKRIQYQLVDISQDNALRDEMRALAGNPKATPPQIVNGDQYCGDYELFVEAVEQNTLQEFLKLALEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCEEEECSSCCSSSHHHHHHHHHHHHHHTTCCCEEEETTSCHHHHHHHHHHTTCTTCCSCEEESSSSEEEEHHHHHHHHTTTCHHHHHTCSCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00660", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "[O-]c1ccccc1", "output": "The molecule contains the following functional groups: aromatic ring.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "[O-]c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 93.1, "LogP": 0.76, "TPSA": 23.06, "QED": 0.46, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring"]}}
{"id": "eval_00661", "category": "Mutation", "instruction": "Describe the effect of mutation P9WKD3-F117I on protein Beta-lactamase.", "input": "### Wild-type:\nMRNRGFGRRELLVAMAMLVSVTGCARHASGARPASTTLPAGADLADRFAELERRYDARLGVYVPATGTTAAIEYRADERFAFCSTFKAPLVAAVLHQNPLTHLDKLITYTSDDIRSFSPVAQQHVQTGMTIGQLCDAAIRYSDGTAANLLLADLGGPGGGTAAFTGYLRSLGDTVSRLDAEEPELNRDPPGDERDTTTPHAIALVLQQLVLGNALPPDKRALLTDWMARNTTGAKRIRAGFPADWKVIDKTGTGDYGRANDIAVVWSPTGVPYVVAVMSDRAGGGYDAEPREALLAEAAT\n\n### Mutant:\nMRNRGFGRRELLVAMAMLVSVTGCARHASGARPASTTLPAGADLADRFAELERRYDARLGVYVPATGTTAAIEYRADERFAFCSTFKAPLVAAVLHQNPLTHLDKLITYTSDDIRSISPVAQQHVQTGMTIGQLCDAAIRYSDGTAANLLLADLGGPGGGTAAFTGYLRSLGDTVSRLDAEEPELNRDPPGDERDTTTPHAIALVLQQLVLGNALPPDKRALLTDWMARNTTGAKRIRAGFPADWKVIDKTGTGDYGRANDIAVVWSPTGVPYVVAVMSDRAGGGYDAEPREALLAEAAT\n\n### Answer:", "output": "Significant decrease in ampicillin resistance. 2-fold and 3-fold decrease in catalytic efficiency with ampicillin and nitrocefin as substrate, respectively, mainly due to a decrease in substrate affinity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00662", "category": "Mutation", "instruction": "Describe the effect of mutation D0ZPH9-I479A on protein E3 ubiquitin-protein ligase SspH2.", "input": "### Wild-type:\nMPFHIGSGCLPATISNRRIYRIAWSDTPPEMSSWEKMKEFFCSTHQTEALECIWTICHPPAGTTREDVINRFELLRTLAYAGWEESIHSGQHGENYFCILDEDSQEILSVTLDDAGNYTVNCQGYSETHRLTLDTAQGEEGTGHAEGASGTFRTSFLPATTAPQTPAEYDAVWSAWRRAAPAEESRGRAAVVQKMRACLNNGNAVLNVGESGLTTLPDCLPAHITTLVIPDNNLTSLPALPPELRTLEVSGNQLTSLPVLPPGLLELSIFSNPLTHLPALPSGLCKLWIFGNQLTSLPVL\n\n### Mutant:\nMPFHIGSGCLPATISNRRIYRIAWSDTPPEMSSWEKMKEFFCSTHQTEALECIWTICHPPAGTTREDVINRFELLRTLAYAGWEESIHSGQHGENYFCILDEDSQEILSVTLDDAGNYTVNCQGYSETHRLTLDTAQGEEGTGHAEGASGTFRTSFLPATTAPQTPAEYDAVWSAWRRAAPAEESRGRAAVVQKMRACLNNGNAVLNVGESGLTTLPDCLPAHITTLVIPDNNLTSLPALPPELRTLEVSGNQLTSLPVLPPGLLELSIFSNPLTHLPALPSGLCKLWIFGNQLTSLPVL\n\n### Answer:", "output": "Increase of ubiquitin ligase activity; when associated with A-481 and A-483. The mutation in the SspH2 protein causes a functional change in the ubiquitination activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00663", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMRVKMHVKKGDTVLVASGKYKGRVGKVKEVLPKKYAVIVEGVNIVKKAVRVSPKYPQGGFIEKEAPLHASKVRPICPACGKPTRVRKKFLENGKKIRVCAKCGGALDTEE\n\n### Answer:", "output": "<SEQ_2D>CCCCCSCCTTSEEEECSSTTTTCEEEEEEEETTTTEEEETTSSEEEEECCSSSSSTTCCEEEEECCEEGGGEEEBCTTTCSBCCEEEEECSSSCEEEEESSSCCBCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00664", "category": "Mutation", "instruction": "Describe the effect of mutation P41180-Q1011E on protein Extracellular calcium-sensing receptor.", "input": "### Wild-type:\nMAFYSCCWVLLALTWHTSAYGPDQRAQKKGDIILGGLFPIHFGVAAKDQDLKSRPESVECIRYNFRGFRWLQAMIFAIEEINSSPALLPNLTLGYRIFDTCNTVSKALEATLSFVAQNKIDSLNLDEFCNCSEHIPSTIAVVGATGSGVSTAVANLLGLFYIPQVSYASSSRLLSNKNQFKSFLRTIPNDEHQATAMADIIEYFRWNWVGTIAADDDYGRPGIEKFREEAEERDICIDFSELISQYSDEEEIQHVVEVIQNSTAKVIVVFSSGPDLEPLIKEIVRRNITGKIWLASEAWA\n\n### Mutant:\nMAFYSCCWVLLALTWHTSAYGPDQRAQKKGDIILGGLFPIHFGVAAKDQDLKSRPESVECIRYNFRGFRWLQAMIFAIEEINSSPALLPNLTLGYRIFDTCNTVSKALEATLSFVAQNKIDSLNLDEFCNCSEHIPSTIAVVGATGSGVSTAVANLLGLFYIPQVSYASSSRLLSNKNQFKSFLRTIPNDEHQATAMADIIEYFRWNWVGTIAADDDYGRPGIEKFREEAEERDICIDFSELISQYSDEEEIQHVVEVIQNSTAKVIVVFSSGPDLEPLIKEIVRRNITGKIWLASEAWA\n\n### Answer:", "output": "iCa was higher in QQ genotype subjects than in the QE group (1. 227 +/- 0. 002 vs. 1. 255 +/- 0. 008 mmol/liter; P = 0. 002). After resolution of phase for the doubly heterozygous subjects, analysis was conducted on haplotypes across all three loci. As expected, subjects with SRQ and ARE haplotypes are relatively hypocalcemic, and those with AGQ are hypercalcemic, relative to subjects with the common ARQ haplotype. The mutation activates one allele of the CaSR gene and causes autosomal dominant familial hypocalciuric hypercalcemia (FHH).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00665", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSVRTKICGLSTEETVATAVRYGADYIGFVFFEKSPRSVTPEKAAMLIRQIPDHVQKMGVFVDPDNNLLERALASGLTGFQLHGHETAERIASIRETFPKVKIWKALSIANSQDLAQAPHYRGLADRLLYDARTDGVLPGGMGRRFDWRLLKEYKHPLPWALSGGLDANNIAQAVAITGAELVDISSGVETSPGIKDMDKIAQFLQAVRLL\n\n### Sequence 2:\nMTTQLPHEKGFHISWDQIHRDSRALAWRLDNHGPDDGAWKAVVAITRGGMAPAMIVARELDIRTVDTISVVSYHSGGGKADERREAKVLKGPDADLMGDGEGVLIIDDLVDSGKTLELVRTLYPKAHFATVYAKPEGEPMVDTFITGVSQDTWIFFPWDMALQYVDPYRGKD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00666", "category": "Mutation", "instruction": "Describe the effect of mutation Q13325-K415E on protein Interferon-induced protein with tetratricopeptide repeats 5.", "input": "### Wild-type:\nMSEIRKDTLKAILLELECHFTWNLLKEDIDLFEVEDTIGQQLEFLTTKSRLALYNLLAYVKHLKGQNKDALECLEQAEEIIQQEHSDKEEVRSLVTWGNYAWVYYHMDQLEEAQKYTGKIGNVCKKLSSPSNYKLECPETDCEKGWALLKFGGKYYQKAKAAFEKALEVEPDNPEFNIGYAITVYRLDDSDREGSVKSFSLGPLRKAVTLNPDNSYIKVFLALKLQDVHAEAEGEKYIEEILDQISSQPYVLRYAAKFYRRKNSWNKALELLKKALEVTPTSSFLHHQMGLCYRAQMIQI\n\n### Mutant:\nMSEIRKDTLKAILLELECHFTWNLLKEDIDLFEVEDTIGQQLEFLTTKSRLALYNLLAYVKHLKGQNKDALECLEQAEEIIQQEHSDKEEVRSLVTWGNYAWVYYHMDQLEEAQKYTGKIGNVCKKLSSPSNYKLECPETDCEKGWALLKFGGKYYQKAKAAFEKALEVEPDNPEFNIGYAITVYRLDDSDREGSVKSFSLGPLRKAVTLNPDNSYIKVFLALKLQDVHAEAEGEKYIEEILDQISSQPYVLRYAAKFYRRKNSWNKALELLKKALEVTPTSSFLHHQMGLCYRAQMIQI\n\n### Answer:", "output": "Reduces binding to RNA and DNA and impairs antiviral activity; when associated with E-384.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00667", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMIPPHRCLTDPLLQGSWPGYLPPMPKRARRMRLTTWRSPVTAVIAALLAAFLTPAAAHGAPRESAPVYSYENAIRESVWVDTRLDGDGDGKTDRVAVDVVRPRELARQGRKIPVIMDASPYYSCCGRGNESQKKTYDANGDVVRMPLFYDNYFVPRGYAFVGVDLAGTNRSDGCVDVGGRSDIQSAKAVIDWLNGRAHGYTSRTGTARAKATWTNGRTGMIGKSWDGTVANGVAATGVKGLKTIVPISAISSWYDYYFAKGAPLYDSGPDWLSDYVDSPDARTKCAAVQRKLVDEAPRTGDWTSLWTERDYVKDASKVRASVFLVHGMQDLNVRAKNFGQWWSALAKNGVERKIWLSQTGHVDPFDFRRTAWVDTLHRWFDHELLGYDNGVDREPTADIERHPDQWVTSTLWPPRGTDAVTLRPGTGTQAGVGTLGLRTGSGTETFTDDPRLSETDWAAHIDESTASKAGFVTAPLAGDVRLSGSSKVTVTATPTTSTAHLSAVLVDLGPDTIRDYADGGEGITTLTDRTCWGASTAGDSACFKNTRATTAAVDYTVLSRGWADLGNHASARKGVPLTPGKAYTITLDLAATDHVVPKGHRLALIVAGTDKDLIDPPSSTPTLTLDLARTSARVPLVGGAAAFTRATAQSGTAADATVLDGVREPHTAHRVPGGGL\n\n### Sequence 2:\nKSLPVFDTSQVIREVVYVQSDLDTDGDGKGDLLPVTVFRPVESQDQ-LKVPALYTASPYFGGIIDNVKTNHNVDENLTDATTWTNPKYVAKPLVKSPAPSGQDVPATELATGQSSYGLNEYLLARGFASVFSGAIGNRHGDGIRITGSPEETISQKEVIEWLTGDRVAYTDRTRRFETKAGWCSGNVGMTGRSYLGTLQIAIATTGVKGLKTVVSEAAISSWYDYYREHGLVIAPSECQGEDMDKLAEVCQSNLWDGGNFTAKKAYEAEQAELLAAQDRATGQYSDFWESRNYRHHADGIKCSWISVHGLNDWNVKPKNVYKIWQKVKQLPVESHLFLHQGPHYNMNNLVSIDFTDLMNLWFVHELLEVENGAYEQWPKVMIQDNLEADKWHAESDWANDLGQASLYSPTAD-----GYLSTVENGTGQLTFTDLGGTEFKKAGISETDWEYQFISGEKKWAKASLRFESEEFLHPTTLVGRPKVQVRVAANKTVGQLSVALVDLGTRQRLTATPKIFARGNQ---------PFGYRFEADSLQEFVPDKATKAK----LITKAHMNLQNYQDMKQPSKLEAGQFVDLEFELQPTYYTLPAGAKLCLIIYSTDQGMTKRPLETEDYTVDLAGTA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00668", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSEVVIKEQLEQYISKIERLEQEKADLSQEVKDIFQDASSHGFDVKAMKSILKLKKLDKDKLAEQDAMLELYRDTLGI\n\n### Sequence 2:\nMVSIRLSRGGSKKRPFYHLTVTDSRSARDGAFIERLGFFNPIARGQEERLRVDNERLEYWKGQGAQMSDRVAKLIKDAAAA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00669", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nACP_LEPIN\n\n### Answer:", "output": "Belongs to the acyl carrier protein (ACP) family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00670", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTAGH_STAAB\n\n### Answer:", "output": "Teichoic acids export ATP-binding protein TagH.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00671", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: N#Cc1nc(NC2CCCC2)c(N)c(N2CCOCC2)n1\nDescriptors: MW=288.36, LogP=1.12, TPSA=100.09, HBD=2, HBA=7", "output": "Descriptors: MW=288.36, LogP=1.12, TPSA=100.09, HBD=2, HBA=7, QED=0.86, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "N#Cc1nc(NC2CCCC2)c(N)c(N2CCOCC2)n1", "compound_name": null, "pubchem_cid": null, "MW": 288.36, "LogP": 1.12, "TPSA": 100.09, "QED": 0.86, "functional_groups": ["ether", "nitrile"], "moleculenet_labels": {"label": 2.59}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00672", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMFDKHTHTLIAQRLDQAEKQREQIRAISLDYPEITIEDAYAVQREWVRLKIAEGRTLKGHKIGLTSKAMQASSQISEPDYGALLDDMFFHDGSDIPTDRFIVPRIEVELAFVLAKPLRGPNCTLFDVYNATDYVIPALELIDARCHNIDPETQRPRKVFDTISDNAANAGVILGGRPIKPDELDLRWISALMYRNGVIEETGVAAGVLNHPANGVAWLANKLAPYDVQLEAGQIILGGSFTRPVPARKGDTFHVDYGNMGSISCRFV\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHHHHHHHHHTCCBCCHHHHSTTCCHHHHHHHHHHHHHHHHHTTCCEEEEEEECHHHHHCTTCSCCSCEEEEEEGGGEEETTEEECGGGCSSCEEEEEEEEEESSCBCCSSCCHHHHHHHEEEEEEEEEEECCSBCCCCCCCCCCCCHHHHHHTGGGEEEEEECSCCBCTTSSCGGGCEEEEEETTEEEEEEEGGGTTTSTTHHHHHHHHHHGGGTCCBCTTCEEECCCSSCCEECCTTCEEEEECGGGCEEEEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00673", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "MBP, PLP1, MOBP", "output": "The expression pattern is consistent with oligodendrocyte in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "oligodendrocyte", "tissue": "liver", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["MBP", "PLP1", "MOBP"]}}
{"id": "eval_00674", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nFOXP3 CD3D CTLA4 CD4\n</cell>", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as CD4, CTLA4, CD3D support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "blood", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["FOXP3", "CD3D", "CTLA4", "CD4"]}}
{"id": "eval_00675", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSDPSTGLPNKLFAFIGKTTASTSYVEKNLLSSTTGAAMVGLPSGGNLLQAQYFVTPEQFGAIGDGVTDDTQAILKTITFANTNNIQVRADKNYRFTSSIAMSGVRWYGGTFTGNGGTMISTVSCWMENVRFEKCYVKMLGGDCRFYRNIFSNATSTAAFLMQAMTSEGTLDFSYNEMYGCKYAILQQGTGEVMTYGRYSNNYIHDIKGDAIELNVVQKHYTEGLIIENNHIANVDASGQGANWGIGIGVAGSGPYGVDVPDSQYVRNFSIVGNRVYNCRQCLHVEMGKNFTIRDNEVYPNTAVSTGTGLTTCGVALYGCQDFEVDGLTGYLLNDPSVSTRMVFIDWGVNNGRYAGPPINFTIKNLDIPESSIEIATSGSDAWENSTIVSNINCNVFKWRGLPSSSTFNNIRCRSIDFIGQHGSGEGSGGGFYTRSQFTYMKWVGCTALSGDETTVSFAKIYTDRCDQVGNNFGVPTAVDGTGHRGPVLTTISEQYFTAYDEFPGGREFPTGTVIHCASGKKHVVTVGGAFFSDNEKIKATVTGQTYLQSNALNWASNGYAKAAGTKIVIPGAGANGGDLVTTIARATYVTNSLYTIDIADPIVTPTAENTQIKALNPVTFVTVNNA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCHHHHTSTTGGGGSBCTTSSBGGGGCCEECGGGGTCCCEEEEECHHHHHHHHHHHHHHTCEEECCSEEECCSCEEEESCEEESCEEEEEEEEEEEEESCEEESCEEEEEEEEEEEEEEEEESCEEECCCSSEEEEEEEESSCEEEEEESCEEESSSEEEEEEEEESCEEEEEEESCEEEEESSCSEEEESCTTCCTTCEEEESCEEEEECCTTSCTTTTEEEEEECCSCCCTTSCGGGSEESCEEESCEEEEESEEEEEESEEEEEEEEEEEEECTTSSTTSCCCCEEEEEESCEEEEEEEEEECCSSCTTTCCEEEEEECEEETTEEECCCEEEEEEEEECTTSEEEEECCCCSSSCCEEEEEEEECSEEEEECCCSEEEEEEEEESEEEEESSCCTTSSTTTTSCCCCSCCEEEEESCEECBTTBCCEEEECCCCSEEEEESCSSCCCCCCSSTTTTSSCCCCCCCEEEESSSSCCCSCCBCTTCEEEETTSCEEEEEECBBCCCTTCCBCCCCTTCCEEEBSSSCTTSSTTCCCTTCEEEETTCSGGGCCEEEEEEEEEEEETTEEEEEESSCCSSCCCTTBCCEESSBCEEEEECCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00676", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTNADPHELQKFSDLAHRWWDPNAEFKPLHDLNPVRLGWIDAHAHLAGKRALDIGCGGGILSESMAGLGAQVKGIDLSTEALGVADLHSLESGITVDYEAIAAEAIAAREPGTYDVVTCMEMLEHVPSPGDIVAACATLVKPGGWVFFSTLNRNLKAYLFAVIGAEYIAQMLPKGTHDYARFIRPSELAGFVRATDLHIVEIKGITYHPIGKRFALSNDTDINYLVACRRGA\n\n### Sequence 2:\nMATGDDIPTFKLVLVGDGGTGKTTFVKRHLTGEFEKKYVATLGVEVHPLIFHTNRGQIRFNVWDTAGQEKFGGLRDGYYIQGQCAIIMFDVTARVTYKNVPNWHRDLVRVCENIPIVLCGNKVDVKDRKVKAKTITFHRKKNLQYYDISAKSNYNFEKPFLWLARKLLGDPNLEFVAMPALAPPEVQMDPTMVAQYEQEIAAAANAELPDDDEDL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00677", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGRIGVLLLNLGGPDQLEDVRPFLYNLFSDPEIIRLPFTWLQKPLAWLISTTRARKSQQNYRLIGGGSPLRRITEEQGKALQAHLASQGQDIQVYIGMRYWHPFTEEAIAAIKQDGITRLVILPLYPQFSISTSGSSFRLLEDLWQRDPQLQAIDYTVIPSWYDRPGYTQAMAELLREELDHFAEPDRVTIFFSAHGVPLSYVTEAGDPYQAEIEGCTALIMQALNRPNPHVLAYQSRVGPVEWLKPYTEEVIPELASQGVNELVVVPISFISEHIETLQEIDMEYRELAEEAGIEHFRRVPALNTHPLFIADLSQLVLEALQGPSRKFDQVVRPQKQVKLYPQERWEWGMTSAAERWNGRLAMLGFLALLLELITGRGPLHLVGLL\n\n### Sequence 2:\nSQTGVLLLNLGGPDRPEDVRPFLYNLFSDPEIIRLPFRWLQKPLAWFISTSRARRSQANYAQIGGGSPLRRITEQQARALKDALEGIGIEANLYIGMRYWHPFTEEAIAQIKADQIRELVILPLYPQFSISTSGSSFRLLESLWNQDPELQKIRYTLIPSWYNHPGYVAAMADLIRQELDRCPNPDEAVIFFSAHGVPKSYVTEAGDPYQEEIEACVRLIMAALNRPNAHVLAYQSRVGPVEWLQPYTEDVILELAAQGVKTLVVVPISFVSEHIETLQEIDIEYREIAAEAGIEVFRRVPALNDHNGFISALAQLVKEALAAPPRTFAEVNQSRKRVKLYPQERWEWGMTSAAERWNGRLAMLGFLALMIELISGQGPLHMLGLL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00678", "category": "Mutation", "instruction": "Describe the effect of mutation Q12094-E111A on protein 18S rRNA aminocarboxypropyltransferase.", "input": "### Wild-type:\nMGKGKNKMHEPKNGRPQRGANGHSSRQNHRRMEMKYDNSEKMKFPVKLAMWDFDHCDPKRCSGKKLERLGLIKSLRVGQKFQGIVVSPNGKGVVCPDDLEIVEQHGASVVECSWARLEEVPFNKIGGKHERLLPYLVAANQVNYGRPWRLNCVEALAACFAIVGRMDWASELLSHFSWGMGFLELNKELLEIYQQCTDCDSVKRAEEEWLQKLEKETQERKSRAKEEDIWMMGNINRRGNGSQSDTSESEENSEQSDLEGNNQCIEYDSLGNAIRIDNMKSREAQSEESEDEESGSKENG\n\n### Mutant:\nMGKGKNKMHEPKNGRPQRGANGHSSRQNHRRMEMKYDNSEKMKFPVKLAMWDFDHCDPKRCSGKKLERLGLIKSLRVGQKFQGIVVSPNGKGVVCPDDLEIVEQHGASVVACSWARLEEVPFNKIGGKHERLLPYLVAANQVNYGRPWRLNCVEALAACFAIVGRMDWASELLSHFSWGMGFLELNKELLEIYQQCTDCDSVKRAEEEWLQKLEKETQERKSRAKEEDIWMMGNINRRGNGSQSDTSESEENSEQSDLEGNNQCIEYDSLGNAIRIDNMKSREAQSEESEDEESGSKENG\n\n### Answer:", "output": "Reduced aminocarboxypropyltransferase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00679", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nRPMNQLYPIDLLTELPPPITDLTLPPPPLVIPPERMLVPSELSNASPDYIRSTLNAVPKNSSLLKKSKLPFGLVIRPYQHLYDDIDPPPLNEDGLIVRCRRCRSYMNPFVTFIEQGRRWRCNFCRLANDVPMQMDQSDPNDPKSRYDRNEIKCAVMEYMAPKEYTLRQPPPATYCFLIDVSQSSIKSGLLATTINTLLQNLDSIPNHDERTRISILCVDNAIHYFKIPLDSENNEESADQINMMDIADLEEPFLPRPNSMVVSLKACRQNIETLLTKIPQIFQSNLITNFALGPALKSAYHLIGGVGGKIIVVSGTLPNLGIGKLQRRNESGVVNTSKETAQLLSCQDSFYKNFTIDCSKVQITVDLFLASEDYMDVASLSNLSRFTAGQTHFYPGFSGKNPNDIVKFSTEFAKHISMDFCMETVMRARGSTGLRMSRFYGHFFNRSSDLCAFSTMPRDQSYLFEVNVDESIMADYCYVQVAVLLSLNNSQRRIRIITLAMPTTESLAEVYASADQLAIASFYNSKAVEKALNSSLDDARVLINKSVQDILATYKKEIVVSNTAGGAPLRLCANLRMFPLLMHSLTKHMAFRSGIVPSDHRASALNNLESLPLKYLIKNIYPDVYSLHDMADEAGLPVQTEDGEATGTIVLPQPINATSSLFERYGLYLIDNGNELFLWMGGDAVPALVFDVFGTQDIFDIPIGKQEIPVVENSEFNQRVRNIINQLRNHDDVITYQSLYIVRGASLSEPVNHASAREVATLRLWASSTLVEDKILNNESYREFLQIMKARISK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCEEETTTSCSCCCSGGGSCCCCCCCCGGGSSSCSSSSSCCTTTEEESBSSEESBHHHHHHHTCCCEEEECSCCSSSTTSSCCCEECSCCCCBCSSSCCBCCTTCEEETTTTEEECTTTCCEEECCGGGTTTSTTCGGGGGGSHHHHCSEEEEECCGGGCSSCCCCCBEEEEEECSHHHHHSCHHHHHHHHHHHTTTTSCCTTSCCEECEEEESSSEEEEECCCCCCCCCCCCCCCEEEEECCCSSCCCSCSSSSSEETTTTHHHHHHHHHHHHHHTSSCCCCCCCHHHHHHHHHHHHTTTCEEEEEEESSCCCSSTTCCCCCCCCCCCCCTTHHHHHTSCSSTHHHHHHHHHHHHTEEEEEEEEESSCCCHHHHHHHHHTTTCCEEEEESCCTTSHHHHHHHHHHHHHHHHSCCEEEEEEEEEECTTEEEEEEESSSBCSSTTEEEEEEECSSCCEEEEEEECSBCCSSEEEEEEEEEEEETTTEEEEEEEEEEEEEESCHHHHHHTBCHHHHHHHHHHHHHHHHHHSCSHHHHHHHHHHHHHHHHHHHHHTCCCTTSSSSCCEEEGGGTTHHHHHHHHHTSTTTCSSCCCHHHHHHHHHHHHHSCHHHHHHHHSCEEEETTTCCSSCSCCCCCCCCCCCCCCCCCCCBCSCSTTCCTTCEEEEECSSEEEEEECTTCCSGGGCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHTTCCSSCCCCCEEEEECCCCCCCCCCCCCCCHHHHHHHHHHHTCTTSSCTTCCCHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00680", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRECF_PSEPF\n\n### Answer:", "output": "Belongs to the RecF family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00681", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "CX3CR1, CSF1R, TMEM119", "output": "This cell is most consistent with microglia, a brain microglia cell. Key markers such as CX3CR1, CSF1R, TMEM119 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "skin", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CX3CR1", "CSF1R", "TMEM119"]}}
{"id": "eval_00682", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nSLC32A1 GAD1 GAD2\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as GAD1, SLC32A1, GAD2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "lymph node", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SLC32A1", "GAD1", "GAD2"]}}
{"id": "eval_00683", "category": "Mutation", "instruction": "Describe the effect of mutation P51787-V541I on protein Potassium voltage-gated channel subfamily KQT member 1.", "input": "### Wild-type:\nMAAASSPPRAERKRWGWGRLPGARRGSAGLAKKCPFSLELAEGGPAGGALYAPIAPGAPGPAPPASPAAPAAPPVASDLGPRPPVSLDPRVSIYSTRRPVLARTHVQGRVYNFLERPTGWKCFVYHFAVFLIVLVCLIFSVLSTIEQYAALATGTLFWMEIVLVVFFGTEYVVRLWSAGCRSKYVGLWGRLRFARKPISIIDLIVVVASMVVLCVGSKGQVFATSAIRGIRFLQILRMLHVDRQGGTWRLLGSVVFIHRQELITTLYIGFLGLIFSSYFVYLAEKDAVNESGRVEFGSYA\n\n### Mutant:\nMAAASSPPRAERKRWGWGRLPGARRGSAGLAKKCPFSLELAEGGPAGGALYAPIAPGAPGPAPPASPAAPAAPPVASDLGPRPPVSLDPRVSIYSTRRPVLARTHVQGRVYNFLERPTGWKCFVYHFAVFLIVLVCLIFSVLSTIEQYAALATGTLFWMEIVLVVFFGTEYVVRLWSAGCRSKYVGLWGRLRFARKPISIIDLIVVVASMVVLCVGSKGQVFATSAIRGIRFLQILRMLHVDRQGGTWRLLGSVVFIHRQELITTLYIGFLGLIFSSYFVYLAEKDAVNESGRVEFGSYA\n\n### Answer:", "output": "In LQT1; unknown pathological significance.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00684", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMVKRKNSDNSGTVAQGNYEAKVAEIEAIISRIESGELELETVFEQFANAVQYLRQCDTFLQQRQQQIDLLIETLNEQDGDQIPRVSNDGIEPKK\n\n### Sequence 2:\nMSRRKREDP------WRYEAAIAEVESLIAQIESGELDLAEVVERFQQAAQTLKRCAEFLEEKRQQVEILIEELNQDDED\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00685", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "O=C1CC(O)C(=O)[OH+][Pt-2]2([NH]CC[NH]2)[OH+]1", "output": "The molecule contains the following functional groups: amine, hydroxyl.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "O=C1CC(O)C(=O)[OH+][Pt-2]2([NH]CC[NH]2)[OH+]1", "compound_name": null, "pubchem_cid": null, "MW": 387.25, "LogP": -3.13, "TPSA": 104.03, "QED": 0.38, "functional_groups": ["amine", "hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amine", "hydroxyl"]}}
{"id": "eval_00686", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMSRPDQAKVTVAKGDIKAIAAALDMYKLDNFAYPSTQQGLEALVKKPTGNPQPKNWNKDGYLKKLPVDPWGNPYQYLAPGTKGPFDLYSLGADGKEGGSDNDADIGNWDN\n\n### Answer:", "output": "<SEQ_3Di>DPDPLLVLQVVVLVFQVQLQVQQVVQCQQQVHGAAQPQRLVCQAPPRPDHTHPHDRHHLNGGDPDDAGPLPWRKDWGADDDPFRTDIWTAANVRDDDDDRRRHTDHDGGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00687", "category": "Mutation", "instruction": "Describe the effect of mutation P16144-C562R on protein Integrin beta-4.", "input": "### Wild-type:\nMAGPRPSPWARLLLAALISVSLSGTLANRCKKAPVKSCTECVRVDKDCAYCTDEMFRDRRCNTQAELLAAGCQRESIVVMESSFQITEETQIDTTLRRSQMSPQGLRVRLRPGEERHFELEVFEPLESPVDLYILMDFSNSMSDDLDNLKKMGQNLARVLSQLTSDYTIGFGKFVDKVSVPQTDMRPEKLKEPWPNSDPPFSFKNVISLTEDVDEFRNKLQGERISGNLDAPEGGFDAILQTAVCTRDIGWRPDSTHLLVFSTESAFHYEADGANVLAGIMSRNDERCHLDTTGTYTQYR\n\n### Mutant:\nMAGPRPSPWARLLLAALISVSLSGTLANRCKKAPVKSCTECVRVDKDCAYCTDEMFRDRRCNTQAELLAAGCQRESIVVMESSFQITEETQIDTTLRRSQMSPQGLRVRLRPGEERHFELEVFEPLESPVDLYILMDFSNSMSDDLDNLKKMGQNLARVLSQLTSDYTIGFGKFVDKVSVPQTDMRPEKLKEPWPNSDPPFSFKNVISLTEDVDEFRNKLQGERISGNLDAPEGGFDAILQTAVCTRDIGWRPDSTHLLVFSTESAFHYEADGANVLAGIMSRNDERCHLDTTGTYTQYR\n\n### Answer:", "output": "In JEB5B. Missense mutation leading to nonlethal phenotype.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00688", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMEMDLNNRLTEDETLEQAYDIFLELAADNLDPADIILFNLQFEERGGAELFDPAEDWQEHIDFDLNPDFFAEVVIGLADTEDGEINDIFARVLLCREKDHKLCHILWRE\n\n### Sequence 2:\nMGLIPKLLAGLVLLTLCVENGSGQYWSYGVRPGGKRNIEPLVDSFQEMAKEIDQLAEPQHFECTLHQPRSPLRDLKGALESLMEEETGQKKI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00689", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nFYPP3_ARATH\n\n### Answer:", "output": "Belongs to the PPP phosphatase family. PP-6 (PP-V) subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00690", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMIRYFTAGESHGPALSAIIEGMPAGVAVSREEIDLQLRRRQQGYGRGGRMKIEQDSAEVLSGIRFGKTIGSPIALVIRNRDWENWTDKMAQFESHEATTEKITIPRPGHADLAGRIKYGFNDIRPVIDRSSARETAARVAAGSLARTFLCQLGIEIGSRISSIGPVTDTGSDKTISALLEKGAECLAGAADRSEVRMLGKKAEEDAIAAIDLAKERGDTLGGIIEIYITGVPVGLGSYVQHDRRLDSQLAAAVMAIQAIKGVEIGPAFENARKPGSEVHDALHLVKGEGVERPTNRSGGIEGSMSSGETIHIRAAMKPISSMQSPLQSFDLATLQPVLSRFERSDTCAVPAAGVVAEAVVAPVIANALFEKFGGDHLEEIRNRLNHYREAVRSAFSG\n\n### Sequence 2:\nMNKFLKEFKDRGFFYQCTGEENLSQLLDKEKIRAYIGFDCTAESLHVGSLLQIMCLRLLQKHGHQPIVLLGGGTTRIGDPSGKDKTRTILSEDEIEKNINNIEKILKNFLDDKDPETKPIFVNNYTWLKNLNYISFLRDVGKHFTINKMLSFDSVKIRLEREQSLSYMEFNYMILQAYDFLELNKKEKCMLQIGGSDQWGNIVNGVDLIKRYSNNHVYGLTTPLITLASGAKMGKTESGAVWLDKKFLSSYDYWQFWRNIDDRDVLKFLKIFTDINVDEIENIKDDNINELKILLANKATSMLHGEDEARKCQETAKQTFSENSLGDNLPTTQINKKMLDDNISILDLVILSKLESSKSEIRRLIKGNGIKINGQAISDEKFLITEDLFKSSLIKLSLGKKKHIKVELI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00691", "category": "Mutation", "instruction": "Describe the effect of mutation P04424-L343F on protein Argininosuccinate lyase.", "input": "### Wild-type:\nMASESGKLWGGRFVGAVDPIMEKFNASIAYDRHLWEVDVQGSKAYSRGLEKAGLLTKAEMDQILHGLDKVAEEWAQGTFKLNSNDEDIHTANERRLKELIGATAGKLHTGRSRNDQVVTDLRLWMRQTCSTLSGLLWELIRTMVDRAEAERDVLFPGYTHLQRAQPIRWSHWILSHAVALTRDSERLLEVRKRINVLPLGSGAIAGNPLGVDRELLRAELNFGAITLNSMDATSERDFVAEFLFWASLCMTHLSRMAEDLILYCTKEFSFVQLSDAYSTGSSLMPQKKNPDSLELIRSKA\n\n### Mutant:\nMASESGKLWGGRFVGAVDPIMEKFNASIAYDRHLWEVDVQGSKAYSRGLEKAGLLTKAEMDQILHGLDKVAEEWAQGTFKLNSNDEDIHTANERRLKELIGATAGKLHTGRSRNDQVVTDLRLWMRQTCSTLSGLLWELIRTMVDRAEAERDVLFPGYTHLQRAQPIRWSHWILSHAVALTRDSERLLEVRKRINVLPLGSGAIAGNPLGVDRELLRAELNFGAITLNSMDATSERDFVAEFLFWASLCMTHLSRMAEDLILYCTKEFSFVQLSDAYSTGSSLMPQKKNPDSLELIRSKA\n\n### Answer:", "output": "In ARGINSA. The mutation in the ASL gene causes a functional change in the argininosuccinate lyase enzyme.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00692", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nNCAP_I96A3\n\n### Answer:", "output": "Belongs to the influenza viruses nucleoprotein family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00693", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSPNSQYLKTRILDIYTPEQRAGIEKSEDWRQFSRRMDTHFPKLMNELDSVYGNNEALLPMLEMLLAQAWQSYSQRNSSLKDIDIARENNPDWILSNKQVGGVCYVDLFAGDLKGLKDKIPYFQELGLTYLHLMPLFKCPEGKSDGGYAVSSYRDVNPALGTIGDLREVIAALHEAGISAVVDFIFNHTSNEHEWAQRCAAGDPLFDNFYYIFPDRRMPDQYDRTLREIFPDQHPGGFSQLEDGRWVWTTFNSFQWDLNYSNPWVFRAMAGEMLFLANLGVDILRMDAVAKIWKQMGTSCENLPQAHALIRAFNAVMRIAAPAVFFKSQAIVHPDQVVQYIGQDECQIGYNPLQMALLWNTLATREVNLLHQALTYRHNLPEHTAWVNYVRSHDDIGWTFADEDAAYLGISGYDHRQFLNRFFVNRFDGSFARGVPFQYNPSTGDCRVSGTAAALVGLAQDDPHAVDRIKLLYSIALSTGGLPLIYLGDEVGTLNDDDWSQDSNKSDDSRWAHRPRYNEALYAQRNDPSTAAGQIYQDLRHMIAVRQSNPRFDGGRLVTFNTNNKHIIGYIRNNALLAFGNFSEYPQTVTAHTLQAMPFKAHDLIGGKTVSLNQDLTLQPYQVMWLEIA\n\n### Answer:", "output": "<SEQ_3Di>DDALVRLVVLLVVVDDPVVSVVQCPDVLNVQLSVLRNVQVVLLVVLVCVQQPPDPVVVVLVSVLVSLLSVLLVPFDVVLVVLLVVCVVPLCPQQFLLAAEEEAECCQEQFALVSVLVCVVVCVVLRHLEYEYEQFADADPQAGPRNLQHQDLQFGDPSRPTLVSLLVSCVSSVVSNHAYEYALSLFWHFCNHPLNVCCVVPPVVSPQQFDKAQACPVVVQLQVQEDQDAVLADRGQWDADPVGMITGHNDHNRIGTTAVSDSVSLSVSLSSVSVVVSSVHQEYEHPQLQQRDDDGPHNSGLDPSSQSSQLSSLSSCSSRRSRYAYEYDDDAALVSQQVQDDSSGGQEYEDLQLQLVLLLCLQVLFCVSVQLCLAPRFDHDPSHFYAFENDDLDWRFNRYDQVSVVVVVDHPLVSVVSSVCLQCVVDPPRLANWHWHRQDPVPRGITIAWALLRGQNPVVPDVCSLLSLLLRQLCSSFQTGHYYHHPVVLLRDTFPPCQLVSVSRVNGRSNRHRHGHDPVSVVCCCPCVDSNVVSNVSSSLSSVCSSPPCLRGHRDKHWADQPPRNWTWIGGPVFKIKIWGSAQAKDKRALVSLVVFDQWWQWSRPRDIDGNNGIDIDHGGDMTITGRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00694", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSASASLPVTRAAAAPRITVLFSTEKPNANTNPYLTQLYDALPDAVQPRFFSMREALLSRYDVLHLHWPEYLLRHPSKMGTLAKQACAALLLMKLQLTGTPVVRTLHNLAPHEDRGWRERALLRWIDQLTRRWIRINATTPVRPPFTDTILHGHYRDWFATMEQSTTLPGRLLHFGLIRPYKGVEVLLDVMRDVQDPRLSLRIVGNPATPXMRTLVETACAQDARISALLAYVEEPVLAREVSACELVVLPYKQMHNSGTLLLALSLARPVLAPWSESNAAIADEVGPGWVFLYEGEFDAALLSGMLDQVRAAPRGPAPDLSQRDWPRIGQLHYRTYLEALGKDGDAAL\n\n### Sequence 2:\nMAELQEVQITEEKPLLPGQTPEAAKEAELAARILLDQGQTHSVETPYGSVTFTVYGTPKPKRPAILTYHDVGLNYKSCFQPLFQFEDMQEIIQNFVRVHVDAPGMEEGAPVFPLGYQYPSLDQLADMIPCVLQYLNFSTIIGVGVGAGAYILARYALNHPDTVEGLVLINIDPNAKGWMDWAAHKLTGLTSSIPEMILGHLFSQEELSGNSELIQKYRNIITHAPNLDNIELYWNSYNNRRDLNFERGGDITLRCPVMLVVGDQAPHEDAVVECNSKLDPTQTSFLKMADSGGQPQLTQPGKLTEAFKYFLQGMGYMASSCMTRLSRSRTASLTSAASVDGNRSRSRTLSQSSESGTLSSGPPGHTMEVSC\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00695", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCASK_BOVIN\n\n### Answer:", "output": "Kappa-casein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00696", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKAKKETTDRFPTWWLFYYVLRKAYFFLGIPFFLFCALGFTEMLCSDRYFGNKVEDYVVTFGSWFLLLAPGIWMYSRAKTRREKIRKVVQTIKESGFYSPEKGYEGLSLTQGAYFGIDLKNGTMLYVRIYPGNIMDVIGFDIHNFTRTVTDDKKLEIHTKYINLPMVPVPSWCAHPETASNTMHAMASRGYDYPVDFPRLIQEKRKEWEQIAGMPVAEVF\n\n### Sequence 2:\nMKAKKETTDRFPTWWLFYYVLRKAYFFLGIPFFLFCALGFTEMLCSDRYFGNKVEDYVVTFGSWFLLLAPGIWMYSRAKTRREKIRKVVQTIKESGFYSPEKGYEGLSLTQGAYFGIDLKNGTMLYVRIYPGNIMDVIGFDIHNFTRTVTDDKKLEIHTKYINLPMVPVPSWCAHPETASNTMHAMASRGYDYPVDFPRLIQEKRKEWEQIAGMPVAEVF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00697", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSMLVVVTENVPPRLRGRLAIWLLEVRAGVYVGDVSAKIREMIWEQIAGLAEEGNVVMAWATNTETGFEFQTFGLNRRTPVDLDGLRLVSFLPVGSSENLYFQ\n\n### Answer:", "output": "<SEQ_2D>CEEEEEEEESCCHHHHHHHHHHSEEEETTEEEEEECHHHHHHHHHHHHHHCTTSEEEEEEECSSSSSEEEEEESCCSEEEEEETTEEEEEECCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00698", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nARSC_ECOLI\n\n### Answer:", "output": "Involved in resistance to arsenate.  Catalyzes the reduction of arsenate [As(V)] to arsenite [As(III)].", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00699", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: Brc1ccccc1Br\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "ESOL", "canonical_smiles": "Brc1ccccc1Br", "compound_name": null, "pubchem_cid": null, "MW": 235.91, "LogP": 3.21, "TPSA": 0.0, "QED": 0.65, "functional_groups": ["aromatic ring", "halide"], "moleculenet_labels": {"label": -3.5}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00700", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "COL1A1, COL3A1, COL1A2", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as COL1A2, COL1A1, COL3A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "spleen", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["COL1A1", "COL3A1", "COL1A2"]}}
{"id": "eval_00701", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nUTS2B_MOUSE\n\n### Answer:", "output": "Belongs to the urotensin-2 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00702", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMARIAGVDLPRGKRVDIALTYIYGIGRATALQILDATGVNWTRNVDDLNADEVNEIRKEIEQNHKVEGDLRREISANIKRLMDIGCYRGLRHRRGLPVRGQRTHTNARTRKGPRRGTVGKKK\n\n### Sequence 2:\nMARIAGVDLPRGKRVDIALTYIYGIGRTTAMKILDTTGVNWERSIDDLSADEVNEIRKELEQHYKVEGDLRREVSSNIKRLMDIGCYRGLRHRRGLPVHGQRTHTNARTRKGPRRGAVGKKK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00703", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY767_MYCTU\n\n### Answer:", "output": "Uncharacterized HTH-type transcriptional regulator Rv0767c.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00704", "category": "Mutation", "instruction": "Describe the effect of mutation P0A769-E154Q on protein Divalent metal cation transporter MntH.", "input": "### Wild-type:\nMTNYRVESSSGRAARKMRLALMGPAFIAAIGYIDPGNFATNIQAGASFGYQLLWVVVWANLMAMLIQILSAKLGIATGKNLAEQIRDHYPRPVVWFYWVQAEIIAMATDLAEFIGAAIGFKLILGVSLLQGAVLTGIATFLILMLQRRGQKPLEKVIGGLLLFVAAAYIVELIFSQPNLAQLGKGMVIPSLPTSEAVFLAAGVLGATIMPHVIYLHSSLTQHLHGGSRQQRYSATKWDVAIAMTIAGFVNLAMMATAAAAFHFSGHTGVADLDEAYLTLQPLLSHAAATVFGLSLVAAGL\n\n### Mutant:\nMTNYRVESSSGRAARKMRLALMGPAFIAAIGYIDPGNFATNIQAGASFGYQLLWVVVWANLMAMLIQILSAKLGIATGKNLAEQIRDHYPRPVVWFYWVQAEIIAMATDLAEFIGAAIGFKLILGVSLLQGAVLTGIATFLILMLQRRGQKPLQKVIGGLLLFVAAAYIVELIFSQPNLAQLGKGMVIPSLPTSEAVFLAAGVLGATIMPHVIYLHSSLTQHLHGGSRQQRYSATKWDVAIAMTIAGFVNLAMMATAAAAFHFSGHTGVADLDEAYLTLQPLLSHAAATVFGLSLVAAGL\n\n### Answer:", "output": "Transports manganese to a level half of that of wild-type.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00705", "category": "Mutation", "instruction": "Describe the effect of mutation P07604-V173H on protein HTH-type transcriptional regulatory protein TyrR.", "input": "### Wild-type:\nMRLEVFCEDRLGLTRELLDLLVLRGIDLRGIEIDPIGRIYLNFAELEFESFSSLMAEIRRIAGVTDVRTVPWMPSEREHLALSALLEALPEPVLSVDMKSKVDMANPASCQLFGQKLDRLRNHTAAQLINGFNFLRWLESEPQDSHNEHVVINGQNFLMEITPVYLQDENDQVVLTGAVVMLRSTIRMGRQLQNVAAQDVSAFSQIVAVSPKMKHVVEQAQKLAMLSAPLLITGDTGTGKDLFAYACHQASPRAGKPYLALNCASIPEDAVESELFGHAPEGKKGFFEQANGGSVLLDEI\n\n### Mutant:\nMRLEVFCEDRLGLTRELLDLLVLRGIDLRGIEIDPIGRIYLNFAELEFESFSSLMAEIRRIAGVTDVRTVPWMPSEREHLALSALLEALPEPVLSVDMKSKVDMANPASCQLFGQKLDRLRNHTAAQLINGFNFLRWLESEPQDSHNEHVVINGQNFLMEITPVYLQDENDQHVLTGAVVMLRSTIRMGRQLQNVAAQDVSAFSQIVAVSPKMKHVVEQAQKLAMLSAPLLITGDTGTGKDLFAYACHQASPRAGKPYLALNCASIPEDAVESELFGHAPEGKKGFFEQANGGSVLLDEI\n\n### Answer:", "output": "Increases the activation function to 73% of the wild-type value. Does not affect repression of aroF.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00706", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BRAF\nLigand SMILES: O=C(c1ccccc1)C(O)(CN1CCCCC1)c1ccccc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and B-Raf kinase (BRAF) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "O=C(c1ccccc1)C(O)(CN1CCCCC1)c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 309.41, "LogP": 3.24, "TPSA": 40.54, "QED": 0.86, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "ketone"], "moleculenet_labels": {"label": 3.41}, "split": "train", "generation_method": "template_protein_ligand", "target": "BRAF", "has_evidence": false}}
{"id": "eval_00707", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD206 SPP1 CD163 MRC1 APOE\n</cell>", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as SPP1, CD206, MRC1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "liver", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD206", "SPP1", "CD163", "MRC1", "APOE"]}}
{"id": "eval_00708", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "C=C1CC2(CCS(=O)(=O)c3ccccc32)OC1=O", "output": "The SMILES represents an unnamed molecule with MW=264.3, LogP=1.56, TPSA=60.44. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "C=C1CC2(CCS(=O)(=O)c3ccccc32)OC1=O", "compound_name": null, "pubchem_cid": null, "MW": 264.3, "LogP": 1.56, "TPSA": 60.44, "QED": 0.53, "functional_groups": ["aromatic ring", "ester", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00709", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nASSY_LACPL\n\n### Answer:", "output": "Belongs to the argininosuccinate synthase family. Type 1 subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00710", "category": "Mutation", "instruction": "Describe the effect of mutation P18910-H213D on protein Atrial natriuretic peptide receptor 1.", "input": "### Wild-type:\nMPGSRRVRPRLRALLLLPPLLLLRGGHASDLTVAVVLPLTNTSYPWSWARVGPAVELALARVKARPDLLPGWTVRMVLGSSENAAGVCSDTAAPLAAVDLKWEHSPAVFLGPGCVYSAAPVGRFTAHWRVPLLTAGAPALGIGVKDEYALTTRTGPSHVKLGDFVTALHRRLGWEHQALVLYADRLGDDRPCFFIVEGLYMRVRERLNITVNHQEFVEGDPDHYPKLLRAVRRKGRVIYICSSPDAFRNLMLLALNAGLTGEDYVFFHLDVFGQSLKSAQGLVPQKPWERGDGQDRSARQ\n\n### Mutant:\nMPGSRRVRPRLRALLLLPPLLLLRGGHASDLTVAVVLPLTNTSYPWSWARVGPAVELALARVKARPDLLPGWTVRMVLGSSENAAGVCSDTAAPLAAVDLKWEHSPAVFLGPGCVYSAAPVGRFTAHWRVPLLTAGAPALGIGVKDEYALTTRTGPSHVKLGDFVTALHRRLGWEHQALVLYADRLGDDRPCFFIVEGLYMRVRERLNITVNDQEFVEGDPDHYPKLLRAVRRKGRVIYICSSPDAFRNLMLLALNAGLTGEDYVFFHLDVFGQSLKSAQGLVPQKPWERGDGQDRSARQ\n\n### Answer:", "output": "Abolishes hormone binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00711", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nOGT1_CAEEL\n\n### Answer:", "output": "Addition of nucleotide-activated sugars directly onto the polypeptide through O-glycosidic linkage with the hydroxyl of serine or threonine.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00712", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "CD79A, IGHG1, IGHA1, CD27", "output": "This profile is most consistent with memory B cell, which is commonly found in lung under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "lung", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CD79A", "IGHG1", "IGHA1", "CD27"]}}
{"id": "eval_00713", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPUR5_PROM0\n\n### Answer:", "output": "Belongs to the AIR synthase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00714", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "COc1ccccc1O", "output": "The molecule contains the following functional groups: aromatic ring, hydroxyl, ether, phenol.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "COc1ccccc1O", "compound_name": null, "pubchem_cid": null, "MW": 124.14, "LogP": 1.4, "TPSA": 29.46, "QED": 0.61, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "hydroxyl", "ether", "phenol"]}}
{"id": "eval_00715", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMFEQRVNSDVLTVSTVNSQDQVTQKPLRDSVKQALKNYFAQLNGQDVNDLYELVLAEVEQPLLDMVMQYTRGNQTRAALMMGINRGTLRKKLKKYGMN\n\n### Sequence 2:\nMAKKITATNPRLVKLIEILKQESYKNQAKIWKDIARRLAKPRRRRAEVNLSKINRYTKEGDVVLVPGKVLGAGKLEHKVVVAAFAFSETAKKLIKEAGGEAITIEELIKRNPKGSNVKIMA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00716", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMRIILASSSPRRRQLMELLGIEFEVEKPDVEEEFLESPEETVRELSLRKAEWVFKKRKEEEILVIGSDTVVVLDGNILGKPESLEEAKGMLKKLSGKWHVVYTGVAFVSSETKDVIVSSTKVRFRELPESVIDYYVEKYRPLDKAGAYGIQDFAAVFVEKIEGDFFTVVGFPLGMVWQYLYEKGWWKVASKREDDKGGARVAFG\n\n### Sequence 2:\nQLILASSSPRRKQLLELANLRFQILASHIDEQIHETKSPEQAVQLLAYRKAKAVADHYPHS--YVIGADTVVVYQDNILGKPKTEEEAAAMLRMLSGNEHHVLTGVAILSPNGQSLFVEKTKVFFWDLTEEEIFDYIATGEPMDKAGAYGIQGRAALFVKRIEGDYFNVVGLPLSRTVREVKRLGW\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00717", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD38 MZB1 XBP1 SDC1 JCHAIN\n</cell>", "output": "This cell is most consistent with plasma cell, an antibody-secreting plasma cell. Key markers such as CD38, JCHAIN, XBP1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "plasma cell", "tissue": "prostate", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD38", "MZB1", "XBP1", "SDC1", "JCHAIN"]}}
{"id": "eval_00718", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nSFTPC:low ABCA3:high SFTPA1:high SFTPB:low\n</cell>", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as ABCA3, SFTPA1, SFTPC support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "ovary", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SFTPC", "ABCA3", "SFTPA1", "SFTPB"]}}
{"id": "eval_00719", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: CSc1c(Cl)nc(NC(C)C)nc1N1CCN(C)CC1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "CSc1c(Cl)nc(NC(C)C)nc1N1CCN(C)CC1", "compound_name": null, "pubchem_cid": null, "MW": 315.87, "LogP": 2.42, "TPSA": 44.29, "QED": 0.68, "functional_groups": ["amine", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00720", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL6_STREM\n\n### Answer:", "output": "This protein binds to the 23S rRNA, and is important in its secondary structure.  It is located near the subunit interface in the base of the L7/L12 stalk, and near the tRNA binding site of the peptidyltransferase center.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00721", "category": "Mutation", "instruction": "Describe the effect of mutation Q8WVY7-D143A on protein Ubiquitin-like domain-containing CTD phosphatase 1.", "input": "### Wild-type:\nMALPIIVKWGGQEYSVTTLSEDDTVLDLKQFLKTLTGVLPERQKLLGLKVKGKPAENDVKLGALKLKPNTKIMMMGTREESLEDVLGPPPDNDDVVNDFDIEDEVVEVENREENLLKISRRVKEYKVEILNPPREGKKLLVLDVDYTLFDHRSCAETGVELMRPYLHEFLTSAYEDYDIVIWSATNMKWIEAKMKELGVSTNANYKITFMLDSAAMITVHTPRRGLIDVKPLGVIWGKFSEFYSKKNTIMFDDIGRNFLMNPQNGLKIRPFMKAHLNRDKDKELLKLTQYLKEIAKLDDF\n\n### Mutant:\nMALPIIVKWGGQEYSVTTLSEDDTVLDLKQFLKTLTGVLPERQKLLGLKVKGKPAENDVKLGALKLKPNTKIMMMGTREESLEDVLGPPPDNDDVVNDFDIEDEVVEVENREENLLKISRRVKEYKVEILNPPREGKKLLVLAVDYTLFDHRSCAETGVELMRPYLHEFLTSAYEDYDIVIWSATNMKWIEAKMKELGVSTNANYKITFMLDSAAMITVHTPRRGLIDVKPLGVIWGKFSEFYSKKNTIMFDDIGRNFLMNPQNGLKIRPFMKAHLNRDKDKELLKLTQYLKEIAKLDDF\n\n### Answer:", "output": "Loss of catalytic activity. No effect on interaction with 19S regulatory particle but loss of catalytic activity; when associated with A-145. The mutation in UBLCP1 disrupts its binding to the 19S regulatory particle (RP) and impairs its ability to dephosphorylate the subunit Rpt1, leading to a disruption in 26S proteasome assembly.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00722", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMMKRLIVLVLLASTLLTGCNTARGFGEDIKHLGNSISRAAS\n\n### Sequence 2:\nMAKKSGSGLQSSAGLMRYYEADKNAIHIQPKTVLIAGALVGIAVIFLSAVNGFWP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00723", "category": "Mutation", "instruction": "Describe the effect of mutation P02649-L46P on protein Apolipoprotein E.", "input": "### Wild-type:\nMKVLWAALLVTFLAGCQAKVEQAVETEPEPELRQQTEWQSGQRWELALGRFWDYLRWVQTLSEQVQEELLSSQVTQELRALMDETMKELKAYKSELEEQLTPVAEETRARLSKELQAAQARLGADMEDVCGRLVQYRGEVQAMLGQSTEELRVRLASHLRKLRKRLLRDADDLQKRLAVYQAGAREGAERGLSAIRERLGPLVEQGRVRAATVGSLAGQPLQERAQAWGERLRARMEEMGSRTRDRLDEVKEQVAEVRAKLEEQAQQIRLQAEAFQARLKSWFEPLVEDMQRQWAGLVEK\n\n### Mutant:\nMKVLWAALLVTFLAGCQAKVEQAVETEPEPELRQQTEWQSGQRWEPALGRFWDYLRWVQTLSEQVQEELLSSQVTQELRALMDETMKELKAYKSELEEQLTPVAEETRARLSKELQAAQARLGADMEDVCGRLVQYRGEVQAMLGQSTEELRVRLASHLRKLRKRLLRDADDLQKRLAVYQAGAREGAERGLSAIRERLGPLVEQGRVRAATVGSLAGQPLQERAQAWGERLRARMEEMGSRTRDRLDEVKEQVAEVRAKLEEQAQQIRLQAEAFQARLKSWFEPLVEDMQRQWAGLVEK\n\n### Answer:", "output": "Found in a patient with hypercholesterolemia; unknown pathological significance; ApoE4 Freiburg. The large OLA sample clearly showed that variation in many, but not all, of OLA-typed SNPs is significantly correlated with the classical protein-coding variants, implying that there may be important substructure within the classical epsilon 2, epsilon 3, and epsilon 4 alleles. The mutation in the LDLR gene is associated with autosomal dominant hypercholesterolemia (ADH). This mutation leads to a phenotypic change characterized by isolated high-cholesterol levels.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00724", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSGGSGKVVKFSYMWTINNFSFCREEMGEVIKSSTFSSGANDKLKWCLRVNPKGLDEESKDYLSLYLLLVSCPKSEVRAKFKFSILNAKGEETKAMESQRAYRFVQGKDWGFKKFIRRDFLLDEANGLLPDDKLTLFCEVSVVQD\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCEEEEEEEEECSCCSCHHHHCSCEECCCBCSSTTCCCCEEEEEESSCSSGGGTTEEEEEEEECSCCCCCEEEEEEEEEECTTSCEEEEEECSSCEEECTTCEEEEEEEEEHHHHHCGGGCCSGGGCEEEEEEEEECCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00725", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nKLF1 GATA1 HBA1 HBB\n</cell>", "output": "This cell is most consistent with erythroid progenitor, an erythroid progenitor cell. Key markers such as HBB, GATA1, KLF1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "brain", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["KLF1", "GATA1", "HBA1", "HBB"]}}
{"id": "eval_00726", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "COc1ccc2c3c1OC1CC(O)C=CC31CCN(C)C2", "output": "The molecule contains the following functional groups: aromatic ring, amine, hydroxyl, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "SIDER", "canonical_smiles": "COc1ccc2c3c1OC1CC(O)C=CC31CCN(C)C2", "compound_name": null, "pubchem_cid": null, "MW": 287.36, "LogP": 1.85, "TPSA": 41.93, "QED": 0.8, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "ether"], "moleculenet_labels": {"Hepatobiliary disorders": 1, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 0, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 1, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amine", "hydroxyl", "ether"]}}
{"id": "eval_00727", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAGHSKFKNIMYRKGAQDKKRSALFSKLSREVTVAAKSGLPDPNANPRLRAAVSAARSGGMPKDNIERAIAKAIGGDGDNYEELRYEGFGPSGVSLIIETLTDNRNRTATNVRTALSKNGGNLGTSGSVTHGFDRMGLITYKAEAGDPDKIFEAALEAGAEDVTSSEDEHEIWTAQADLHEVAGKLEAILGEPEGVKLAWRPQTLINVDEESAGTLLRLIETLEDDDDVQTVWANYDIDDEVMERLGQE\n\n### Sequence 2:\nMLTLIDPNDERQAFPPPDAALKEPNGLLAVGGSLSVARLERAYRRGIFPWYGEGDPILWWSPDPRLVLFPEKLRISRSLRKTLRRRLFRFSFDRAFRAVITACAERREKSEGTWLTADMQSAYLAFHHAGFAHSFEAWQDGALVGGLYGVAMGRIFYGESMFHRVTDASKAALAFAVGCLSHWGYRMIDCQVYSSHLVSLGASTIPRSEFQALVSEYSETSVDPEAWKHAPGDFL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00728", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: PARP1\nLigand SMILES: CC1OC(OC2C(O)COC(OC3C(C)OC(OC4C(OC(=O)C56CCC(C)(C)CC5C5=CCC7C8(C)CC(O)C(OC9OC(CO)C(O)C(OC%10OC(CO)C(O)C(O)C%10O)C9O)C(C)(CO)C8CCC7(C)C5(C)CC6)OCC(O)C4O)C(O)C3O)C2O)C(O)C(O)C1O\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and poly(ADP-ribose) polymerase 1 (PARP1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "CC1OC(OC2C(O)COC(OC3C(C)OC(OC4C(OC(=O)C56CCC(C)(C)CC5C5=CCC7C8(C)CC(O)C(OC9OC(CO)C(O)C(OC%10OC(CO)C(O)C(O)C%10O)C9O)C(C)(CO)C8CCC7(C)C5(C)CC6)OCC(O)C4O)C(O)C3O)C2O)C(O)C(O)C1O", "compound_name": null, "pubchem_cid": null, "MW": 1369.51, "LogP": -5.11, "TPSA": 491.97, "QED": 0.04, "functional_groups": ["carboxylic acid", "ester", "hydroxyl", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "PARP1", "has_evidence": false}}
{"id": "eval_00729", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: CDK2\nLigand SMILES: COC(=O)c1ccc(C)c(NS(=O)(=O)c2ccc3c(c2)c(=O)sn3C)c1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and cyclin-dependent kinase 2 (CDK2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "COC(=O)c1ccc(C)c(NS(=O)(=O)c2ccc3c(c2)c(=O)sn3C)c1", "compound_name": null, "pubchem_cid": null, "MW": 392.46, "LogP": 2.5, "TPSA": 94.47, "QED": 0.69, "functional_groups": ["aromatic ring", "ester", "ether", "sulfonamide"], "moleculenet_labels": {"label": 2.6}, "split": "train", "generation_method": "template_protein_ligand", "target": "CDK2", "has_evidence": false}}
{"id": "eval_00730", "category": "Mutation", "instruction": "Describe the effect of mutation Q05323-Q229R on protein Transcriptional activator VP30.", "input": "### Wild-type:\nMEASYERGRPRAARQHSRDGHDHHVRARSSSRENYRGEYRQSRSASQVRVPTVFHKKRVEPLTVPPAPKDICPTLKKGFLCDSSFCKKDHQLESLTDRELLLLIARKTCGSVEQQLNITAPKDSRLANPTADDFQQEEGPKITLLTLIKTAEHWARQDIRTIEDSKLRALLTLCAVMTRKFSKSQLSLLCETHLRREGLGQDQAEPVLEVYQRLHSDKGGSFEAALWQQWDRQSLIMFITAFLNIALQLPCESSAVVVSGLRTLVPQSDNEEASTNPGTCSWSDEGTP\n\n### Mutant:\nMEASYERGRPRAARQHSRDGHDHHVRARSSSRENYRGEYRQSRSASQVRVPTVFHKKRVEPLTVPPAPKDICPTLKKGFLCDSSFCKKDHQLESLTDRELLLLIARKTCGSVEQQLNITAPKDSRLANPTADDFQQEEGPKITLLTLIKTAEHWARQDIRTIEDSKLRALLTLCAVMTRKFSKSQLSLLCETHLRREGLGQDQAEPVLEVYQRLHSDKGGSFEAALWQRWDRQSLIMFITAFLNIALQLPCESSAVVVSGLRTLVPQSDNEEASTNPGTCSWSDEGTP\n\n### Answer:", "output": "Complete loss of interaction with nucleoprotein/NP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00731", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSNVVAVEFDTYLNPDYGDPNYIHIGIDVNSIRSKVTAKWDWQNGKIATAHISYNSVSKRLSVTSYYAGSKPATLSYDIELHTVLPEWVRVGLSASTGQDKERNTVHSWSFTSSLWTN\n\n### Answer:", "output": "<SEQ_3Di>DFDWDWAFDDDDDVVQQRDPAGWIAIDGPGPNGPDIWHDDDDPPFDKDKDWDDDQVVQKIKMWMDGPPDHIIIDMDGHNVCVRDPPDDDDDDDDDDDPDDDDDDDDDDDDDDDDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00732", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTYW31_PYRAB\n\n### Answer:", "output": "S-adenosyl-L-methionine-dependent methyltransferase that acts as a component of the wyosine derivatives biosynthesis pathway.  Probably methylates N-4 position of wybutosine-86 to produce wybutosine-72.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00733", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nMTND_XYLFM\n\n### Answer:", "output": "Belongs to the acireductone dioxygenase (ARD) family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00734", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: HIV protease\nLigand SMILES: CC(C)(C)Cl\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CC(C)(C)Cl", "compound_name": null, "pubchem_cid": null, "MW": 92.57, "LogP": 2.02, "TPSA": 0.0, "QED": 0.4, "functional_groups": ["halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00735", "category": "Mutation", "instruction": "Describe the effect of mutation Q5AQ33-S576E on protein Heat shock transcription factor.", "input": "### Wild-type:\nMIMNMTTDYRDPLLDLFGTESNSGNETSSPSDIPVINRSGTFNQQQFSPLLTQQSLYNTPNSGSTPNIFDPNYTQMQEEQTSPSSNKLQPEDPPRKKRNTRSQTKIHQQSEGDEYNSNDYKDSIDLDKPPVVEPSPPFFVESDTTPEFVIPTPTSEQQQQQHHELIAQDYQRSNNSNQFGNLTHYEPNLPPLPPLSESILPQTNTFHPLVLPHDPRHAITAGPANNSQQQQQQQQQDSSIPSDGISSKIQQLHAPSLSNNQSASQRKKKESSGPKTRPAFVMKIWSMVNDPANHEYIRWN\n\n### Mutant:\nMIMNMTTDYRDPLLDLFGTESNSGNETSSPSDIPVINRSGTFNQQQFSPLLTQQSLYNTPNSGSTPNIFDPNYTQMQEEQTSPSSNKLQPEDPPRKKRNTRSQTKIHQQSEGDEYNSNDYKDSIDLDKPPVVEPSPPFFVESDTTPEFVIPTPTSEQQQQQHHELIAQDYQRSNNSNQFGNLTHYEPNLPPLPPLSESILPQTNTFHPLVLPHDPRHAITAGPANNSQQQQQQQQQDSSIPSDGISSKIQQLHAPSLSNNQSASQRKKKESSGPKTRPAFVMKIWSMVNDPANHEYIRWN\n\n### Answer:", "output": "Mimics phosphorylation and increases thermotolerance; when associated with E-570, E-556 and E-577. This mutation affects the activation of Hsf1 and its ability to activate HSE-containing genes in response to heat shock. The mutant is thermosensitive and unable to contribute to the thermal adaptation and virulence of Candida albicans.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00736", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSANELTSGDFTESGEPFKLFAEWLKEAEASEPNDPNAVALATVDEHGLPNVRMVLLKGFDDDGFVFYTNFESQKGREILGQKKAAMCFHWKSLRRQVRLRGPVEIVSDAEADAYFKTRARGSRIGAWASKQSRPLESRFALEKAVAEYTARYAIGEIPRPPHWSGFRIRPTSIEFWKDQQFRLHDRIEFRRPSPVGAWEKVRMYP\n\n### Sequence 2:\nMSANELTSGDFTESGEPFKLFAEWLKEAEASEPNDPNAVALATVDEHGLPNVRMVLLKGFDDDGFVFYTNFESQKGREILGQKKAAMCFHWKSLRRQVRLRGPVEIVSDAEADAYFKTRARGSRIGAWASKQSRPLESRFALEKAVAEYTARYAIGEIPRPPHWSGFRIRPTSIEFWKDQQFRLHDRIEFRRPSPVGAWEKVRMYP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00737", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "TOX, CD8A, PDCD1", "output": "This cell is most consistent with exhausted CD8 T cell, an exhausted CD8+ T cell. Key markers such as TOX, CD8A, PDCD1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "prostate", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TOX", "CD8A", "PDCD1"]}}
{"id": "eval_00738", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL1_CALBD\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uL1 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00739", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKKHLLPLALLFSGISPAQALDVGDISSFMNSDSSTLSKTIQNSTDSGRLINIRLERLSSPLDDGQVIAMDKPDELLLTPASLLLPAQASEVIRFFYKGPADEKERYYRIVWFDQALSDAQRDNANRSAVATASARIGTILVVAPRQANYHFQYANGSLTNTGNATLRILAYGPCLKAANGKECKENYYLMPGKSRRFTRVDTADNKGRVALWQGDKFIPVK\n\n### Sequence 2:\nMVLFTLEDFAVFKATTLATRMHLIREQLDPKFAEAATVIVPLLQTDQQQAIYSHIAKHQRRYRNPPPNTWVAFSTSSRGYKMVPHLALGFWDDRLFLWLSVLRESKPASRVLTGITAMATTLPGKWQVAGEHTDKAMLPLTSANLATVGARFQTIKKAEFLLGKVYLADDPIWADPVRLWQDIQQRVVALKPMFDQLVQNV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00740", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAKKVKGNRVQVVLECAEHKESGLPGTSRYITTKNKKNTTTRLELKKYNPILKKMTLHKEIK\n\n### Sequence 2:\nMAKKAKGNRVQVILECTEHKESGMPGTSRYITTKNRKNTTERLELKKYNPILKRVTVHKEIK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00741", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nITP_HHV8P\n\n### Answer:", "output": "Plays an essential role in cytoplasmic secondary envelopment during viral egress.  Interacts with the capsid via the large tegument protein/LTP and participates in its transport to the host trans-Golgi network (TGN) where secondary envelopment occurs.  Modulates tegumentation and capsid accumulation at the viral assembly complex.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00742", "category": "Mutation", "instruction": "Describe the effect of mutation A1X8D8-S25C on protein Alpha-conotoxin-like Pu1.2.", "input": "### Wild-type:\nLDGRNAAADFETSDLLAMTIRGGCSSYPPCIANNPLCGGKR\n\n### Mutant:\nLDGRNAAADFETSDLLAMTIRGGCCSYPPCIANNPLCGGKR\n\n### Answer:", "output": "No change in [C4S]Pu1.2(1-9); No gain of HVA calcium currents inhibition in rat DRG neurons (this effect could be facilitated by the selective GABA(B) receptor antagonist CGP55845). No facilitation of human alpha-7/CHRNA7 and alpha-9-alpha-10/CHRNA9-CHRNA10 AChR. The mutation in α-conotoxins has been identified to demodulate voltage-gated calcium channels by acting as a G protein-uncoupled GABA(B) receptor (GABA(B)R) agonist.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00743", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CC(C)(C)NS(=O)(=O)c1cncc(-c2ccc3nc(N)nn3c2)c1", "output": "The molecule contains the following functional groups: amine, sulfonamide, pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "CC(C)(C)NS(=O)(=O)c1cncc(-c2ccc3nc(N)nn3c2)c1", "compound_name": null, "pubchem_cid": null, "MW": 346.42, "LogP": 1.45, "TPSA": 115.27, "QED": 0.74, "functional_groups": ["amine", "sulfonamide", "pyridine"], "moleculenet_labels": {"label": 1.41}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["amine", "sulfonamide", "pyridine"]}}
{"id": "eval_00744", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "IAPP, MAFA, INS", "output": "This profile is most consistent with pancreatic beta cell, which is commonly found in breast under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "breast", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["IAPP", "MAFA", "INS"]}}
{"id": "eval_00745", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL20_PROM0\n\n### Answer:", "output": "Binds directly to 23S ribosomal RNA and is necessary for the in vitro assembly process of the 50S ribosomal subunit.  It is not involved in the protein synthesizing functions of that subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00746", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nHXC6_SHEEP\n\n### Answer:", "output": "Belongs to the Antp homeobox family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00747", "category": "Mutation", "instruction": "Describe the effect of mutation P42212-H145Y on protein Green fluorescent protein.", "input": "### Wild-type:\nMSKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKQHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNHNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYK\n\n### Mutant:\nMSKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTGKLPVPWPTLVTTFSYGVQCFSRYPDHMKQHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYK\n\n### Answer:", "output": "Increases mut1.3; shifts fluorescence lifetime from 2.78 to 3.03 ns. Decreases mut1.5; shifts fluorescence lifetime from 2.72 to 3.03 ns; when associated with A-193. Decreases mut1.27; shifts fluorescence lifetime from 2.85 to 3.03 ns; when associated with E-103. Decreases mut1.28; shifts fluorescence lifetime from 2.76 to 3.03 ns; when associated with R-30. Decreases mut2.1; shifts fluorescence lifetime from 2.50 to 3.03 ns; when associated with A-176 and I-198. Increases mut2.2; shifts fluorescence lifetime from 1.94 to 3.03 ns; when associated with R-30 and H-69. Increases mut3.3; shifts fluorescence lifetime from 1.88 to 3.03 ns; when associated with R-30; L-46 and H-69. Mutation of Y145 to either histidine or cysteine was found to decrease the fluorescence lifetime of GFPuv from 2.78 +/- 0.05 to 3.03 +/- 0.03 ns for the mutant and to 2.74 +/- 0.05 ns for . Some of the longer-lifetime mutants exhibited excitation peaks that were blue-shifted relative to their maximal absorption, indicating that the mutations restricted the adoption of additional conformations relative to wtGFPuv.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00748", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKNYHDIVLALAGVCQSAKLVHQLATESRADSETFLTALNSLFITQPQRIEDVFGGEVRHLKLGLETLIHQLNAQGDQNLTRYWLSLLALEGKLSKNPDAKQTLGNRISRLKEQEIHYARDSEIMLSIMANIYSDIISPLGKKIHILGSPDYLRQELVQNKIRAVLLAGIRSAVLWKQMGGTKWQILFFRRKLLATAKQIYSSIY\n\n### Sequence 2:\nMQNPQNLIWIDLEMTGLDPDRDVIIEMATIVTDSDLNTLAEGPVIAIHQPEEILAGMDEWNTRQHGQSGLTQRVRESTVSMAEAEAQTLAFLEQWVPKRSSPICGNSICQDRRFLYRHMPRLEGYFHYRNLDVSTLKELAARWAPQVRESFKKGNTHLALDDIRESIAELRHYRDHFIKL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00749", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nMYH7:medium MYH6:high TNNI3:high\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as TNNI3, MYH6, MYH7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "bone marrow", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MYH7", "MYH6", "TNNI3"]}}
{"id": "eval_00750", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGENLYFQGHMTQQPQAKYRHDYRAPDYQITDIDLTFDLDAQKTVVTAVSQAVRHGASDAPLRLNGEDLKLVSVHINDEPWTAWKEEEGALVISNLPERFTLKIINEISPAANTALEGLYQSGDALCTQCEAEGFRHITYYLDRPDVLARFTTKIIADKIKYPFLLSNGNRVAQGELENGRHWVQWQDPFPKPCYLFALVAGDFDVLRDTFTTRSGREVALELYVDRGNLDRAPWAMTSLKNSMKWDEERFGLEYDLDIYMIVAVDFFNMGAMENKGLNIFNSKYVLARTDTATDKDYLDIERVIGHEYFHNWTGNRVTCRDWFQLSLKEGLTVFRDQEFSSDLGSRAVNRINNVRTMRGLQFAEDASPMAHPIRPDMVIEMNNFYTLTVYEKGAEVIRMIHTLLGEENFQKGMQLYFERHDGSAATCDDFVQAMEDASNVDLSHFRRWYSQSGTPIVTVKDDYNPETEQYTLTISQRTPATPDQAEKQPLHIPFAIELYDNEGKVIPLQKGGHPVNSVLNVTQAEQTFVFDNVYFQPVPALLCEFSAPVKLEYKWSDQQLTFLMRHARNDFSRWDAAQSLLATYIKLNVARHQQGQPLSLPVHVADAFRAVLLDEKIDPALAAEILTLPSVNEMAELFDIIDPIAIAEVREALTRTLATELADELLAIYNANYQSEYRVEHEDIAKRTLRNACLRFLAFGETHLADVLVSKQFHEANNMTDALAALSAAVAAQLPCRDALMQEYDDKWHQNGLVMDKWFILQATSPAANVLETVRGLLQHRSFTMSNPNRIRSLIGAFAGSNPAAFHAEDGSGYLFLVEMLTDLNSRNPQVASRLIEPLIRLKRYDAKRQEKMRAALEQLKGLENLSGDLYEKITKALA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCEEGGGCCCCSEEEEEEEEEEECCSSCEEEEEEEEEEECSCTTSCEEEECSSCEEEEEEETTEECCSEEEETTEEEECSCCSEEEEEEEEEECGGGCCSCCEEEEETTEEEEECTTTTGGGTSCBCCSTTCCEEEEEEEEEETTTCCEEEESSEEEEEEECSTTEEEEEEEEEEEECGGGCCEEEECCEEEEEEEECTTSCEEEEEEEECTTCGGGCHHHHHHHHHHHHHHHHHHCCCCCSSEEEEEEESSCSSSEECCTTEEEEEGGGTCCCTTTCCHHHHHHHHHHHHHHHHTTTBTTTBEESSGGGHHHHHHHHHHHHHHHHHHHSCHHHHHHHHHHHHHHTHHHHHTSTTCCCSSCSEESCGGGGCCHHHHHHHHHHHHHHHHHHCHHHHHHHHHHHHHHHTTSEECHHHHHHHHHHHHCCCCTTTTHHHHCCSCCEEEEEEEEETTTTEEEEEEEEECCCBTTBSCCCCCCEEEEEEEECTTSCBCCCEETTEECCSEEEECSSEEEEEESCCCSCCEEEESTTCCSSCEEECCCCHHHHHHHHHHCSSHHHHHHHHHHHHHHHHHHHHHHHHTTCCCCCCHHHHHHHHHHHHCSSSCHHHHHHHTSCCCHHHHHTTCSSCCHHHHHHHHHHHHHHHHHHTHHHHHHHHHHTCCSSCCCSHHHHHHHHHHHHHHHHHHTSSHHHHHHHHHHHHHHCSSHHHHHHHHHHHHHTTCTTHHHHHHHHHHHHTTCHHHHHHHHHHHHTCCCTTHHHHHHHHTTSTTCCTTCHHHHHHHHHHHHHHCHHHHTCTTCHHHHHHHHHHHHHHHHCHHHHHHHHHHHHTGGGSCHHHHHHHHHHHHHHHTCTTCCHHHHHHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00751", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nMBP:low PLP1:high MOG:medium\n</cell>", "output": "This cell is most consistent with oligodendrocyte, an oligodendrocyte. Key markers such as MBP, PLP1, MOG support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "oligodendrocyte", "tissue": "colon", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MBP", "PLP1", "MOG"]}}
{"id": "eval_00752", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMARLKARYNDELKAKLQEELSIKNVMEIPRITKITLNMGVGAAATDKKLLDGAVADMQLIAGQKPVVTLARKSIAGFKIRDGWPIGCKVTLRGDQMYEFLDRLISIAIPRIRDFRGFSAKSFDGRGNYSMGLKEQIVFPEIDFDKIDRIRGMDITITTTARTDDEGRALMRAFGFPFK\n\n### Sequence 2:\nMPVDQEKLAKLHKLSAANKVGGTRRKINKKGNLYNNNDKDNTKLQAELHKLHPMTIENVAEANFFKKNGKVLHFNSAVVQIAPQCNLTMIHGQPKENTLNGLYPSVASQLGSQELEYLTGLAHNLENEQTVLDQLGDRCSETKQQVMNS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00753", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: O=C(O)C1C(C(=O)O)C2(Cl)C(Cl)=C(Cl)C1(Cl)C2(Cl)Cl\nDescriptors: MW=388.85, LogP=3.23, TPSA=74.6, HBD=2, HBA=2", "output": "Descriptors: MW=388.85, LogP=3.23, TPSA=74.6, HBD=2, HBA=2, QED=0.71, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "O=C(O)C1C(C(=O)O)C2(Cl)C(Cl)=C(Cl)C1(Cl)C2(Cl)Cl", "compound_name": null, "pubchem_cid": null, "MW": 388.85, "LogP": 3.23, "TPSA": 74.6, "QED": 0.71, "functional_groups": ["hydroxyl", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": null, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00754", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSVADIKKSEVVAQFARGTNDTGSPEVQVALLTARIVELTGHFKTHAKDHHSRRGLLRMVSRRRKLLDYLKGKDADRYRALIEKLGLRK\n\n### Sequence 2:\nMHEMSIALSIVDAVDAQARVEGAGRISRVELVIGRLAGIEPESLRFCFPAAATGTLAEGAELDIEEIAAVAVCGACGFRFSVTFPVAECPECRSLRISVVSGEEFVIRSITIEEGD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00755", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRF1_AZOVD\n\n### Answer:", "output": "Belongs to the prokaryotic/mitochondrial release factor family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00756", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "LTB, CCR7, IL7R", "output": "This cell is most consistent with naive CD4+ T cell, a T-cell subtype with naive/central-memory-like profile. Key markers such as LTB, CCR7, IL7R support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "ovary", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["LTB", "CCR7", "IL7R"]}}
{"id": "eval_00757", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAMKIRLARGGSKKRPFYRIVAADSRMPRDGRFIEKLGTYNPLLPKDSEERVKMDIEKIQAWLDKGAQPTDRVARMLEAAGVREKTERNNPNKAKPGKKAQERAEEKAAKAAEAAEAADAE\n\n### Sequence 2:\nMAMKIRLARGGSKKRPFYRIVASDSRMPRDGRFIEKLGTYNPLLPKDSEDRVKMDMERVQHWLDQGAQPTDRISRFLEAAGHTPKKERANMKKAQPGKKAVER\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00758", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSVSAFNRRWAAVILEALTRHGVRHICIAPGSRSTPLTLAAAENSAFIHHTHFDERGLGHLALGLAKVSKQPVAVIVTSGTAVANLYPALIEAGLTGEKLILLTADRPPELIDCGSNQAIRQPGMFASHPTHSISLPRPTQDIPARWLVSTIDHALGTLHAGGVHINCPFAEPLYGEMDDTGISWQQRLGDWWQDDKPWLREAPRRESEKQRDWFFWRQKRGVVVAGRMSAEEGKKVALWAQTLGWPLIGDVLSQTGQPLPCADLWLGNAKATSELQQAQIVVQLGSSLTGKRLLQWQASCEPEEYWIVDDIEGRLDPAHHRGRRLIANIADWLELHPAEKRQPWCVEIPRLAEQAMQAVIARRDAFGEAQLAHRISDYLPEQGQLFVGNSLVVRLIDALSQLPAGYPVYSNRGASGIDGLLSTAAGVQRASGKPTLAIVGDLSALYDLNALALLRQVSAPLVLIVVNNNGGQIFSLLPTPKSERERFYLMPQNVHFEHAAAMFELKYHRPQNWQELETTLVDAWRTPTTTVIEMVVNDTDGAQTLQQLLAQVSHL\n\n### Sequence 2:\nMDEFHRCGKEDSFWQQCFLYPLFFQEDLYAISHDHYLDVSSSSRPMEHLSSNDQLSFLTVKRLIGQIRQQNHSIVLFVNCDPNPLADRKKSFYSESVLEALTLVLEVPFSIWSKSSVEGMNECKSFRSIHSIFPFLEDKFPHSNSILDARIPYSIHPEILVRTFRRWIRDAPSLHPLRSVLYDYRNSPENLQRSIIVVPRVNTRFFLFLLNYYVCECESILFSRLKRSSHSRSLDHGSFPQRTHFHRKIKHIIIFSRRNSLKSIWSLKDPKIHYVRYGERPIIAIKGADLLVKKCRYYLLIFRQFYFHLWSEPYRVCSHQLSKNCSSSPGYFLRVRMNPLLVRTKTLDELFIPVLITNEMDPIVPIVPIIGLLATEKFCDISGRPISKLSWTSLTDDDILDRFDQIWRNLFHYYSGSFDRDGLYRIKYILLLSCAKTLACKHKSTIRVVRKELGPELFKKSFSKEREFDSLPFSSKAAARSQRERIWHSDIPQINPLANSWQKIQDLKIENLFDQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00759", "category": "Mutation", "instruction": "Describe the effect of mutation Q04837-E111Q on protein Single-stranded DNA-binding protein, mitochondrial.", "input": "### Wild-type:\nMFRRPVLQVLRQFVRHESETTTSLVLERSLNRVHLLGRVGQDPVLRQVEGKNPVTIFSLATNEMWRSGDSEVYQLGDVSQKTTWHRISVFRPGLRDVAYQYVKKGSRIYLEGKIDYGEYMDKNNVRRQATTIIADNIIFLSDQTKEKE\n\n### Mutant:\nMFRRPVLQVLRQFVRHESETTTSLVLERSLNRVHLLGRVGQDPVLRQVEGKNPVTIFSLATNEMWRSGDSEVYQLGDVSQKTTWHRISVFRPGLRDVAYQYVKKGSRIYLQGKIDYGEYMDKNNVRRQATTIIADNIIFLSDQTKEKE\n\n### Answer:", "output": "In OPA13; reduced ability to stimulate POLG-dependent DNA synthesis in vitro. The mutation in SSBP1 affects the amount of SSBP1 protein and alters multimer formation, but not the binding to ssDNA. This mutation impairs mtDNA, nucleoids, and 7S-DNA amounts as well as mtDNA replication, affecting replisome machinery. The variable mtDNA depletion in cells is reflected in the severity of mitochondrial dysfunction, including respiratory efficiency, OXPHOS subunits, and complex amount and assembly.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00760", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nATAT2_DROME\n\n### Answer:", "output": "Alpha-tubulin N-acetyltransferase 2, Alpha-TAT 2, TAT 2, Acetyltransferase mec-17 homolog 2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00761", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGPGSEFDPGLPSTEDVILKTEQVTKNIQELLRAAQEFKHDSFVPCSEKIHLAVTEMASLFPKRPALEPVRSSLRLLNASAYRLQSECRKTVPPEPGAPVDFQLLTQQVIQCAYDIAKAAKQLVTITTREKKQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHHTCGGGHHHHHHHHHHHHHHHHTTSCSSCSSHHHHHHHHHHHHHHHHHHHHHGGGCCCCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHTCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00762", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDIKGLLVILFFVLLITGEVENMKPKKHKYKYKNRKRFLENFEEPLQKRDMNEESYELFNQ\n\n### Sequence 2:\nMLNYIHQLVKPQRICTFASVAWTAQCTRVSSIHWRAGKSFTFQNQPKWKSLAFWRAQKAWLHSHNG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00763", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAKRDYYEILGVSKTAEEREIKKAYKRLAMKYHPDRNQGDKEAEAKFKEIKEAYEVLTDAQKRAAYDQYGHAAFEQGGMGGGFGGGFNGGADFSDIFGDVFGDIFGGGRGRQRAARGADLRYNMDLTLEEAVRGVTKEIRIPTLEECDVCHGSGAKAGTQPQTCPTCHGSGQVQMRQGFFAVQQTCPHCQGRGTLIKDPCHKCHGHGRVEKSKTLSVKIPAGVDTGDRIRLAGEGEAGEHGAPAGDLYVQVQVKQHPIFEREGNNLYCEVPINFAMAALGGEIEVPTLDGRVMLKVPSETQTGKLFRMRGKGVKSVRGGAQGDLLCRVVVETPVGLSEKQKQLLKDLQESFGGPTGEKNSPRSKSFFDGVKKFFDDLTR\n\n### Sequence 2:\nMAKQDYYEILGVSKTAEEREIKKAYKRLAMKYHPDRNQGDKEAEAKFKEIKEAYEILTDAQKRAAYDQYGHAAFEQGGMGGG--GGFGGGADFSDIFGDVFGDIFGGGRGRQRAARGADLRYNMELTLEEAVRGVTKEIRIPTLEECDVCHGSGAKAGSKPQTCPTCHGAGQVQMRQGFFAVQQTCPHCQGRGTLIKDPCNKCHGHGRVEKTKTLSVKIPAGVDTGDRIRLAGEGEAGEHGAPAGDLYVQVQVKQHAIFEREGNNLYCEVPINFTMAALGGEIEVPTLDGRVSLKVPGETQTGKLFRMRGKGVKSVRGGAQGDLLCRVVVETPVGLNEKQKQLLKELQESFGGPTGENNSPRSKSFFDGVKKFFDDLTR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00764", "category": "Mutation", "instruction": "Describe the effect of mutation Q13435-K509R on protein Splicing factor 3B subunit 2.", "input": "### Wild-type:\nMATEHPEPPKAELQLPPPPPPGHYGAWAAQELQAKLAEIGAPIQGNREELVERLQSYTRQTGIVLNRPVLRGEDGDKAAPPPMSAQLPGIPMPPPPLGLPPLQPPPPPPPPPPGLGLGFPMAHPPNLGPPPPLRVGEPVALSEEERLKLAQQQAALLMQQEERAKQQGDHSLKEHELLEQQKRAAVLLEQERQQEIAKMGTPVPRPPQDMGQIGVRTPLGPRVAAPVGPVGPTPTVLPMGAPVPRPRGPPPPPGDENREMDDPSVGPKIPQALEKILQLKESRQEEMNSQQEEEEMETDA\n\n### Mutant:\nMATEHPEPPKAELQLPPPPPPGHYGAWAAQELQAKLAEIGAPIQGNREELVERLQSYTRQTGIVLNRPVLRGEDGDKAAPPPMSAQLPGIPMPPPPLGLPPLQPPPPPPPPPPGLGLGFPMAHPPNLGPPPPLRVGEPVALSEEERLKLAQQQAALLMQQEERAKQQGDHSLKEHELLEQQKRAAVLLEQERQQEIAKMGTPVPRPPQDMGQIGVRTPLGPRVAAPVGPVGPTPTVLPMGAPVPRPRGPPPPPGDENREMDDPSVGPKIPQALEKILQLKESRQEEMNSQQEEEEMETDA\n\n### Answer:", "output": "Moderately diminished formation of omega-N monomethylarginine but greatly reduced formation of symmetrical dimethylarginine; when associated with R-507. Abolishes formation of omega-N monomethylarginine and formation of symmetrical dimethylarginine; when associated with K-508 and R-507.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00765", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMHISKLPAALLIFSVALNHLKATPVRSGTNHQMDKRKCNTATCATQRLANFLVHSNNNLGPVLSPTNVGSNTYGKRSAAEIPDGDSLDLFLL\n\n### Sequence 2:\nMGILKLQVFLIVLSVALNHLKATPIES---HQVEKRKCNTATCATQRLANFLVHSSNNFGAILSSTNVGSNTYGKRNAVEVLKREPLN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00766", "category": "Mutation", "instruction": "Describe the effect of mutation P32578-G2A on protein SNF1 protein kinase subunit beta-1.", "input": "### Wild-type:\nMGNSPSTQDPSHSTKKEHGHHFHDAFNKDRQGSITSQLFNNRKSTHKRRASHTSEHNGAIPPRMQLLASHDPSTDCDGRMSSDTTIDKGPSHLFKKDYSLSSAADVNDTTLANLTLSDDHDVGAPEEQVKSPSFLSPGPSMATVKRTKSDLDDLSTLNYTMVDETTENERNDKPHHERHRSSIIALKKNLLESSATASPSPTRSSSVHSASLPALTKTDSIDIPVRQPYSKKPSIHAYQYQYLNNDETFSENSQMDKEGNSDSVDAEAGVLQSEDMVLNQSLLQNALKKDMQRLSRVNSS\n\n### Mutant:\nMANSPSTQDPSHSTKKEHGHHFHDAFNKDRQGSITSQLFNNRKSTHKRRASHTSEHNGAIPPRMQLLASHDPSTDCDGRMSSDTTIDKGPSHLFKKDYSLSSAADVNDTTLANLTLSDDHDVGAPEEQVKSPSFLSPGPSMATVKRTKSDLDDLSTLNYTMVDETTENERNDKPHHERHRSSIIALKKNLLESSATASPSPTRSSSVHSASLPALTKTDSIDIPVRQPYSKKPSIHAYQYQYLNNDETFSENSQMDKEGNSDSVDAEAGVLQSEDMVLNQSLLQNALKKDMQRLSRVNSS\n\n### Answer:", "output": "Prevents relocalization to the vacuolar membrane.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00767", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: HIV protease\nLigand SMILES: Nc1ccccc1C(=O)OCC(O)CO\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and HIV-1 protease (HIV protease) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "Nc1ccccc1C(=O)OCC(O)CO", "compound_name": null, "pubchem_cid": null, "MW": 211.22, "LogP": -0.22, "TPSA": 92.78, "QED": 0.47, "functional_groups": ["aromatic ring", "ester", "hydroxyl", "ether"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": null, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": false}}
{"id": "eval_00768", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMARKPLIAGNWKMNLNHLEAIALVQKIAFALPEKYFEKVDVAVIPPFVDIRSVQTLVEGDKLLLTYGAQDVSVHESGAYTGEISASMLAKLGCTFVVVGHSERRQYHHEDDATVLGKAKKALEHGLTPIVCIGEGLNVREAGTHVEYNLEQLRGSLKGLSAEQIAKVVIAYEPVWAIGTGKVASAADAQEVCGAIRAELAELAGPEVAAQVRVLYGGSVNAKNVGELVAQPDVDGALVGGASLKGDEFATLSAIAAGGPLP\n\n### Sequence 2:\nMSAATFRGYAPAKVNLALHVTGRRVDGYHELDSLVVFAGVGDRLEIAPADALSLTVTGPRAEGVPDDARNLVWKAADWLAPGRGAAMTLDKHLPHAGGIGGGSADAACALRGLAEIWDVDVPDGAEALGADVPVCLHGQPVRMRGIGDVLDAVPPLPPMWIVLVNAGVEVPTGAVFKAMERVDNPPLPAPAWDGFDSFLVWLERTRNDMESSARSQAPVIGMVLERLRALPGCRFTRMSGSGGTCFGLFETEIAARDAALGLPKHWWVTYAPVLRA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00769", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSWDSYIDNLIAQTKDASGTGHSDKACIIGIDGGAPWTTAGHANALKLEGQEGPNIARCFKSKDFTPFMSSGIVADGTKYQFLREEDGKLVLAKKKGQGALTLQSSKTAIVIGHAPEGGQQGNTNKGVAVIAEYLESLGM\n\n### Sequence 2:\nMSWDSYVDNLIAQSKDASGTTHCDKACIIGKDGSA-WTTMPTSDTSNNLKLTPEEMANIAKCFKSKDFAAFMSSGIYVNGTKYQFLREEDSKLVLGKKKGEGSLTLQSSKTAIVIGHCPEGGQQGNLNKAVGVIAEYLESLSM\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00770", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMDTGAKLKKGAGERKGGGPKKKPVSRSVKAGLQFPVGRIGRFLKKGRYAQRVGSGAPVYLAAVLEYLAAEVLELAGNAARDNKKNRIIPRHVLLAVRNDEELGKLLAGVTIAHGGVLPNINPVLLPKKAEKAAAAATKEPKSPAKATKSPKKA\n\n### Sequence 2:\nMIIGLTGGIGVGKSFIANCFKEFGAAVFDADFIVHQLYRVDKNIISYAEKNFPGAIANGEIDKTVLSKYFLDYDENWKQFQSLVHSAVQNELEIFIAQDKEINRKLLVLDVPLLLETRFHLYCDFIIFIYADSAAQAQRLSERNIDKEKLDLISSIQLPVKEKRQMSDFTIDTSTSKEHVLSQVKDIVDSLSLSS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00771", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: CCOC(=O)COc1ccc2c(c1)CC(NCC(O)c1cccc(Cl)c1)CC2\nDescriptors: MW=403.91, LogP=3.46, TPSA=67.79, HBD=2, HBA=5", "output": "Descriptors: MW=403.91, LogP=3.46, TPSA=67.79, HBD=2, HBA=5, QED=0.66, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "CCOC(=O)COc1ccc2c(c1)CC(NCC(O)c1cccc(Cl)c1)CC2", "compound_name": null, "pubchem_cid": null, "MW": 403.91, "LogP": 3.46, "TPSA": 67.79, "QED": 0.66, "functional_groups": ["aromatic ring", "ester", "amine", "hydroxyl", "ether", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00772", "category": "Mutation", "instruction": "Describe the effect of mutation P33690-D67E on protein Tryptophan prenyltransferase ComQ.", "input": "### Wild-type:\nMKEIVEQNIFNEDLSQLLYSFIDSKETFSFAESTILHYVVFGGENLDVATRLGAGIEILILSSDIMDDLEDEDNHHALWMKINRSESLNAALSLYTVGLTSIYSLNNNPLIFKYVLKYVNEAMQGQHDDITNKSKTEDESLEVIRLKCGSLIALANVAGVLLATGEYNETVERYSYYKGIIAQISGDYYVLLSGNRSDIEKNKHTLIYLYLKRLFNDASEDLLYLISHKDLYYKSLLDKEKFQEKLIKAGVTQYISVLLEIYKQKCISAIEQLNLDKEKKELIKECLLSYTKGDTRCKT\n\n### Mutant:\nMKEIVEQNIFNEDLSQLLYSFIDSKETFSFAESTILHYVVFGGENLDVATRLGAGIEILILSSDIMEDLEDEDNHHALWMKINRSESLNAALSLYTVGLTSIYSLNNNPLIFKYVLKYVNEAMQGQHDDITNKSKTEDESLEVIRLKCGSLIALANVAGVLLATGEYNETVERYSYYKGIIAQISGDYYVLLSGNRSDIEKNKHTLIYLYLKRLFNDASEDLLYLISHKDLYYKSLLDKEKFQEKLIKAGVTQYISVLLEIYKQKCISAIEQLNLDKEKKELIKECLLSYTKGDTRCKT\n\n### Answer:", "output": "Loss of activity. Cannot complement the null mutant.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00773", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nASTE_SALHS\n\n### Answer:", "output": "Transforms N(2)-succinylglutamate into succinate and glutamate.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00774", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAKTAKGIRPGELNLKEKLVHINRTAKVVKGGKRFGFNAIVVVGDKEGHVGYGLGKANEVQDAIAKGVEDGKKNVVKVPIIKGTIPHQIVAKYGSAKVLMKPATPGTGLIAGGAVRAVLEMAGVRDVLTKSLGSSNPHNVVKAAILGLKSISDANDVAERRSKSLKEVFES\n\n### Sequence 2:\nMSDDLNDFFAKKDSTKKVVKKPTSTTPKVVVKPVSPTTPVVSPTTVTSPVSPTTPVVEPKKVVDLSQLNKSKDTTPVVDVKESKTEQINIPTMRWADKNEQTTTTTQTKVYKNYPSLKKEETKVDEDDDQIYEDKEEEKEVKKDEENSTADEEPVKKSNKKVAPKQKKKSKQELEAEALMASLGIIDEKPVPKQKKK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00775", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMGFEWPWQYNFPPFFTLQPNVDTRQKQLSAWSSLVLSYCRQNKLYTMNLMEIQESPLFNNKKIQRKLSLESVQVVLEELRKKGNLEWIDKNKSRFLIMWRRPDEWGKVIYQWVSKNGMTNSVFTLYELISGDDTEGEEFHGLDEAMLLRSLEALQQEHKAEIITLNESRGVKFF\n\n### Sequence 2:\nMNPRRKKRLTLAVALIVGVAGAASLLLYALNSNLNLFYTPNEIINGKNDTGIKPEIGQRIRIGGMVTMGSMVRDPDSLHVEFAVHDSTGTQVIVTYDDLLPDLFREGQGIVAQGVLTDSGKLEATEVLAKHDENYMPPEVAEAMGKQHKKLEYQEGVEKTAQY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00776", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMDNEIIIIVIVIIIFFFYLKQKKLTNCETQVVKVQKDIDEINLKLKKLNK\n\n### Sequence 2:\nMDNEIIIIVIVIIIFFFYLKQKKLTNCETQVVKVQKDIDEINLKLKKLNK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00777", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMLDAATPTHQRVRGRAMARFTGRRLRDLHQSGSAKVLIPRVFGRAPEVVFLNTAGGITGGDRLDYALHVDDGTCVGTTQTAERAYRSHGPMGRVSTRLTIGQGATLHWVPQEMILFDGVALERDLSVEMAGDAELVMLETLVLGRAAMGEVLRDLHLRDRRRVTRGGKLAMVEAIGLGPDDFASSSPAGLNGAVATASLTLMAQDAEDRLNALRAVLPTDVRSAASAWDGRLTARFLAPQAYPLRRAVARAVEQVTCGALPRVWQI\n\n### Sequence 2:\nMSVISMKQLLEAGVHFGHQTRRWNPKMKPYIFTERNGIHVIDLQKTVKLVDDAYNYVKNASQEGAVVLFVGTKKQAAEAVKEEALRAGQYYVNHRWLGGMLTNWNTIQTRVTRLKEINKMEEEGTFEVLPKKEVVLLNKERERLEKFIGGIADMPRIPDVMYIVDPHAEQIAVKEAKTLGIPVVAMVDTNADPEPIDVVIPANDDAIRAVKLITAKMADAIIEGRQGEDAADDFVAEEAASEESLEELAEIVEGK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00778", "category": "Mutation", "instruction": "Describe the effect of mutation Q641K1-H912A on protein Cytosolic carboxypeptidase 1.", "input": "### Wild-type:\nMSKLKVVGEKSLTNSSRVVGLLAQLEKINTDSTESDTARYVTSKILHLAQSQEKTRREMTTKGSTGMEVLLSTLENTKDLQTVLNILSILIELVSSGGGRRASFLVAKGGSQILLQLLMNASKDSPPHEEVMVQTHSILAKIGPKDKKFGVKARVNGALTVTLNLVKQHFQNYRLVLPCLQLLRVYSTNSVNSVSLGKNGVVELMFKIIGPFSKKNSGLMKVALDTLAALLKSKTNARRAVDRGYVQVLLTIYVDWHRHDNRHRNMLIRKGILQSLKSVTNIKLGRKAFIDANGMKILYN\n\n### Mutant:\nMSKLKVVGEKSLTNSSRVVGLLAQLEKINTDSTESDTARYVTSKILHLAQSQEKTRREMTTKGSTGMEVLLSTLENTKDLQTVLNILSILIELVSSGGGRRASFLVAKGGSQILLQLLMNASKDSPPHEEVMVQTHSILAKIGPKDKKFGVKARVNGALTVTLNLVKQHFQNYRLVLPCLQLLRVYSTNSVNSVSLGKNGVVELMFKIIGPFSKKNSGLMKVALDTLAALLKSKTNARRAVDRGYVQVLLTIYVDWHRHDNRHRNMLIRKGILQSLKSVTNIKLGRKAFIDANGMKILYN\n\n### Answer:", "output": "Abolishes ability to rescue Purkinje cell degeneration in pcd mice when expressed in a transgene. Mutation of the site in the Nna1 gene affects its ability to support neuron survival. Loss of function of Nna1 causes Purkinje cell degeneration (pcd) in mice. The pcd mice display altered proteolytic processing of Nna1 interacting proteins and mitochondrial abnormalities. The mutation in CCP1 catalyzes the shortening of polyglutamate chains.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00779", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "CC(C)C(NC(=O)Cn1c(-c2ccccc2)ncc(NC(=O)OCC(F)(F)F)c1=O)C(=O)C(F)(F)F", "output": "The SMILES represents an unnamed molecule with MW=522.4, LogP=3.29, TPSA=119.39. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Lipophilicity", "canonical_smiles": "CC(C)C(NC(=O)Cn1c(-c2ccccc2)ncc(NC(=O)OCC(F)(F)F)c1=O)C(=O)C(F)(F)F", "compound_name": null, "pubchem_cid": null, "MW": 522.4, "LogP": 3.29, "TPSA": 119.39, "QED": 0.51, "functional_groups": ["aromatic ring", "ester", "amide", "ether", "ketone", "halide"], "moleculenet_labels": {"label": 2.37}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00780", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nERGWFDILDDWLKRDRFVFVGWSGILLFPCAYLALGGWLTGTTFVTSWYTHGLASSYLEGCNFLTVAVSTPANSMGHSLLLLWGPEAQGDFTRWCQLGGLWTFIALHGAFGLIGFMLRQFEIARLVGVRPYNAIAFSAPIAVFVSVFLIYPLGQSSWFFAPSFGVAAIFRFLLFFQGFHNWTLNPFHMMGVAGVLGGALLCAIHGATVENTLFQDGEGASTFRAFNPTQAEETYSMVTANRFWSQIFGIAFSNKRWLHFFMLFVPVTGLWMSAIGVVGLALNLRSYDFISQEIRAAEDPEFETFYTKNLLLNEGIRAWMAPQDQPHENFVFPEEVLPRGNAL\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHTBCCSSCCHHHHHHHHHHHHHHHHHHHHHHHHCCBTTTTSBCCSGGGTCCTTTCEECCCCGGGTTCCCCTTSTTTTTCHHHHHHHTHHHHHHHHHHHHHHHHHHHHHHHHHHHHTCCCHHHHHTHHHHHHHHHHHTHHHHHSSSGGGSCCEEHHHHHHHHHHHHHHHCGGGCHHHHHHHHHHHHHHHHHHHHHHHHHTSBBCCSSCSSCSTTCCTTCCBCSBCHHHHHHHHHHHHSCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHTTTCBCCCCHHHHHHHHHCTTCCCHHHHHHHHHHHHHHHHHHHHSGGGCCCCCGGGSCCCSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00781", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nASVDTMNARNTQFTKAFHALKQNAGSHSPSMEDLKKMFPTLEIKIDACYLSNPYASELVLDYIDRELIQTNAYKKVLTHYPSQQRSLQKVMAESLHVKPENIFIGNGATEIIQMLLQQEEVQKVALMIPTFSSYYEFVGKGCEVVYFPLNERDDYSFDADKYCQFIENEQPDTVVLINPNNPNGAYLSLEKMHILLKRLAFVPRIIIDESFIHFAYEDEALTCLSSTVLFDMYPNVIIVKSLSKDFGIAGVRLGYALMDSRKIDALLEHGFLWNINGIGEYCLRLFVREDFLKRYEEARKQYIKEMCRFKEALLGIENVYVYPSMANFVMLKLPSRIKASFVISALLVEYGIYVRTMADKIGVEGECIRIAGRTREENNCIVMALKSILKDSK\n\n### Answer:", "output": "<SEQ_2D>CCHHHHTTTTHHHHHHHHHSCCCSSSSSSCHHHHHHHSTTCCCCEECSSCCCHHHHHHHHHHHHHHTTTTSTHHHHHHSCCCCHHHHHHHHHHHHTSCGGGEEEESCHHHHHHHHHTSTTCCEEEEEESCSSCCGGGSCTTCEEEEEECCGGGTTCCCHHHHHHHHHHHCCSEEEEESSCTTTCCCCCHHHHHHHHHHTTTSSEEEEECTTGGGCCSSTTCCCCCTHHHHHHCTTEEEEEESSTTTTCTTSCCEEEECCHHHHHHHTSSCCCCCCCHHHHHHHHHHTCHHHHHHHHHHHHHHHHHHHHHHHHHHTCTTCEECCCSSSEEEEECCTTCCHHHHHHHHHHHHCEECEECTTSTTCCSSEEEEECCCHHHHHHHHHHHHHHTCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00782", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nH4_DROYA\n\n### Answer:", "output": "Histone H4.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00783", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UBF8-V55A on protein Phosphatidylinositol 4-kinase beta.", "input": "### Wild-type:\nMGDTVVEPAPLKPTSEPTSGPPGNNGGSLLSVITEGVGELSVIDPEVAQKACQEVLEKVKLLHGGVAVSSRGTPLELVNGDGVDSEIRCLDDPPAQIREEEDEMGAAVASGTAKGARRRRQNNSAKQSWLLRLFESKLFDISMAISYLYNSKEPGVQAYIGNRLFCFRNEDVDFYLPQLLNMYIHMDEDVGDAIKPYIVHRCRQSINFSLQCALLLGAYSSDMHISTQRHSRGTKLRKLILSDELKPAHRKRELPSLSPAPDTGLSPSKRTHQRSKSDATASISLSSNLKRTASNPKVEN\n\n### Mutant:\nMGDTVVEPAPLKPTSEPTSGPPGNNGGSLLSVITEGVGELSVIDPEVAQKACQEALEKVKLLHGGVAVSSRGTPLELVNGDGVDSEIRCLDDPPAQIREEEDEMGAAVASGTAKGARRRRQNNSAKQSWLLRLFESKLFDISMAISYLYNSKEPGVQAYIGNRLFCFRNEDVDFYLPQLLNMYIHMDEDVGDAIKPYIVHRCRQSINFSLQCALLLGAYSSDMHISTQRHSRGTKLRKLILSDELKPAHRKRELPSLSPAPDTGLSPSKRTHQRSKSDATASISLSSNLKRTASNPKVEN\n\n### Answer:", "output": "Loss of interaction with ACBD3.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00784", "category": "Mutation", "instruction": "Describe the effect of mutation P18597-Q130N on protein Potassium-transporting ATPase subunit beta.", "input": "### Wild-type:\nMAALQEKKSCSQRMEEFRHYCWNPDTGQMLGRTLSRWVWISLYYVAFYVVMTGLFALCIYVLMQTIDPYTPDYQDQLKSPGVTLRPDVYGEKGLEIHYNISDNRTWTSLTHTLRSFLAGYSPAAQVDNIQCTSKTYFFQESFGAPNHTKFSCKFTADMLENCSGLTDPSFGFKEGKPCFIIKMNRIVRFLPSNSTPPRVDCTFLDMPHQALTPLQVEYYPPNGTFSLHYFPYYGKKAQPHYSNPLVAAKLLNVPTNTEVVVLCKILADHVTFDNPHDPYEGKVEFKLKIQK\n\n### Mutant:\nMAALQEKKSCSQRMEEFRHYCWNPDTGQMLGRTLSRWVWISLYYVAFYVVMTGLFALCIYVLMQTIDPYTPDYQDQLKSPGVTLRPDVYGEKGLEIHYNISDNRTWTSLTHTLRSFLAGYSPAAQVDNINCTSKTYFFQESFGAPNHTKFSCKFTADMLENCSGLTDPSFGFKEGKPCFIIKMNRIVRFLPSNSTPPRVDCTFLDMPHQALTPLQVEYYPPNGTFSLHYFPYYGKKAQPHYSNPLVAAKLLNVPTNTEVVVLCKILADHVTFDNPHDPYEGKVEFKLKIQK\n\n### Answer:", "output": "Enhances targeting to the plasma membrane; when associated with Q-99 and Q-222. Gain of pump ATPase activity; when associated with Q-99; Q-103; Q-146; Q-161; Q-193 and Q-222. K(+)-ATPase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00785", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "O=[N+]([O-])c1ccc2c(c1)S(=O)(=O)N1Cc3ccccc3CCN21", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, amine, nitro, sulfonamide. Its descriptor profile: MW=331.35, LogP=2.08, TPSA=83.76, HBD=0, HBA=5, QED=0.59.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "O=[N+]([O-])c1ccc2c(c1)S(=O)(=O)N1Cc3ccccc3CCN21", "compound_name": null, "pubchem_cid": null, "MW": 331.35, "LogP": 2.08, "TPSA": 83.76, "QED": 0.59, "functional_groups": ["aromatic ring", "amine", "nitro", "sulfonamide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00786", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMASKNSINKPKIKINAHSHAKTLGRKRAARARKNPSTKSSTSRYAEVSGTATKPSDSTAIALYTGEAPKPTGVVTNTTLSNKRAKKLARNQKYIDQRKNEGLGKEHEMDVEVETEEQRQSRLEQIKKALWSVVQNQSEGLYKVAVDAEGTTLGVQAF\n\n### Sequence 2:\nMTTIVSVRREGKVVVAGDGQASQGDMIAKGNVKKVRRLYNDSVLVGFAGSTADAFILFDLCERKLEMHQGNLTKAAVELAKEWRSDRALRRLEAMLIVADKTTSLIISGTGDLINADNDLLTIGSGGYFARSAATALLENTDLDAREIATKALTIAGDIDVYTNHNHTVEELNA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00787", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPPSIGIAVILVGTSDEVAIKDLSVVPRVELVAMNETDPKSIITRICDLMSDRKIQGVVFADDTDQEAIAQILDFISAQTLTPILGIHGGSSMIMADKDESSMFFQFGPSIEQQASVMLNIMEEYDWYIFSIVTTYFPGYQDFVNKIRSTIENSFVGWELEEVLLLDMSLDDGDSKIQNQLKKLQSPIILLYCTKEEATYIFEVANSVGLTGYGYTWIVPSLVAGDTDTVPSEFPTGLISVSYDEWDYGLPARVRDGIAIITTAASDMLSEHSFIPEPKSSCQSNMLNRYLINVTFEGRDLSFSEDGYQMHPKLVIILLNKERKWERV\n\n### Answer:", "output": "<SEQ_3Di>DEEAEEEEEEEDDDCLDVLPDDQYHYHYDYHYFPDDDLVRVVVVVLVVVVVDWHAAYEYEYQELQLCSCVNVLVVQLPWLHAYEQEYFSSQQDAPPGDPRGNYFYAWDDLLVFLVVVVVVCVVVVLLEEEEEEEPHPCVVVNVVSVVVVCVPDPDNRCNPYYDYDYPVDDCPVVPLLVRVVSDDRQEYEYEDAQVVVLVSLLSCVVVCQLPPSHEYEYEARNCGNLADDDPSDRFFYKYKTQPCVVCDDSNSVVVVRQLVRQLVSVCCVPVVDDDTGDSPCPDSVSVVSNQFDDDPPDGQHADPSRYGPDGHMFMWGCDPVRGTDGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00788", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "NC1=c2[nH]c(-c3ccc([N+](=O)[O-])o3)nc2=NS(=O)(O)=N1", "output": "This molecule is medium-sized and polar, with hydrophilic character. It contains amine, hydroxyl, nitro, imidazole. Its descriptor profile: MW=298.24, LogP=-0.91, TPSA=173.0, HBD=3, HBA=7, QED=0.47.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "NC1=c2[nH]c(-c3ccc([N+](=O)[O-])o3)nc2=NS(=O)(O)=N1", "compound_name": null, "pubchem_cid": null, "MW": 298.24, "LogP": -0.91, "TPSA": 173.0, "QED": 0.47, "functional_groups": ["amine", "hydroxyl", "nitro", "imidazole"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00789", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nB2MG_HORSE\n\n### Answer:", "output": "Belongs to the beta-2-microglobulin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00790", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nK1A_ANEER\n\n### Answer:", "output": "Inhibits voltage-gated potassium channels (Kv1/KCNA) (PubMed:16905168).  Blocks the binding of dendrotoxin to rat synaptosomal membrane in a dose-dependent manner (IC(50)=91 nM) (PubMed:16905168).  Inhibits Kv1. 1/KCNA1 and Kv1. 3/KCNA3 with high potency (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00791", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nSLC32A1 GAD2 GAD1\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as GAD1, SLC32A1, GAD2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "lung", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SLC32A1", "GAD2", "GAD1"]}}
{"id": "eval_00792", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "MYH6, TNNT2, TNNI3", "output": "The expression pattern is consistent with a healthy cardiomyocyte without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "pancreas", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["MYH6", "TNNT2", "TNNI3"]}}
{"id": "eval_00793", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKTVFAILFLAFIALTYARSYEDVKEEIKNEVEKEILEDLEEESDELNDKRKEINDAKPWRWVRRIRWKKLIPYIPVVVAAAGKK\n\n### Sequence 2:\nQDLPPLDPSAEQAQALNYVKDTAEAADHPAHQEGEQCDNCMFFQADSQGCQLFPQNSVEPAGWCQSWTAQN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00794", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLEPFYVERYSWSQLKKLLADTRKYHGYMMAKAPHDFMFVKRNDPDGPHSDRIYYLAMSGENRENTLFYSEIPKTINRAAVLMLSWKPLLDLFQVGIASYDYHQGSGTFLFQAGSGIYHVKDGGPQGFTQQPLRPNLVETSCPNIRMDPKLCPADPDWIAFIHSNDIWISNIVTREERRLTYVHNELANMEEDARSAGVATFVLQEEFDRYSGYWWCPKAETTPSGGKILRILYEENDESEVEIIHFRYPKTGTANPKVTFKMSEIMIDAEGRIIDVIDKELIQPFEILFEGVEYIARAGWTPEGKYAWSILLDRSQTRLQIVLISPELFIPVEDDVMERQRLIESVPDSVTPLIIYEETTDIWINIHDIFHVFPQSHEEEIEFIFASECKTGFRHLYKITSILKESKYKRSSGGLPAPSDFKCPIKEEIAITSGEWEVLGRHGSNIQVDEVRRLVYFEGTKDSPLEHHLYVVSYVNPGEVTRLTDRGYSHSCCISQHCDFFISKYSNQKNPHCVSLYKLSSPEDDPTCKTKEFWATILDSAGPLPDYTPPEIFSFESTTGFTLYGMLYKPHDLQPGKKYPTVLFIYGGPQVQLVNNRFKGVKYFRLNTLASLGYVVVVIDNRGSCHRGLKFEGAFKYKMGQIEIDDQVEGLQYLASRYDFIDLDRVGIHGWSYGGYLSLMALMQRSDIFRVAIAGAPVTLWIFYDTGYTERYMGHPDQNEQGYYLGSVAMQAEKFPSEPNRLLLLHGFLDENVHFAHTSILLSFLVRAGKPYDLQIYPQPESGEHYELHLLHYLQENLGSRIAALKV\n\n### Answer:", "output": "<SEQ_3Di>DAAAFFDADDLVVLLVVCVVPDDAFPLDAWFDWDPKAWFAQPDVPDFFGTKIKTWTAPHRPGAIAIWIWTFGPDDDVVDFFATDTAGQADDPCAHWPDWDAAQAFQWIWTDGPQWIWIDGERHPVHHDPHHDYTHTLDDPAPDGWADKDAQLADRQKIWTDDQQFIKMAGSPVGDIGTQDDQDDVVDDCQVRVKHKFADDDCLVPQLVDGTFKEWQNHWAADPQQWTKTKMKGKIWHQNQADDDCPGFFFFQGDDTDIAIKMWIWTAGNPGHTPDIFIWGFLDDPCQQPPQFDDFNDWYAAPNRQWIWTWTAHLLQFKIWIKTAGPLRIDGDDPPPVVSVVSSVPRFQLRGIAGQDMAGDPAHFRDFQEKHWDDDPDPFKTWIWDWDCRPPQATFIKIWMWGRDRFPDGCVVVDTDDPCRRYTHIPDIDTQDDDQKYWDGDDQWYWAADVPLQKIWTWIQVVHLQAIFIWMAHRPDHDDIGTADDPPKHWDKDAYNNRQKMWIFIAFAQWATWIFIKGWDDDPVDRRDIYIDTRHTHGDTPGRDPLFDGWDKDWDQDPVGDIKIKTKDAAPPDDPPYAAAEEEEDQQAAPDDPHHGHHPCSSPVCVRVVSPSTYIYMYIHQFLYDGRGSVSNSPCALAALNPRVVRVVVVVVVVVVVDVRHDQVFYAYEYEESSLLNLLSCQLVPVVRHQAGERELYQQASSRAHSSPCSSRNNRCVVRVPSRCNSHNLNVLLSHDQDELRAEYEYECPDPHRHCVRVVSSVVSNVVSNHYYHYDYDYPSCVVSVVSSVVVSSCSCQGRRPNNVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00795", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRSMA_MESH2\n\n### Answer:", "output": "Belongs to the class I-like SAM-binding methyltransferase superfamily. rRNA adenine N(6)-methyltransferase family. RsmA subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00796", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSELTTVARPYAKAAFDFALEQGALDKWAEMLSFAAAVAQDETIASFLSSSSTVGKTTEVFLGVCGDELDDNAKNFVKVLAENERLPVLPAVSELYQTLRAEHDKEVTVDVKSAVKLLKAQQTALIKALEKRLQRKIKLNCSVDKSIIGGLVIEAGDTVIDGTLRGKLDRLAYALQS\n\n### Sequence 2:\nMQVILLDKVANLGSLGDQVNVKAGYARNFLVPQGKAVPATKKNVEYFEARRAELEAKLADVLAAANARAEKINALETVTIASKAGDEGKLFGSIGTRDIADAVTAAGVDVAKSEVRLPNGVLRTTGEHEVNFQVHSEVFAKVIINVVAE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00797", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nISPF_FUSNN\n\n### Answer:", "output": "Involved in the biosynthesis of isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP), two major building blocks of isoprenoid compounds.  Catalyzes the conversion of 4-diphosphocytidyl-2-C-methyl-D-erythritol 2-phosphate (CDP-ME2P) to 2-C-methyl-D-erythritol 2,4-cyclodiphosphate (ME-CPP) with a corresponding release of cytidine 5-monophosphate (CMP).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00798", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nUP355_STAEQ\n\n### Answer:", "output": "UPF0355 protein SERP0066.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00799", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nTFF3 CLCA1 MUC2\n</cell>", "output": "This cell is most consistent with goblet cell, an intestinal goblet cell. Key markers such as TFF3, CLCA1, MUC2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "goblet cell", "tissue": "pancreas", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TFF3", "CLCA1", "MUC2"]}}
{"id": "eval_00800", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMLDINQIKEIIPHRYPFLLVDRILSVEDGRKAVGLKNVSANEPYFQGHFPGYPVMPGVLIIEAMAQVGAVAVLRLPEFAGKMAFFAGIDRARFRRQVVPGDQLRIEVELQKLRGTVGKARGAAYVGQELAAEAELMFAVR\n\n### Sequence 2:\nMDLKAKLREVPDFPKEGINFIDITTVLQDPEALKECIDSMKKKVESFGEFDIIVGAESRGFIFGAPLAYALGKGFVPIRKKGKLPYKTISVEYELEYGTDILEMHIDAIKKGQRVVIVDDLLATGGTTKSNIKLVEQLGGEILGIVYFVELTFLNGREKLKGYNVESIVQY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00801", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "CD163, APOE, SPP1", "output": "The expression pattern is consistent with tumor-associated macrophage in a context associated with COVID-19, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "pancreas", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CD163", "APOE", "SPP1"]}}
{"id": "eval_00802", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: C[C@H](Nc1ncc(F)c(Nc2cc(C3CC3)[nH]n2)n1)c1ncc(F)cn1\nDescriptors: MW=358.36, LogP=3.06, TPSA=104.3, HBD=3, HBA=7", "output": "Descriptors: MW=358.36, LogP=3.06, TPSA=104.3, HBD=3, HBA=7, QED=0.62, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "C[C@H](Nc1ncc(F)c(Nc2cc(C3CC3)[nH]n2)n1)c1ncc(F)cn1", "compound_name": null, "pubchem_cid": null, "MW": 358.36, "LogP": 3.06, "TPSA": 104.3, "QED": 0.62, "functional_groups": ["halide"], "moleculenet_labels": {"label": 2.47}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00803", "category": "Mutation", "instruction": "Describe the effect of mutation Q66K64-D453V on protein DDB1- and CUL4-associated factor 15.", "input": "### Wild-type:\nMAPSSKSERNSGAGSGGGGPGGAGGKRAAGRRREHVLKQLERVKISGQLSPRLFRKLPPRVCVSLKNIVDEDFLYAGHIFLGFSKCGRYVLSYTSSSGDDDFSFYIYHLYWWEFNVHSKLKLVRQVRLFQDEEIYSDLYLTVCEWPSDASKVIVFGFNTRSANGMLMNMMMMSDENHRDIYVSTVAVPPPGRCAACQDASRAHPGDPNAQCLRHGFMLHTKYQVVYPFPTFQPAFQLKKDQVVLLNTSYSLVACAVSVHSAGDRSFCQILYDHSTCPLAPASPPEPQSPELPPALPSFCP\n\n### Mutant:\nMAPSSKSERNSGAGSGGGGPGGAGGKRAAGRRREHVLKQLERVKISGQLSPRLFRKLPPRVCVSLKNIVDEDFLYAGHIFLGFSKCGRYVLSYTSSSGDDDFSFYIYHLYWWEFNVHSKLKLVRQVRLFQDEEIYSDLYLTVCEWPSDASKVIVFGFNTRSANGMLMNMMMMSDENHRDIYVSTVAVPPPGRCAACQDASRAHPGDPNAQCLRHGFMLHTKYQVVYPFPTFQPAFQLKKDQVVLLNTSYSLVACAVSVHSAGDRSFCQILYDHSTCPLAPASPPEPQSPELPPALPSFCP\n\n### Answer:", "output": "Increased interaction with DDA1 and RBM39 in presence of indisulam.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00804", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSLTQLQERIEDLECKLAFQEQTIETLNDALTQQQLLLSKMQDQMKYVVGKVKNMDTSTLADPAHETPPPHY\n\n### Sequence 2:\nTIALLENRINDLECQVAFQEQTIEELNDALTQQQLLIAKMQDQMKYVVGKMKNMDSSNMVDPAKEPPPPHY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00805", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CC(C)CC(NC(=O)C(CCCCN)NC(=O)CNC(=O)C(CC(C)C)NC(=O)C(NC(=O)C1CSSCC(N)C(=O)NC(CO)C(=O)NC(CC(N)=O)C(=O)NC(CC(C)C)C(=O)NC(CO)C(=O)NC(C(C)O)C(=O)N1)C(C)C)C(=O)NC(CO)C(=O)NC(CCC(N)=O)C(=O)NC(CCC(=O)O)C(=O)NC(CC(C)C)C(=O)NC(Cc1cnc[nH]1)C(=O)NC(CCCCN)C(=O)NC(CC(C)C)C(=O)NC(CCC(N)=O)C(=O)NC(C(=O)NC(Cc1ccc(O)cc1)C(=O)N1CCCC1C(=O)NC(CCCNC(=N)N)C(=O)NC(C(=O)NC(CC(N)=O)C(=O)NC(C(=O)NCC(=O)NC(CO)C(=O)NCC(=O)NC(C(=O)N1CCCC1C(N)=O)C(C)O)C(C)O)C(C)O)C(C)O", "output": "This molecule is relatively large and polar, with hydrophilic character. It contains aromatic ring, amide, amine, hydroxyl, phenol. Its descriptor profile: MW=3431.91, LogP=-21.53, TPSA=1508.21, HBD=52, HBA=54, QED=0.01.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "SIDER", "canonical_smiles": "CC(C)CC(NC(=O)C(CCCCN)NC(=O)CNC(=O)C(CC(C)C)NC(=O)C(NC(=O)C1CSSCC(N)C(=O)NC(CO)C(=O)NC(CC(N)=O)C(=O)NC(CC(C)C)C(=O)NC(CO)C(=O)NC(C(C)O)C(=O)N1)C(C)C)C(=O)NC(CO)C(=O)NC(CCC(N)=O)C(=O)NC(CCC(=O)O)C(=O)NC(CC(C)C)C(=O)NC(Cc1cnc[nH]1)C(=O)NC(CCCCN)C(=O)NC(CC(C)C)C(=O)NC(CCC(N)=O)C(=O)NC(C(=O)NC(Cc1ccc(O)cc1)C(=O)N1CCCC1C(=O)NC(CCCNC(=N)N)C(=O)NC(C(=O)NC(CC(N)=O)C(=O)NC(C(=O)NCC(=O)NC(CO)C(=O)NCC(=O)NC(C(=O)N1CCCC1C(N)=O)C(C)O)C(C)O)C(C)O)C(C)O", "compound_name": null, "pubchem_cid": null, "MW": 3431.91, "LogP": -21.53, "TPSA": 1508.21, "QED": 0.01, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "phenol", "imidazole"], "moleculenet_labels": {"Hepatobiliary disorders": 1, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 1, "General disorders and administration site conditions": 1, "Endocrine disorders": 1, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 1, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 1, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00806", "category": "Mutation", "instruction": "Describe the effect of mutation P61212-T31N on protein ADP-ribosylation factor-like protein 1.", "input": "### Wild-type:\nMGGFFSSIFSSLFGTREMRILILGLDGAGKTTILYRLQVGEVVTTIPTIGFNVETVTYKNLKFQVWDLGGQTSIRPYWRCYYSNTDAVIYVVDSCDRDRIGISKSELVAMLEEEELRKAILVVFANKQDMEQAMTPSEMANALGLPALKDRKWQIFKTSATKGTGLDEAMEWLVETLKSRQ\n\n### Mutant:\nMGGFFSSIFSSLFGTREMRILILGLDGAGKNTILYRLQVGEVVTTIPTIGFNVETVTYKNLKFQVWDLGGQTSIRPYWRCYYSNTDAVIYVVDSCDRDRIGISKSELVAMLEEEELRKAILVVFANKQDMEQAMTPSEMANALGLPALKDRKWQIFKTSATKGTGLDEAMEWLVETLKSRQ\n\n### Answer:", "output": "Accumulation of GDP-bound form. Over-expression of Arl1(), which is expected to be restricted to the GDP-bound form and thus function as a dominant-negative mutant, causes the disappearance of the Golgi apparatus (marked by Golgi SNARE GS28), suggesting that Arl1 is necessary for maintaining normal Golgi structure.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00807", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "SFTPB, SFTPC, SFTPA1, ABCA3", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as SFTPA1, SFTPC, SFTPB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "tumor", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SFTPB", "SFTPC", "SFTPA1", "ABCA3"]}}
{"id": "eval_00808", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSFNIQSDDLLHHFEADSNDTLLSAALRAELVFPYECNSGGCGACKIELLEGEVSNLWPDAPGLAARELRKNRFLACQCKPLSDLKIKVINRAEGRASHPPKRFSTRVVSKRFLSDEMFELRLEAEQKVVFSPGQYFMVDVPELGTRAYSAANPVDGNTLTLIVKAVPNGKVSCALANETIETLQLDGPYGLSVLKTADETQSVFIAGGSGIAPMVSMVNTLIAQGYEKPITVFYGSRLEAELEAAETLFGWKENLKLINVSSSVVGNSESSYPTGYVHEIIPEYMEGLLGAEFYLCGPPQMINSVQKLLMIENKVPFEAIHFDRFF\n\n### Answer:", "output": "<SEQ_2D>CEEEECTTSSSCEEECTTSCHHHHHHHTTCCCCCSCSSSSSCTTEEEEEESCEEESCTTCTTSCHHHHHTTEEEGGGEEESSCEEEEECSCCCCCCSSCCEEEEEEEEEEEESSSSEEEEEEECSSCCCCCTTCEEEEEETTTEEEEEECCSCCCTTEEEEEEECCTTCHHHHHHHHSCCSEEEEEEEECSCCCCCSSCSCEEEEEEGGGHHHHHHHHHHHHHTTCCSCEEEEEEESCHHHHHHHHHHSCCCTTEEEEEEECCSSCCTTSCCCBSCHHHHGGGGGGGGTTCEEEEEECHHHHHHHHHHHHTTSCCCGGGEEEEECC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00809", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPMIVFVRFNSSHGFPVEVDSDTSIFQLKEVVAKRQGVPADQLRVIFAGKELRNDWTVQNCDLDQQSIVHIVQRQSLTRVDLSYNSFYVYCKGPCQRVQPGKLRVQCSTCRQATLTLTQGPSCWDDVLIPNRMSGECQSPHCPGTSAEFFFKCGAHPTSDKETSVALHLIATNSRNITCITCTDVRSPVLVFQCNSRHVICLDCFHLYCVTRLNDRQFVHDPQLGYSLPCVAGCPNSLIKELHHFRILGEEQYNRYQQYGAEECVLQMGGVLCPRPGCGAGLLPEPDCRKVTCENGLGCGFAFCRECKEAYHEGECS\n\n### Answer:", "output": "<SEQ_3Di>DAWEWEQELNHDTDIDDDDQFAFPLNVLVVVCVVVVHDSVFKWKAFQQATGDRNDGCVNLVPDHHEYIYIYGNVGQQVAAAVVKAWWFQQVPDVLGIAGKGKGKDFPPPRHRQWHWPDDDHHPCCQPPFQNTFTARPNPPDRGTGIHMFMPGPPDHDDPPNTTGTPHQKIFPPVQQAAPPPRHGDPIWGFWPAPVRHIHRLVVVLVVVVVCLVVVNFDQDPVQFTFDAPPVPGPPGTDRPVCVNSSNPDPVVVSSVVVRVVVVLVVLVWAQQPPPPGRDRDDDPNPDQWDAQDVPRGDGATAGPVPNHGDDPDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00810", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: VEGFR2\nLigand SMILES: CCOC(=O)Cc1cccc2ccccc12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CCOC(=O)Cc1cccc2ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 214.26, "LogP": 2.95, "TPSA": 26.3, "QED": 0.73, "functional_groups": ["aromatic ring", "ester", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_00811", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nDHPLDRPVWNSLGGPQSELDVASGNLRRLDPAYGPFAAAAPGAEAGLASLLQGDADEIWLVEPEPVAPPPGTRVIRVAPLLQMIADGPVPSFDDPGIVALGETDVPEMTALALATEPGPWASGTWRYGQFYGVRIDGRLAAMAGERMRPAPNLAEVSGVCTWPEYRGRGLAARLIRKVIAGMAARGEVPYLHSYASNASAIRLYESLGFRARRAMTATLLGKST\n\n### Answer:", "output": "<SEQ_3Di>DDLVQFQQVLLCVAVVVVQFPDDDQKTAGNVQQFQEIDGHVVGVLCVLVVADDQFRKHKGKHADEDDAHPQKDWPDKFKKWKKWQPDDQDDDDDVQKAKDDDVCQVLLVVQCVQWPQDTHHNNNPVQFTKIFGDDPNHGFWMKGWHGDSYPQEIEITRTTGPPVPPPPCGSLVRVSVHVVVSVVSNHMYMYMDGPPCPPVVVSVVVRRIDTDGMIMMTMMHGDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00812", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNLYESCVRGALRAYKLTLSPLIGRQCRFFPTCSEYAAEALIGHGPVRGSWLTVRRLCRCHPWGPSGWDPPPPPRRKGAKWTCET\n\n### Sequence 2:\nMNLYESCVRGALRAYKLTLSPLIGRQCRFFPTCSEYAAEALIGHGPVRGSWLTVRRLCRCHPWGPSGWDPPPPPRRKGAKWTCET\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00813", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMGTKREAILKVLENLTPEELKKFKMKLGTVPLREGFERIPRGALGQLDIVDLTDKLVASYYEDYAAELVVAVLRDMRMLEEAARLQRAA\n\n### Sequence 2:\nMFLNGISLRLTRRSGMWHLLLNMGPNSISIPLDSSSSIRRRPRWYSVRRHNQVLILHLVASSLHWLITIFPNTQQILCQTLPSLSIVSQDI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00814", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "IKZF2, IL2RA, CD3D, FOXP3", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as CD3D, IKZF2, FOXP3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "spleen", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IKZF2", "IL2RA", "CD3D", "FOXP3"]}}
{"id": "eval_00815", "category": "Mutation", "instruction": "Describe the effect of mutation O00400-S113R on protein Acetyl-coenzyme A transporter 1.", "input": "### Wild-type:\nMSPTISHKDSSRQRRPGNFSHSLDMKSGPLPPGGWDDSHLDSAGREGDREALLGDTGTGDFLKAPQSFRAELSSILLLLFLYVLQGIPLGLAGSIPLILQSKNVSYTDQAFFSFVFWPFSLKLLWAPLVDAVYVKNFGRRKSWLVPTQYILGLFMIYLSTQVDRLLGNTDDRTPDVIALTVAFFLFEFLAATQDIAVDGWALTMLSRENVGYASTCNSVGQTAGYFLGNVLFLALESADFCNKYLRFQPQPRGIVTLSDFLFFWGTVFLITTTLVALLKKENEVSVVKEETQGITDTYKL\n\n### Mutant:\nMSPTISHKDSSRQRRPGNFSHSLDMKSGPLPPGGWDDSHLDSAGREGDREALLGDTGTGDFLKAPQSFRAELSSILLLLFLYVLQGIPLGLAGSIPLILQSKNVSYTDQAFFRFVFWPFSLKLLWAPLVDAVYVKNFGRRKSWLVPTQYILGLFMIYLSTQVDRLLGNTDDRTPDVIALTVAFFLFEFLAATQDIAVDGWALTMLSRENVGYASTCNSVGQTAGYFLGNVLFLALESADFCNKYLRFQPQPRGIVTLSDFLFFWGTVFLITTTLVALLKKENEVSVVKEETQGITDTYKL\n\n### Answer:", "output": "In SPG42; significant increase in the amount of nuclear phosphorylated SMAD1-SMAD5-SMAD8 protein complex; marked increase of the BMPR1A protein level; no change for BMPR2 protein level; decrease of BMPR1A degradation. The missense mutation, causing the change of the highly conserved serine to arginine at the codon 113 (p. ), disrupts the second transmembrane domain in the transporter and reverses the orientation of all of the descending domains. Knockdown of Slc33a1 in zebrafish caused a curve-shaped tail and defective axon outgrowth from the spinal cord. Although the wild-type human SLC33A1 was able to rescue the phenotype caused by Slc33a1 knockdown in zebrafish, the mutant SLC33A1 (p. ) was not, suggesting that mutation renders SLC33A1 nonfunctional and one that wild-type allele is not sufficient for sustaining the outgrowth and maintenance of long motor axons in human heterozygotes. Thus, our study illustrated a critical role of acetyl-CoA transporter in motor-neuron development and function. SLC33A1 p. was found to be the most plausible causal variant in this family. It acts in a dominant-negative manner and can negatively regulate BMP signaling.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00816", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCC(C)SSC(C)CC", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains no standard functional groups detected. Its descriptor profile: MW=178.37, LogP=3.96, TPSA=0.0, HBD=0, HBA=2, QED=0.58.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CCC(C)SSC(C)CC", "compound_name": null, "pubchem_cid": null, "MW": 178.37, "LogP": 3.96, "TPSA": 0.0, "QED": 0.58, "functional_groups": [], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00817", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGSAPWAPVLLLALGLRGLQAGARRAPDPGFQERFFQQRLDHFNFERFGNKTFPQRFLVSDRFWVRGEGPIFFYTGNEGDVWAFANNSAFVAELAAERGALLVFAEHRYYGKSLPFGAQSTQRGHTELLTVEQALADFAELLRALRRDLGAQDAPAIAFGGSYGGMLSAYLRMKYPHLVAGALAASAPVLAVAGLGDSNQFFRDVTADFEGQSPKCTQGVREAFRQIKDLFLQGAYDTVRWEFGTCQPLSDEKDLTQLFMFARNAFTVLAMMDYPYPTDFLGPLPANPVKVGCDRLLSEAQRITGLRALAGLVYNASGSEHCYDIYRLYHSCADPTGCGTGPDARAWDYQACTEINLTFASNNVTDMFPDLPFTDELRQRYCLDTWGVWPRPDWLLTSFWGGDLRAASNIIFSNGNLDPWAGGGIRRNLSASVIAVTIQGGAHHLDLRASHPEDPASVVEARKLEATIIGEWVKAARREQQPALRGGPRLSL\n\n### Sequence 2:\nAPWATIALRPALRALGSLHLPTNPTSLPAVAKNYSVLYFQQKVDHFGFNTV--KTFNQRYLVADKYWKKNGGSILFYTGNEGDIIWFCNNTGFMWDVAEELKAMLVFAEHRYYGESLPFGDNSFKDSRHLNFLTSEQALADFAELIKHLKRTIPGAENQPVIAIGGSYGGMLAAWFRMKYPHMVVGALAASAPIWQFEDLVPCGVFMKIVTTDFRKSGPHCSESIHRSWDAINRLSNTGS--GLQWLTGALHLCSPLTSQ-DIQHLKDWISETWVNLAMVDYPYASNFLQPLPAWPIKVVCQYLKNPNVSDSLLLQNIFQALNVYYNYSGQVKCLNISETATSSLGTLG---------WSYQACTEVVMPFCTNGVDDMFEPHSWNLKELSDDCFQQWGVRPRPSWITTMYGGKNISSHTNIVFSNGELDPWSGGGVTKDITDTLVAVTISEGAHHLDLRTKNALDPMSVLLARSLEVRHMKNWIR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00818", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNSESVLEEIVLELYRNKLFKIGEYRLTSGKISPYYIDLRILPSYYDIYSKIIDISLSKLEKLNFDIVVGIESAGIIHASFIACKTHKPVGYVRKKPKQHGTKRLVEGIVADRNVLVVDDVATTGGSLEHAVNAVRSMGGIVEKAFVFVDREEGARERLKRLGVELISLMNIWFIINVLVKHRLIDEQTYYLIKNYLNREKHS\n\n### Sequence 2:\nMSIKYVATSKLPTPWGVFAMHGFEDTETGKEHVALTFGTLSADEPVLGRIHSECLTGDALFSLRCDCGFQLQTAMQNIAETGCGFILYLRQEGRGIGLLNKIRAYELQDKGANTVEANEQLGFPADMRKYDMIKPMLEQIGVKHVRLMTNNPRKVKAMKEFGIEVVERVPLQVGKNRYNEAYLKTKSTELGHMMSEYHFVDED\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00819", "category": "Mutation", "instruction": "Describe the effect of mutation Q8VZS8-N117R on protein Abscisic acid receptor PYL1.", "input": "### Wild-type:\nMANSESSSSPVNEEENSQRISTLHHQTMPSDLTQDEFTQLSQSIAEFHTYQLGNGRCSSLLAQRIHAPPETVWSVVRRFDRPQIYKHFIKSCNVSEDFEMRVGCTRDVNVISGLPANTSRERLDLLDDDRRVTGFSITGGEHRLRNYKSVTTVHRFEKEEEEERIWTVVLESYVVDVPEGNSEEDTRLFADTVIRLNLQKLASITEAMNRNNNNNNSSQVR\n\n### Mutant:\nMANSESSSSPVNEEENSQRISTLHHQTMPSDLTQDEFTQLSQSIAEFHTYQLGNGRCSSLLAQRIHAPPETVWSVVRRFDRPQIYKHFIKSCNVSEDFEMRVGCTRDVNVISGLPARTSRERLDLLDDDRRVTGFSITGGEHRLRNYKSVTTVHRFEKEEEEERIWTVVLESYVVDVPEGNSEEDTRLFADTVIRLNLQKLASITEAMNRNNNNNNSSQVR\n\n### Answer:", "output": "Forms homodimer and exhibits an enhanced ABA affinity; when associated with R-112. The mutation in PYLs/RCARs protein may affect the ABA-binding affinity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00820", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "ALB, HNF4A, APOB, TTR", "output": "The marker profile is enriched for xenobiotic metabolism and plasma protein synthesis, consistent with hepatocyte function.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "colon", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_pathway", "top_genes": ["ALB", "HNF4A", "APOB", "TTR"]}}
{"id": "eval_00821", "category": "Mutation", "instruction": "Describe the effect of mutation Q76K71-Q143E on protein Aliphatic aldoxime dehydratase.", "input": "### Wild-type:\nMESAIGEHLQCPRTLTRRVPDTYTPPFPMWVGRADDALQQVVMGYLGVQFRDEDQRPAALQAMRDIVAGFDLPDGPAHHDLTHHIDNQGYENLIVVGYWKDVSSQHRWSTSTPIASWWESEDRLSDGLGFFREIVAPRAEQFQTLYAFQEDLPGVGAVMDGISGEINEHGYWGSMRERFPISQTDWMQASGELRVIAGDPAVGGRVVVRGHDNIALIRSGQDWADAEADERSLYLDEILPTLQSGMDFLRDNGPAVGCYSNRFVRNIDIDGNFLDLSYNIGHWASLDQLERWSESHPTHL\n\n### Mutant:\nMESAIGEHLQCPRTLTRRVPDTYTPPFPMWVGRADDALQQVVMGYLGVQFRDEDQRPAALQAMRDIVAGFDLPDGPAHHDLTHHIDNQGYENLIVVGYWKDVSSQHRWSTSTPIASWWESEDRLSDGLGFFREIVAPRAEQFETLYAFQEDLPGVGAVMDGISGEINEHGYWGSMRERFPISQTDWMQASGELRVIAGDPAVGGRVVVRGHDNIALIRSGQDWADAEADERSLYLDEILPTLQSGMDFLRDNGPAVGCYSNRFVRNIDIDGNFLDLSYNIGHWASLDQLERWSESHPTHL\n\n### Answer:", "output": "Retains 86% of wild-type activity with Z-phenylacetaldoxime as substrate. This mutation may lead to functional changes in the enzyme's substrate recognition and catalysis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00822", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTRMFO_STRPB\n\n### Answer:", "output": "Belongs to the MnmG family. TrmFO subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00823", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDSBA_BORA1\n\n### Answer:", "output": "Belongs to the thioredoxin family. DsbA subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00824", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "C1QB, CD163, CD68, C1QA", "output": "This cell is most consistent with macrophage, a tissue macrophage. Key markers such as CD68, C1QA, CD163 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "prostate", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["C1QB", "CD163", "CD68", "C1QA"]}}
{"id": "eval_00825", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "COc1cc(OC)c(S(=O)(=O)N(c2ccc(C)cc2)c2ccc(C)cc2)cc1NC(=O)CCC(=O)O", "output": "This molecule is relatively large and polar, with lipophilic character. It contains aromatic ring, amide, hydroxyl, ether, sulfonamide. Its descriptor profile: MW=512.58, LogP=4.65, TPSA=122.24, HBD=2, HBA=6, QED=0.41.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "COc1cc(OC)c(S(=O)(=O)N(c2ccc(C)cc2)c2ccc(C)cc2)cc1NC(=O)CCC(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 512.58, "LogP": 4.65, "TPSA": 122.24, "QED": 0.41, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether", "sulfonamide"], "moleculenet_labels": {"label": 0.06}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00826", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMDQQRFTHLLQLEAESIHIIREVAAEFSNPVMMYSIGKDSSVMLHLARKAFHPGKLPFPLLHVDTGWKFREMYAFRDHTAKAYGFELLVHKNPEGVAMGINPFVHGSAKHTDIMKNEGLKQALNQYGFDAAFGGARRDEEKSRAKERIYSFRDRFHRWDPKNQRPELWHNYNGQINKGESIRVFPLSNWTELDIWQYIFLENIEIVPLYLAKKRPVLERDGMLIMVDDDRIDLQPGEVIEQRMVRFRTLGCWPLTGAVASDAQTLPEIIEEMLVSTTSERQGRVIDRDQSGSMEMKKRQGYF\n\n### Sequence 2:\nMLIIETLPLLRQHIRRLRQEGKRIALVPTMGNLHDGHMKLVDEARARADVVVVSIFVNPMQFDRADDLARYPRTLQEDCEKLKKRHVDFVFSPVPSDIYPQGTEGATYVDVPGISTMLEGASRPGHFRGVSTIVSKLFNLVQPDIACFGEKDFQQLALIRKMVADMGYDIEIVGVPIVRAKDGLALSSRNGYLTADQRKIAPGLCKVMNAMAEQLKAKELTTEEIVALAEQELNDKGLRADDIQIRDADTLLELSATSKRAVILVAAWLGQARLIDNKVVELAE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00827", "category": "Mutation", "instruction": "Describe the effect of mutation P19447-T119P on protein General transcription and DNA repair factor IIH helicase subunit XPB.", "input": "### Wild-type:\nMGKRDRADRDKKKSRKRHYEDEEDDEEDAPGNDPQEAVPSAAGKQVDESGTKVDEYGAKDYRLQMPLKDDHTSRPLWVAPDGHIFLEAFSPVYKYAQDFLVAIAEPVCRPTHVHEYKLTAYSLYAAVSVGLQTSDITEYLRKLSKTGVPDGIMQFIKLCTVSYGKVKLVLKHNRYFVESCHPDVIQHLLQDPVIRECRLRNSEGEATELITETFTSKSAISKTAESSGGPSTSRVTDPQGKSDIPMDLFDFYEQMDKDEEEEEETQTVSFEVKQEMIEELQKRCIHLEYPLLAEYDFRND\n\n### Mutant:\nMGKRDRADRDKKKSRKRHYEDEEDDEEDAPGNDPQEAVPSAAGKQVDESGTKVDEYGAKDYRLQMPLKDDHTSRPLWVAPDGHIFLEAFSPVYKYAQDFLVAIAEPVCRPTHVHEYKLPAYSLYAAVSVGLQTSDITEYLRKLSKTGVPDGIMQFIKLCTVSYGKVKLVLKHNRYFVESCHPDVIQHLLQDPVIRECRLRNSEGEATELITETFTSKSAISKTAESSGGPSTSRVTDPQGKSDIPMDLFDFYEQMDKDEEEEEETQTVSFEVKQEMIEELQKRCIHLEYPLLAEYDFRND\n\n### Answer:", "output": "In TTD2; mild. The causative mutation in the XPB protein is associated with Trichothiodystrophy (TTD) and is part of the TFIIH complex, a basal transcription factor with a second function in DNA repair. This mutation results in a missense mutation in a conserved region of the XPB protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00828", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "CX3CR1, P2RY12, CSF1R", "output": "This profile is most consistent with microglia, which is commonly found in heart under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "heart", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CX3CR1", "P2RY12", "CSF1R"]}}
{"id": "eval_00829", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "O=C(NCCCCCN1CCC2(CCCCC2)CC1)C(c1ccccc1)C1CCCCC1.O=C(O)C(=O)O", "output": "The SMILES represents an unnamed molecule with MW=528.73, LogP=5.84, TPSA=106.94. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "O=C(NCCCCCN1CCC2(CCCCC2)CC1)C(c1ccccc1)C1CCCCC1.O=C(O)C(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 528.73, "LogP": 5.84, "TPSA": 106.94, "QED": 0.27, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00830", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRK22_FAGEA\n\n### Answer:", "output": "The globular domain of the protein is located near the polypeptide exit tunnel on the outside of the subunit, while an extended beta-hairpin is found that lines the wall of the exit tunnel in the center of the 70S ribosome.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00831", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nVNPD_OXYSC\n\n### Answer:", "output": "Natriuretic peptide OsNP-d.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00832", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nASLAIGGVVIIGGGGHAKVVIESLRACGETVAAIVDADPTRAVLGVPVVGDDLALPMLREQGLSRLFVAIGDNRLRQKLGRKARDHGFSLVNAIHPSAVVSPSVRLGEGVAVMAGVAINADSWIGDLAIINTGAVVDADCRLGAACHLGPASALAGGVSVGERAFLGVGARVIPGVTIGADTIVGAGGVVVRDLPDSVLAIGVPAKIKGD\n\n### Answer:", "output": "<SEQ_3Di>DDWDPDAEEEEDQDPLVLQLCVLCVLVVGHYTAYAYPVQDDAHVNHYRHYHPVCLLVCVVVPHAEYAYRDLQQVVSVVVVVVSVVSRHFAAERAHPQEAEDPQEAEEGNEYHGHQEYAEHNEYAEHNEYEDANEYHYYQEYEYASDYEEHCEYEEYQEYEERNEYEYAHEYEYHNEYEYHCEYEYHNEYDDHHHYHNFYFYDVPTDTDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00833", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "ALB, TTR, HNF4A", "output": "This profile is most consistent with hepatocyte, which is commonly found in breast under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "breast", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["ALB", "TTR", "HNF4A"]}}
{"id": "eval_00834", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nAGLQFPVGRIARFLKAGKYAERVGAGAPVYLAAVLEYLAAEVLELAGNAARDNKKTRIVPRHIQLAVRNDEELSKLLGDVTIA\n\n### Answer:", "output": "<SEQ_3Di>DDQDFDLPVQQVCVCVVCPDPDDDSVNSVVVSVVVRVVVVQLVVQLVVQCVVVVHDDRDQVSSQCSQVVDPVSCVVPPHRDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00835", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "Cc1nc2sccn2c(=O)c1CCN1CCC(=C(c2ccc(F)cc2)c2ccc(F)cc2)CC1", "output": "The SMILES represents an unnamed molecule with MW=477.58, LogP=5.48, TPSA=37.61. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "Cc1nc2sccn2c(=O)c1CCN1CCC(=C(c2ccc(F)cc2)c2ccc(F)cc2)CC1", "compound_name": null, "pubchem_cid": null, "MW": 477.58, "LogP": 5.48, "TPSA": 37.61, "QED": 0.38, "functional_groups": ["aromatic ring", "amine", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": null, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 1.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00836", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nATPB_ECO81\n\n### Answer:", "output": "ATP synthase subunit beta, ATP synthase F1 sector subunit beta, F-ATPase subunit beta.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00837", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGN_SHV21\n\n### Answer:", "output": "Envelope glycoprotein N.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00838", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSQRKIQKVMVQPVNLIFRYLQNRTRVQIWLYEDVTHRLEGYIIGFDEFMNVVFDEAEEVNMKTKGRNKIGRILLKGDNITLIHAAAQEA\n\n### Sequence 2:\nMSNKVKTKAMVPPINCIFNFLQQQTPVTIWLFEQIGIRIKGKIVGFDEFMNVVIDEAVEIPVNSADGKEDVEKG-TPLGKILLKGDNITLITSA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00839", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "LUM, PDGFRA, COL1A2, DCN, COL1A1", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as DCN, PDGFRA, COL1A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["LUM", "PDGFRA", "COL1A2", "DCN", "COL1A1"]}}
{"id": "eval_00840", "category": "Mutation", "instruction": "Describe the effect of mutation B2UQG6-D326A on protein Beta-hexosaminidase Amuc_0868.", "input": "### Wild-type:\nMISKCTFSATVFSLFSLCWGAPSSPVLEAPHTIPLPAAMRVQTGESGFSLKNGVRLPEKNPLSRQAERIFRDNGINTALVKNNADIIFTEDASLGREGYRLAVTPDSISIASGSVNGTLYALQSLVQSIAADKNGAPALPRMDVKDQPRFSWRGLMVDSCRHMMPVRDIKKVLDLMERYKFNTLHWHLTDDQGWRLPIAKYPRLTTVGGARAQSPVIGNRNKGDGIPYSGHYTADEIRDVVRYARDRGITVIPEVEMPGHASAAIAAYPELGNTDIPGYEPRVQETWGVHSYTFSPTEKT\n\n### Mutant:\nMISKCTFSATVFSLFSLCWGAPSSPVLEAPHTIPLPAAMRVQTGESGFSLKNGVRLPEKNPLSRQAERIFRDNGINTALVKNNADIIFTEDASLGREGYRLAVTPDSISIASGSVNGTLYALQSLVQSIAADKNGAPALPRMDVKDQPRFSWRGLMVDSCRHMMPVRDIKKVLDLMERYKFNTLHWHLTDDQGWRLPIAKYPRLTTVGGARAQSPVIGNRNKGDGIPYSGHYTADEIRDVVRYARDRGITVIPEVEMPGHASAAIAAYPELGNTDIPGYEPRVQETWGVHSYTFSPTEKT\n\n### Answer:", "output": "Specific activity is decreased by 99% with a 145-fold reduction in kcat/Km value compared to that of wild-type. mutation affects the catalytic activity of Am0868.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00841", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRNPA_SHIBS\n\n### Answer:", "output": "Ribonuclease P protein component, RNase P protein, RNaseP protein, Protein C5.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00842", "category": "Mutation", "instruction": "Describe the effect of mutation Q2LKU9-R660H on protein NACHT, LRR and PYD domains-containing protein 1a.", "input": "### Wild-type:\nMEESQSKQESSTKVAQHEGQEDVDPTFKTKKLMEVELMKHRVQLERNLKLRTFPGARTKQVKEALYPLLTWSSKSKNLFQNFTKLLLFKKLCQRGSENLVRESWYPCVPEEEAHMIDIQDLFGPNLGTQKKPQLVIIEGAAGIGKSTLARLVKRAWKEGKLYRNDFHHVFFFSCRELAQYEQLSLAELIVQGQEVPTAPIRQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQTQPVHTLLGSLLGKSILPGASFLLTTRTTALQKFIPSLEQPCQVEVLGFTLFERKNYFYKYFG\n\n### Mutant:\nMEESQSKQESSTKVAQHEGQEDVDPTFKTKKLMEVELMKHRVQLERNLKLRTFPGARTKQVKEALYPLLTWSSKSKNLFQNFTKLLLFKKLCQRGSENLVRESWYPCVPEEEAHMIDIQDLFGPNLGTQKKPQLVIIEGAAGIGKSTLARLVKRAWKEGKLYRNDFHHVFFFSCRELAQYEQLSLAELIVQGQEVPTAPIRQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQTQPVHTLLGSLLGKSILPGASFLLTTRTTALQKFIPSLEQPCQVEVLGFTLFERKNYFYKYFG\n\n### Answer:", "output": "In strain: AKR/J. Little is known about the functional changes caused by this mutation in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00843", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRRLLSVAPVLLWLITPLAFAQLPGITSQPLPGGGQSWSLPVQTLVFITSLTFIPAILLMMTSFTRIIIVFGLLRNALGTPSAPPNQVLLGLALFLTFFIMSPVIDKIYVDAYQPFSEEKISMQEALEKGAQPLREFMLRQTREADLGLFARLANTGPLQGPEAVPMRILLPAYVTSELKTAFQIGFTIFIPFLIIDLVIASVLMALGMMMVPPATIALPFKLMLFVLVDGWQLLVGSLAQSFYS\n\n### Sequence 2:\nVFLLSLCPSAHADMPGLTSHILDDGSQTWSVPVQTLVFLTSLTFLPAFLLMMTSFTRIVIVFGLLRNALGTPYAPPNQILLGLALFLTFFIMSPTFEKIYKDAYVPFSQEKMNMEDAILKGSMPLKKFMLNQIRTPDLELFSKLAHISSYKNKNDIPMRILLPSFITSELKTAFQIGFTIFIPFLIIDLVVASVLMALGMMMVPPSTISLPFKLMLFVLVDGWQLLITSLAHSFNT\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00844", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSFVKDFKPQALGDTNLFKPIKIGNNELLHRAVIPPLTRMRALHPGNIPNRDWAVEYYTQRAQRPGTMIITEGAFISPQAGGYDNAPGVWSEEQMVEWTKIFNAIHEKKSFVWVNLWVLGWAAFPDNLARDGLRYDSASDNVFMDAEQEAKAKKANNPQHSLTKDEIKQYIKEYVQAAKNSIAAGADGVEIHSANGYLLNQFLDPHSNTRTDEYGGSIENRARFTLEVVDALVEAIGHEKVGLRLSPYGVFNSMSGGAETGIVAQYAYVAGELEKRAKAGKRLAFVHLVEPRVTNPFLTEGEGEYEGGSNDFVYSIWKGPVIRAGNFALHPEVVREEVKDKRTLIGYGRFFISNPDLVDRLEKGLPLNKYDRDTFYQMSAHGYIDYPTYEEALKLGWDKK\n\n### Answer:", "output": "<SEQ_2D>CBCSSCCCCCCTTSGGGSCEEETTEEESSCEEECCCCCCCCBTTTTBCCTTTHHHHHHHTTCSTTCEEEEEEEESSTTTCCCTTSCBSSSHHHHHHHHHHHHHHHHTTCEEEEEEECCGGGSCHHHHHHTTCCEEESCSSCCSCHHHHHHHHHTTCCEEECCHHHHHHHHHHHHHHHHHHHHTTCSEEEEECCTTSHHHHHHSTTTCCCCSTTSSSHHHHTHHHHHHHHHHHHHHCGGGEEEEECTTCCTTTCCGGGSTTHHHHHHHHHHHHHHHHHTTCCCSEEEEECTTCSCTTSCTTSSCCCSCCSTTHHHHCCSCEEEESSCTTCHHHHHHHTTSTTEEEECCHHHHHCTTHHHHHHHTCCCBCCCGGGSSCSSSTTTTCCCCHHHHHHTTTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00845", "category": "Mutation", "instruction": "Describe the effect of mutation Q6ZJK7-A298D on protein Tryptophan decarboxylase 1.", "input": "### Wild-type:\nMGSLDTNPTAFSAFPAGEGETFQPLNADDVRSYLHKAVDFISDYYKSVESMPVLPNVKPGYLQDELRASPPTYSAPFDVTMKELRSSVVPGMTHWASPNFFAFFPSTNSAAAIAGDLIASAMNTVGFTWQASPAATEMEVLALDWLAQMLNLPTSFMNRTGEGRGTGGGVILGTTSEAMLVTLVAARDAALRRSGSDGVAGLHRLAVYAADQTHSTFFKACRLAGFDPANIRSIPTGAETDYGLDPARLLEAMQADADAGLVPTYVCATVGTTSSNAVDPVGAVADVAARFAAWVHVAAA\n\n### Mutant:\nMGSLDTNPTAFSAFPAGEGETFQPLNADDVRSYLHKAVDFISDYYKSVESMPVLPNVKPGYLQDELRASPPTYSAPFDVTMKELRSSVVPGMTHWASPNFFAFFPSTNSAAAIAGDLIASAMNTVGFTWQASPAATEMEVLALDWLAQMLNLPTSFMNRTGEGRGTGGGVILGTTSEAMLVTLVAARDAALRRSGSDGVAGLHRLAVYAADQTHSTFFKACRLAGFDPANIRSIPTGAETDYGLDPARLLEAMQADADAGLVPTYVCATVGTTSSNAVDPVGAVADVAARFAAWVHVDAA\n\n### Answer:", "output": "Enables enzymatic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00846", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMALYEHVLLARQDISQQQVDALVEQFKGVLEANGGKFGKVENWGLRPLTYRIKKNRKAYYTLVNIDAPAAAVAEMERQMRINEDVLRFLTVRVEEHEEGQSAMLTRRDDRRERDGDDRPRRREGGFDRGDRGDRGPRRPRDTEAGEGA\n\n### Sequence 2:\nMALYEHVLLARQDISQQQVDALVEQFKGVLEANGGKFGKVENWGLRPLTYRIKKNRKAYYTLVNIDAPAAAVAEMERQMRINEDVLSFLTVRVEEHEEGQSAMLTRRDDRRERDGDDRPRRREGGFDRGDRGDRGPRRPRDNEAGEGA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00847", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRAACLFLLFMPGLLAQGQYDLDPLPPFPDHVQYNHYGDQIDNADYYDYQEVSPRTPEEQFQSQQQVQQEVIPAPTPEPAAAGDLETEPTEPGPLDCREEQYPCTRLYSIHKPCKQCLNEVCFYSLRRVYVVNKEICVRTVCAHEELLRADLCRDKFSKCGVMAVSGLCQSVAASCARSCGGC\n\n### Sequence 2:\nMRAACLFLLFMPGLLAQGQYDLDPLPPFPDHVQYNHYGDQIDNADYYDYQEVSPRTPEEQFQSQQQVQQEVIPAPTPEPAAAGDLETEPTEPGPLDCREEQYPCTRLYSIHKPCKQCLNEVCFYSLRRVYVVNKEICVRTVCAHEELLRADLCRDKFSKCGVMAVSGLCQSVAASCARSCGGC\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00848", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSYFA_BLOFL\n\n### Answer:", "output": "Belongs to the class-II aminoacyl-tRNA synthetase family. Phe-tRNA synthetase alpha subunit type 1 subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00849", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPIAQIHILEGRSDEQKETLIREVSEAISRSLDAPLTSVRVIITEYAKGHAGIGGELA\n\n### Answer:", "output": "<SEQ_2D>CEEEEEEESCCCHHHHHHHHHHHHHHHHHHTTCCGGGCEEEEEEECGGGEEESSSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00850", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nALB TTR APOB HNF4A\n</cell>", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as APOB, TTR, HNF4A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "skin", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["ALB", "TTR", "APOB", "HNF4A"]}}
{"id": "eval_00851", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "NKG7, NCAM1, FCGR3A, KLRD1", "output": "This profile is most consistent with natural killer cell, which is commonly found in prostate under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "natural killer cell", "tissue": "prostate", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["NKG7", "NCAM1", "FCGR3A", "KLRD1"]}}
{"id": "eval_00852", "category": "Mutation", "instruction": "Describe the effect of mutation P77202-C129S on protein Thiol:disulfide interchange protein DsbG.", "input": "### Wild-type:\nMLKKILLLALLPAIAFAEELPAPVKAIEKQGITIIKTFDAPGGMKGYLGKYQDMGVTIYLTPDGKHAISGYMYNEKGENLSNTLIEKEIYAPAGREMWQRMEQSHWLLDGKKDAPVIVYVFADPFCPYCKQFWQQARPWVDSGKVQLRTLLVGVIKPESPATAAAILASKDPAKTWQQYEASGGKLKLNVPANVSTEQMKVLSDNEKLMDDLGANVTPAIYYMSKENTLQQAVGLPDQKTLNIIMGNK\n\n### Mutant:\nMLKKILLLALLPAIAFAEELPAPVKAIEKQGITIIKTFDAPGGMKGYLGKYQDMGVTIYLTPDGKHAISGYMYNEKGENLSNTLIEKEIYAPAGREMWQRMEQSHWLLDGKKDAPVIVYVFADPFCPYSKQFWQQARPWVDSGKVQLRTLLVGVIKPESPATAAAILASKDPAKTWQQYEASGGKLKLNVPANVSTEQMKVLSDNEKLMDDLGANVTPAIYYMSKENTLQQAVGLPDQKTLNIIMGNK\n\n### Answer:", "output": "No loss of chaperone activity; when associated with S-126. This mutation may affect the protein's function as a molecular chaperone and its ability to prevent thermal aggregation of substrate proteins. This mutation affects the ability of DsbG to control the global sulfenic acid content of the periplasm and protect single cysteine residues from oxidation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00853", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nIQRTPKIQVYSRHPAENGKSNFLNCYVSGFHPSDIEVDLLKNGERIEKVEHSDLSFSKDWSFYLLYYTEFTPDEYACRVNHVTLSQPKIVKWDRDM\n\n### Answer:", "output": "<SEQ_3Di>DWWDKDKDWAFCDLDDAQDWTKTKMKIWTGDDDDKDKAKDDVHDGDDQKDKDDWDHDVVPIIMMMIIDIDHHHWIWMWIDDDPPPDIDIDTDDNVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00854", "category": "Mutation", "instruction": "Describe the effect of mutation Q9JIH7-S382D on protein Serine/threonine-protein kinase WNK1.", "input": "### Wild-type:\nMSDGTAEKQSGTPGFLSPPAPVPKNGSSSDSSVGEKLGAAVADSGIGRTEEYRRRRHTMDKDSRGAAATTTPTEHRFFRRSVICDSNATALELPGLPLSIPQPSVPAVVPQSAPPEPHREETLTATVASQVSQQPSAAASPGEQAVVGSATATVPSSTSKDRPVSQPSLVGSKEEPPPSRSGSGSGGASAKEPQEERNQQQDDIEELETKAVGMSNDGRFLKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQDRKLTKSERQRFKEEAEMLKGLQHPNIVRFYDSWESTVKGKKCIVLV\n\n### Mutant:\nMSDGTAEKQSGTPGFLSPPAPVPKNGSSSDSSVGEKLGAAVADSGIGRTEEYRRRRHTMDKDSRGAAATTTPTEHRFFRRSVICDSNATALELPGLPLSIPQPSVPAVVPQSAPPEPHREETLTATVASQVSQQPSAAASPGEQAVVGSATATVPSSTSKDRPVSQPSLVGSKEEPPPSRSGSGSGGASAKEPQEERNQQQDDIEELETKAVGMSNDGRFLKFDIEIGRGSFKTVYKGLDTETTVEVAWCELQDRKLTKSERQRFKEEAEMLKGLQHPNIVRFYDSWESTVKGKKCIVLV\n\n### Answer:", "output": "Loss of kinase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00855", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nOX26_RAT\n\n### Answer:", "output": "Stimulates feeding and grooming behavior, metabolic rate and locomotor activity and increases blood pressure.  May have orexigenic activity.  May promote aldosterone secretion by the adrenal gland.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00856", "category": "Mutation", "instruction": "Describe the effect of mutation P15391-Y421F on protein B-lymphocyte antigen CD19.", "input": "### Wild-type:\nMPPPRLLFFLLFLTPMEVRPEEPLVVKVEEGDNAVLQCLKGTSDGPTQQLTWSRESPLKPFLKLSLGLPGLGIHMRPLAIWLFIFNVSQQMGGFYLCQPGPPSEKAWQPGWTVNVEGSGELFRWNVSDLGGLGCGLKNRSSEGPSSPSGKLMSPKLYVWAKDRPEIWEGEPPCLPPRDSLNQSLSQDLTMAPGSTLWLSCGVPPDSVSRGPLSWTHVHPKGPKSLLSLELKDDRPARDMWVMETGLLLPRATAQDAGKYYCHRGNLTMSFHLEITARPVLWHWLLRTGGWKVSAVTLAYL\n\n### Mutant:\nMPPPRLLFFLLFLTPMEVRPEEPLVVKVEEGDNAVLQCLKGTSDGPTQQLTWSRESPLKPFLKLSLGLPGLGIHMRPLAIWLFIFNVSQQMGGFYLCQPGPPSEKAWQPGWTVNVEGSGELFRWNVSDLGGLGCGLKNRSSEGPSSPSGKLMSPKLYVWAKDRPEIWEGEPPCLPPRDSLNQSLSQDLTMAPGSTLWLSCGVPPDSVSRGPLSWTHVHPKGPKSLLSLELKDDRPARDMWVMETGLLLPRATAQDAGKYYCHRGNLTMSFHLEITARPVLWHWLLRTGGWKVSAVTLAYL\n\n### Answer:", "output": "No effect on the ability to complement impaired B cell development and functions; when associated with F-461.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00857", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nFCER1A:low HLA-DQA1:low CD1C:medium\n</cell>", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as FCER1A, CD1C, HLA-DQA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "prostate", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["FCER1A", "HLA-DQA1", "CD1C"]}}
{"id": "eval_00858", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMETIHFTKVHGSQNDFFLVDEEENHITEWSDEKRANFAIKLCDRKHSLGGADGILYVTKSSEVGPIGQMRVVNSDGSIASMCGNGLRTVARYLLEKHALTDAKVETMKAILDVKKATSLGFDIPTYQVEISPVKFAAETLPMHVGVEKLFNQVIPELDAELAFSAVSVPNPHLITFVDQAVLDSNKQEKLASYLNSENPYFPDGVNVSFVKRLSDDAIYVRTFERGVGFTNACGTAMSACSLIKKMLDNDILETPLNVYNDGGRVQVTAKKDAAGEISLQLIGNATFVSKGSVRYENDVVTELTNEATDEQGQYQALVKEVKEFLKTTE\n\n### Sequence 2:\nMAVKVAINGFGRIGRLAFRRIQEVEGLEVVAVNDLTDDDMLAHLLKYDTMQGRFTGEVEVVDGGFRVNGKEVKSFSEPDASKLPWKDLNIDVVLECTGFYTDKDKAQAHIEAGAKKVLISAPATGDLKTIVFNTNHQELDGSETVVSGASCTTNSLAPVAKVLNDDFGLVEGLMTTIHAYTGDQNTQDAPHRKGDKRRARAAAENIIPNSTGAAKAIGKVIPEIDGKLDGGAQRVPVATGSLTELTVVLEKQDVTVEQVNEAMKNASNESFGYTEDEIVSSDVVGMTYGSLFDATQTRVMSVGDRQLVKVAAWYDNEMSYTAQLVRTLAYLAELSK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00859", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDUS3_KLULA\n\n### Answer:", "output": "tRNA-dihydrouridine(47) synthase [NAD(P)(+)], mRNA-dihydrouridine synthase DUS3, tRNA-dihydrouridine synthase 3.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00860", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMPRVKRGVQARARHKKVLKQAKGYYGARSRVYRVAFQAVTKAGQYAYRDRRQKKRTFRALWIARINAAARQNGLSYSRLINGLKKASIEIDRKILADIAVFDKATFTVLVEKAKAAL\n\n### Sequence 2:\nMKISLSTSLFSIEQKVEYFNLNYQSLSDFSVVAKLNYTFTWYGNDFSIGFAPKKGEKLYFDLFFTFKASPNHPFAAENFKPDSEAIDFYVSFSWRLEGKDEVSKKLFELSVFGRARAFQIDDYKINLFSYLVYVIR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00861", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nFEFTLMVVGESGLGKSTLINSLFLVQVEQSKVLILLLTIVDTPGFGDAVDNSNCWQPVIDYIDSKFEDYLNAESRVNRRQMPDNRVQCCLYFIAPSGHGLKPLDIEFMKRLHEKVNIIPLIAKADTLTPEECQQFKKQIMKEIQEHKIKIYEFPETLVKKIKDRLPLAVVGNGEHCDFTILRNMLIRTHMQDLKDVTNNVHYENYRSRKLAA\n\n### Answer:", "output": "<SEQ_3Di>DEAEEEEDDAPPLCVVLLVCLLVCVAKDWDWDDVPIYIYIYNPDPHPPPPPPDDLVRVVCVQVVQLVVLVVLCPDPPNPDDDRPHHLEYEYEADLDDLAGPPVRLVNQQVCQLSYAYQYEYTDCVVDDPVRVVSHFVRYVCVDVVVPGDHDDDDDDVCVCVVVVDPAYDDSPCPVHVSVVVVVCPVPVNCVVSVVCNVPRRSVVVVVVVPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00862", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKFSFVYGLTGFLAATSSALPSEILSTGYERSALEKRGDGYLIMCKNCDPNTGSCDWKQNWNTCVGIGANVHWMVTGGSTDGKQGCATIWEGSGCVGRSTTMCCPANTCCNINTGFYIRSYRRVE\n\n### Sequence 2:\nMKFSFVYGLTGFLAATSSALPSEILSTGYERSALEKRGDGYLIMCKNCDPNTGSCDWKQNWNTCVGIGANVHWMVTGGSTDGKQGCATIWEGSGCVGRSTTMCCPANTCCNINTGFYIRSYRRVE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00863", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nATPG_STRA5\n\n### Answer:", "output": "ATP synthase gamma chain, ATP synthase F1 sector gamma subunit, F-ATPase gamma subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00864", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLTPKETCDLCQIALRTVFGHFGGNIPSRRKLVHQLKHECKRHFNYRRRCLLLMKVNSDLIFREMTDGSFKPMEVCLIMRECNPHDSPL\n\n### Answer:", "output": "<SEQ_3Di>DPLQVLLQVQLVLLVVLLVVCPLDDDDLVVSLVSQLVSLVVDVVRNVVSNVLCVVCSVVSVVCSNVPPDDSSVSCVVSVSHPPPYDGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00865", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: PARP1\nLigand SMILES: Cc1ccccc1S(=O)(=O)[O-]\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and poly(ADP-ribose) polymerase 1 (PARP1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "Cc1ccccc1S(=O)(=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 171.2, "LogP": 0.9, "TPSA": 57.2, "QED": 0.59, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "PARP1", "has_evidence": false}}
{"id": "eval_00866", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nMTNA_LEIIN\n\n### Answer:", "output": "Belongs to the eIF-2B alpha/beta/delta subunits family. MtnA subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00867", "category": "Mutation", "instruction": "Describe the effect of mutation P02545-R482L on protein Prelamin-A/C [Cleaved into: Lamin-A/C.", "input": "### Wild-type:\nMETPSQRRATRSGAQASSTPLSPTRITRLQEKEDLQELNDRLAVYIDRVRSLETENAGLRLRITESEEVVSREVSGIKAAYEAELGDARKTLDSVAKERARLQLELSKVREEFKELKARNTKKEGDLIAAQARLKDLEALLNSKEAALSTALSEKRTLEGELHDLRGQVAKLEAALGEAKKQLQDEMLRRVDAENRLQTMKEELDFQKNIYSEELRETKRRHETRLVEIDNGKQREFESRLADALQELRAQHEDQVEQYKKELEKTYSAKLDNARQSAERNSNLVGAAHEELQQSRIRID\n\n### Mutant:\nMETPSQRRATRSGAQASSTPLSPTRITRLQEKEDLQELNDRLAVYIDRVRSLETENAGLRLRITESEEVVSREVSGIKAAYEAELGDARKTLDSVAKERARLQLELSKVREEFKELKARNTKKEGDLIAAQARLKDLEALLNSKEAALSTALSEKRTLEGELHDLRGQVAKLEAALGEAKKQLQDEMLRRVDAENRLQTMKEELDFQKNIYSEELRETKRRHETRLVEIDNGKQREFESRLADALQELRAQHEDQVEQYKKELEKTYSAKLDNARQSAERNSNLVGAAHEELQQSRIRID\n\n### Answer:", "output": "In FPLD2; abnormal nuclear localization in a honeycomb expression pattern in about 10% of cultured skin fibroblasts from heterozygous patients; no effect on protein level. The missense mutation in the LMNA gene has been identified in individuals with partial lipodystrophy (PLD). This mutation affects the lamin A/C protein, which is a component of the nuclear envelope. The mutation leads to a regional loss of fat in the trunk and limbs during the peri-pubertal phase. Additionally, it may cause variable degrees of resistance to insulin action and a hyperlipidaemic state,", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00868", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQAYLQQSGAELVRPGASVKMSCKASGYTFTSYNMHWVKQTPRQGLEWIGAIYPGNGDTSYNQKFKGKATLTVDKSSSTAYMQLSSLTSEDSAVYFCARVVYYSNSYWYFDVWGTGTTVTVSAASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCD\n\n### Answer:", "output": "<SEQ_3Di>DWAFAWDAEDEDEAQAKDKTKTATDDDDLQVWKKWKWWADPPPGIGTAKIARPVPGDIGGDPVQPPFWDKDDDDPSRMIMIMGGNDDQVPQTWMKMFIWDCDPPHGIDGDHIYPTYGYGYHPDDFDAWDKDWFAWEVVQDDPQKGKIWIKTHFGDDDDKDKDKPNVPDDPQKDWDDWDQDPVRGTITMIMGMDGNVCLVPDWIWIWMDDVRVPDIDIDTRHHDYPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00869", "category": "Mutation", "instruction": "Describe the effect of mutation P09592-P52Q on protein Structural polyprotein.", "input": "### Wild-type:\nMFPFQPMYPMQPMPYRNPFAAPRRPWFPRTDPFLAMQVQELTRSMANLTFKPRRDAPPEGPSAKKPKKEASQKQKGGGQGKKKKNQGKKKAKTGPPNPKAQNGNKKKTNKKPGKRQRMVMKLESDKTFPIMLEGKINGYACVVGGKLFRPMHVEGKIDNDVLAALKTKKASKYDLEYADVPQNMRADTFKYTHEKPQGYYSWHHGAVQYENGRFTVPKGVGAKGDSGRPILDNQGRVVAIVLGGVNEGSRTALSVVMWNEKGVTVKYTPENCEQWSLVTTMCLLANVTFPCAQPPICYDR\n\n### Mutant:\nMFPFQPMYPMQPMPYRNPFAAPRRPWFPRTDPFLAMQVQELTRSMANLTFKQRRDAPPEGPSAKKPKKEASQKQKGGGQGKKKKNQGKKKAKTGPPNPKAQNGNKKKTNKKPGKRQRMVMKLESDKTFPIMLEGKINGYACVVGGKLFRPMHVEGKIDNDVLAALKTKKASKYDLEYADVPQNMRADTFKYTHEKPQGYYSWHHGAVQYENGRFTVPKGVGAKGDSGRPILDNQGRVVAIVLGGVNEGSRTALSVVMWNEKGVTVKYTPENCEQWSLVTTMCLLANVTFPCAQPPICYDR\n\n### Answer:", "output": "Enhancing cytopathic effects, capable of promoting host transcription. The results of the present study suggest that the N-terminal, approximately 35-amino-acid-long peptide of VEEV and EEEV capsid proteins plays the least critical role in the upregulation of cellular transcription and prevention of a cytopathic effect. The identified VEEV-specific peptide C(VEE)33-68 includes two domains with indistinct functions: the alpha-helix domain, helix I, which is minimally involved in hindering the balance between the presence of the protein in the cytoplasm and nucleus, and the downstream peptide, which might lack a functional nuclear localization signal(s). The integrity of both domains not only determines the intracellular distribution of the VEEV capsid but is also unessential for direct capsid protein functioning in the promotion of transcription. Our results suggest that the VEEV capsid protein dissociates from the nuclear pore complex, and this interaction inversely correlates with the protein's ability to cause transcriptional enhancement and, ultimately, cell survival.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00870", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CX3CR1, TMEM119, P2RY12", "output": "This cell is most consistent with microglia, a brain microglia cell. Key markers such as CX3CR1, P2RY12, TMEM119 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "ovary", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CX3CR1", "TMEM119", "P2RY12"]}}
{"id": "eval_00871", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nAIM3_YEAS2\n\n### Answer:", "output": "Belongs to the AIM3 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00872", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSPEE_ECOK1\n\n### Answer:", "output": "Catalyzes the irreversible transfer of a propylamine group from the amino donor S-adenosylmethioninamine (decarboxy-AdoMet) to putrescine (1,4-diaminobutane) to yield spermidine.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00873", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CCOCCOCC", "output": "The SMILES represents an unnamed molecule with MW=118.18, LogP=1.06, TPSA=18.46. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "CCOCCOCC", "compound_name": null, "pubchem_cid": null, "MW": 118.18, "LogP": 1.06, "TPSA": 18.46, "QED": 0.5, "functional_groups": ["ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_00874", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nTQQLTIEVLVTVDGVNFRTVVLNNKNTYRSQLGCVFFNGADISDTIPDEKQNGHSLYLADNLTADETKALKELYGPVDPTFLHRFYSLKAAVHGWKMVVCDKVRSLKLSDNNCYLNAVIMTLDLLKDIKFVIPALQHAFMKHKGGDSTDFIALIMAYGNCTFGAPDDASRLLHTVLAKAELCCSARMVWREWCNVCGIKDVVLQGLKACCYVGVQTVEDLRARMTYVCQCGGERHRQLVEHTTPWLLLSGTPNEKLVTTSTAPDFVAFNVFQGIETAVGHYVHARLKGGLILKFDSGTVSKTSDWKCKVTDVLFPGQKYSSDCN\n\n### Answer:", "output": "<SEQ_2D>CCCCEEEEEEESSSSCCEEEEEESSSCHHHHTCSEEETTEECTTSCCCGGGTTCEEEECSSCCHHHHHHHHHHHCSCCTTHHHHHHHHHHHGGGCCEEEETTEEEECCCTTCHHHHHHHHHHTTCTTEEESSHHHHHHHHHHHHTCCHHHHHHHHHHTTCCTTCCCCHHHHHHHHHTTEEESSCCEEEEEEEETTTEEEEEEEEGGGGGEEESCCSHHHHSSCEEEECTTSSEEEEEEEEEEESCEEEEEEEEEEEEESSSCCSCSEEEEEESCTTSBCEEEEEEECCCCEEEEETTEEEEESEEEEEEEEEEECCEEEECCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00875", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPIMT_XANC5\n\n### Answer:", "output": "Protein-L-isoaspartate O-methyltransferase, L-isoaspartyl protein carboxyl methyltransferase, Protein L-isoaspartyl methyltransferase, Protein-beta-aspartate methyltransferase, PIMT.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00876", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMFLLHEYDIFWAFLIISSVIPILAFLISGVLAPISEGPEKLSSYESGIEPMGDAWLQFRIRYYMFALVFVVFDVETVFLYPWAMSFDVLGVPVFIEALIFVLILIVGSVYAWRKGALEWS\n\n### Sequence 2:\nMFLLHEYDIFWAFLIISSVIPILAFLISGVLAPINEGPEKLSSYESGIEPMGDAWLQFRIRYYMFALVFVVFDVETVFLYPWAMSFDVLGVSVFVEALIFVLILIVGLVYAWRKGALEWS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00877", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nIKZF2 CD4 CD3D FOXP3\n</cell>", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as FOXP3, CD3D, CD4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "blood", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IKZF2", "CD4", "CD3D", "FOXP3"]}}
{"id": "eval_00878", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "MS4A1, CD79A, TCL1A", "output": "The expression pattern is consistent with naive B cell in a context associated with rheumatoid arthritis, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "colon", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["MS4A1", "CD79A", "TCL1A"]}}
{"id": "eval_00879", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPDXB_PSEPG\n\n### Answer:", "output": "Belongs to the D-isomer specific 2-hydroxyacid dehydrogenase family. PdxB subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00880", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCH101_VIBVU\n\n### Answer:", "output": "Belongs to the GroES chaperonin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00881", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nAKKWVTVYYGVPVWKEATTTLFCASDAKAYDTEVHNVWATHACVPTDPNPQEIVLGNVTENFNMWKNNMVEQMHEDIISLWDQSLKPCVKLTPLCVTLNCNNVNTNNGEMKNCSFNVTTSIRDKIKKEYALFYKLDVVPLENITNYRLINCNTSVITQACPKVSFEPIPIHYCAPAGFAILKCNSKTFNGSGPCTNVSTVQCTHGIRPVVSTQLLLNGSLAEEEIVIRSENITDNAKTIIVQLNEAVEINCTRPNNNTRKSIHIGPGRAFYATGDIIGNIRQAHCNISKARWNETLGQIVAKLEEQFPNKTIIFNHSSGGDPEIVTHSFNCGGEFFYCNTTPLFNSTWNNTRTDDYPTGGEQNITLQCRIKQIINMWQGVGKAMYAPPIRGQIRCSSNITGLLLTRDGGRDQNGTETFRPGGGNMRDNWRSELYKYKVVKIEPLGIAPTACKRRV\n\n### Answer:", "output": "<SEQ_3Di>DDKDKDKFFFDFDWFFDDDDWAWADAVVPPPDPAQAQLGVVPHHHDDPDFDKAWDDQDKDWFALVDDVVLVVVVVVVVVSVVVALPPFDWCLQQFDKWFWDPPPDVVDFKTWIKDKGQAQDNPDIDIDIFIDGPVQWDFDDVGTMTGGDCVVFATETAGDPSHHDFFFFIKIFDDPQKWKKFADDQADEGGGMGRTIMIGGIFDRHTPWFKAFFFWADDAGPPAWHKYWDDVVDQVIKIKIFHHAWWKKKKKDDWQWDWDWGDHHDPHIDIGIDDTDDDDDFIKIWTWDVVVLVVVLRVVVVVCVVQVPFKEAEAADDDDDLRFRFIWGDFGQWIKRWHCRVPRGFIDDPSDGPPDPPDDDTTDMITIGTDAWDDPVDDRHYIYGGHDDGDMTMHTIITTIGMKGWDDDDCDDRYTYIYRGPPDRSVNVSRRCRTMGMIGDDRMDMDMDPDDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00882", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMMISILLVFAVSTAYCRDYGREMRATRMSVVDFWKNETKGDGLNVRVAHGVNSWPDLVDQLHEPFLNKSMMIEGDVFMQAHRRPRHRAVPVMKADTKLADRITFKEWLREVATMNKAIKINFRSNEVVRPVLQDLYASQADPTSPVLQYPVILHANVFRSPRSVETEVDPSTFVEKAKDLFPDATLSLGWTKQSNFSHLHPKFKKLSWRQLFHILEYISRLDQPVMLSVRLSVAAHSKEQLLWLLGMDQSISLLLWSDAEDHVTNWTPIVELRRSTTKNRILYDLDPKHRKFLQTDSNEPIATPPTFSLQDWQAVEFSSSGSQLSTVVRSEKGPAFLGEPTALLLSKIPPPNFPSEQKISGKVHFLPKKGMENVDLDENSGVSIYLMDKVQDIDSPKIKNSLEVFIGFDGKIKITNGQLKNLPYYETKSVGQLPDSDCYGFEVTDMGWRVKLDAWTAECGNQKLRKRRTVSIELDTPFQKNRHLRNVVVSKKGDSAVDFLLEELQHKNAYNYRNPWITVVSTILLMIYLL\n\n### Sequence 2:\nMSTTTATPAGADDTVAAARTYQVRTFGCQMNVHDSERLSGLLEDAGYVRFDEAGRPEGSDAPVEPDVVVFNTCAVRENADNKLYGNLGHLAPVKERRPGMQIAVGGCLAQKDRGEIVRKAPWVDVVFGTHNVGSLPVLLERARHNAEAQVEILESLETFPSTLPTRRESPYAAWVSISVGCNNTCTFCIVPALRGKEKDRRPGDVLAEIEALVGEGVLEVTLLGQNVNTYGVEFGDKLAFGKLLRATGGIEGLERVRFTSPHPSSFTDDVVDAMAETPNVMPSLHMPLQSGSDRVLKAMRRSYRQSRFLGIIDRVRSSIPDAAITTDIIVGFPGETDEDFEQTLHVVEQARFSSAFTFQYSPRPGTPAATMGDQIPKRVVQERYERLTALQDRITYEDNQAQTGRTLEVLVAEGEGRKDAATRRLSGRAPDNRLVHFALPEGAQAPRPGDVATVTVTRGAPHYLEADDVSGFAVRRTRAGDAWEARQARPEPESTGPRPVGLGLPTLRRA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00883", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nASNA_DROGR\n\n### Answer:", "output": "Belongs to the arsA ATPase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00884", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nARGB_CUPPJ\n\n### Answer:", "output": "Acetylglutamate kinase, N-acetyl-L-glutamate 5-phosphotransferase, NAG kinase, NAGK.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00885", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "GATA1, HBB, HBA1", "output": "The expression pattern is consistent with erythroid progenitor in a context associated with Alzheimer disease, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "breast", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["GATA1", "HBB", "HBA1"]}}
{"id": "eval_00886", "category": "Mutation", "instruction": "Describe the effect of mutation F2XG53-K316A on protein CRISPR-associated nuclease/helicase Cas3.", "input": "### Wild-type:\nMKHINDYFWAKKTEENSRLLWLPLTQHLEDTKNIAGLLWEHWLSEGQKVLIENSINVKSNIENQGKRLAQFLGAVHDIGKATPAFQTQKGYANSVDLDIQLLEKLERAGFSGISSLQLASPKKSHHSIAGQYLLSHYGVDEDIATIIGGHHGRPVDDLDGLNSQKSYPSNYYQDEKKDSLVYQKWKSNQEAFLNWALTETGFNSVSQLPKIKQPAQVILSGLLIMSDWIASNEHFFPLLSLDETDVKNKSQRIETGFKKWKKSNLWQPETFVDLVTLYQERFGFSPRNFQLILSQTIEKT\n\n### Mutant:\nMKHINDYFWAKKTEENSRLLWLPLTQHLEDTKNIAGLLWEHWLSEGQKVLIENSINVKSNIENQGKRLAQFLGAVHDIGKATPAFQTQKGYANSVDLDIQLLEKLERAGFSGISSLQLASPKKSHHSIAGQYLLSHYGVDEDIATIIGGHHGRPVDDLDGLNSQKSYPSNYYQDEKKDSLVYQKWKSNQEAFLNWALTETGFNSVSQLPKIKQPAQVILSGLLIMSDWIASNEHFFPLLSLDETDVKNKSQRIETGFKKWKKSNLWQPETFVDLVTLYQERFGFSPRNFQLILSQTIEKT\n\n### Answer:", "output": "Loss of ssDNA-dependent ATPase.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00887", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "ALB, TTR, HNF4A, APOB", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as HNF4A, ALB, APOB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "tumor", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["ALB", "TTR", "HNF4A", "APOB"]}}
{"id": "eval_00888", "category": "Mutation", "instruction": "Describe the effect of mutation Q8TJK1-S150C on protein Arsenite methyltransferase.", "input": "### Wild-type:\nMDAAEKKEVIKKKYQEIATLGGSCCSGGGCCGDLSAADLSRSLGYSEADVQAVPDANLGLGCGNPTAFAELKPGDIVLDLGSGAGFDSFLAAQRVGSLGKVIGVDMTQEMVKKAQDNARKYGYSNVEFRQGDIEALPLDDRSVDVIISNSVINLAPDKEKVFREAFRVLKPGGRMYVSDMVLLEDLPEDLKNDCDLLAGCVAGALLKEEYLGLLKKAGFSFKILAEDSDVSKRQYEGLPVESLKLKAWV\n\n### Mutant:\nMDAAEKKEVIKKKYQEIATLGGSCCSGGGCCGDLSAADLSRSLGYSEADVQAVPDANLGLGCGNPTAFAELKPGDIVLDLGSGAGFDSFLAAQRVGSLGKVIGVDMTQEMVKKAQDNARKYGYSNVEFRQGDIEALPLDDRSVDVIISNCVINLAPDKEKVFREAFRVLKPGGRMYVSDMVLLEDLPEDLKNDCDLLAGCVAGALLKEEYLGLLKKAGFSFKILAEDSDVSKRQYEGLPVESLKLKAWV\n\n### Answer:", "output": "Increases activity. Mutation at site in MaArsM enables the methylation of monomethylarsenic.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00889", "category": "Mutation", "instruction": "Describe the effect of mutation Q980R9-Y125A on protein N-alpha-acetyltransferase.", "input": "### Wild-type:\nMELAEKDKGRDFTLRNARMDDIDQIIKINRLTLPENYPYYFFVEHLKEYGLAFFVAIVDNSVVGYIMPRIEWGFSNIKQLPSLVRKGHVVSIAVLEEYRRKGIATTLLEASMKSMKNDYNAEEIYLEVRVSNYPAIALYEKLNFKKVKVLKGYYADGEDAYLMARPL\n\n### Mutant:\nMELAEKDKGRDFTLRNARMDDIDQIIKINRLTLPENYPYYFFVEHLKEYGLAFFVAIVDNSVVGYIMPRIEWGFSNIKQLPSLVRKGHVVSIAVLEEYRRKGIATTLLEASMKSMKNDYNAEEIALEVRVSNYPAIALYEKLNFKKVKVLKGYYADGEDAYLMARPL\n\n### Answer:", "output": "Same catalytic efficiency and 1.7-fold decrease of the affinity for Ser-N-terminal peptide compared to the wild-type. 1.5- and 2.5-fold decrease of the catalytic efficiency and affinity for Met-N-terminal peptide compared to the wild-type, respectively. The mutation in the ssNAT protein has been shown to alter the substrate specificity of the enzyme.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00890", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "Cn1cccc1", "output": "No standard functional groups from the preset library were detected in this molecule.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "FreeSolv", "canonical_smiles": "Cn1cccc1", "compound_name": null, "pubchem_cid": null, "MW": 81.12, "LogP": 1.03, "TPSA": 4.93, "QED": 0.44, "functional_groups": [], "moleculenet_labels": {"label": -2.89}, "split": "train", "generation_method": "template_from_smarts", "fgs": []}}
{"id": "eval_00891", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "GAD1, SLC32A1, GAD2", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as GAD2, GAD1, SLC32A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["GAD1", "SLC32A1", "GAD2"]}}
{"id": "eval_00892", "category": "Mutation", "instruction": "Describe the effect of mutation A0A411MR89-E215A on protein Nitrosourea synthase.", "input": "### Wild-type:\nMSHVPPHVPFELSGAELRDAIVQYATNPIYHDNLDWLNHDNPYRRQLRPQVLPHLDYDKVPGRENILNYASLAVQRLLTSVYEADLVFFPKSGLKGKEEDFRAFYSPANRALGERIRPALERYAFGFLDDEVETSGTWTAQSLDAYLDSLDTAGGAEQSPVEKAILGSADRERAARMWLVQFAPDFLSEASPMMRNVLGYYGPAQSEWFKVVIDEYGYGVHDTKHSTLFERTLESVGLESDLHRYWQYYLNSSLLLNNYFHYLGKNHELFFRYVGALYYTESSLVDFCRRADHLLREVFG\n\n### Mutant:\nMSHVPPHVPFELSGAELRDAIVQYATNPIYHDNLDWLNHDNPYRRQLRPQVLPHLDYDKVPGRENILNYASLAVQRLLTSVYEADLVFFPKSGLKGKEEDFRAFYSPANRALGERIRPALERYAFGFLDDEVETSGTWTAQSLDAYLDSLDTAGGAEQSPVEKAILGSADRERAARMWLVQFAPDFLSEASPMMRNVLGYYGPAQSEWFKVVIDAYGYGVHDTKHSTLFERTLESVGLESDLHRYWQYYLNSSLLLNNYFHYLGKNHELFFRYVGALYYTESSLVDFCRRADHLLREVFG\n\n### Answer:", "output": "Loss of activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00893", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPDUK_CITFR\n\n### Answer:", "output": "The 1,2-PD-specific bacterial microcompartment (BMC) concentrates low levels of 1,2-PD catabolic enzymes, concentrates volatile reaction intermediates thus enhancing pathway flux and keeps the level of toxic, mutagenic propionaldehyde low.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00894", "category": "Mutation", "instruction": "Describe the effect of mutation P78330-T182S on protein Phosphoserine phosphatase.", "input": "### Wild-type:\nMVSHSELRKLFYSADAVCFDVDSTVIREEGIDELAKICGVEDAVSEMTRRAMGGAVPFKAALTERLALIQPSREQVQRLIAEQPPHLTPGIRELVSRLQERNVQVFLISGGFRSIVEHVASKLNIPATNVFANRLKFYFNGEYAGFDETQPTAESGGKGKVIKLLKEKFHFKKIIMIGDGATDMEACPPADAFIGFGGNVIRQQVKDNAKWYITDFVELLGELEE\n\n### Mutant:\nMVSHSELRKLFYSADAVCFDVDSTVIREEGIDELAKICGVEDAVSEMTRRAMGGAVPFKAALTERLALIQPSREQVQRLIAEQPPHLTPGIRELVSRLQERNVQVFLISGGFRSIVEHVASKLNIPATNVFANRLKFYFNGEYAGFDETQPTAESGGKGKVIKLLKEKFHFKKIIMIGDGASDMEACPPADAFIGFGGNVIRQQVKDNAKWYITDFVELLGELEE\n\n### Answer:", "output": "Reduces L-phosphoserine phosphatase activity by about 99%.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00895", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMQSSPLLEHLIENLRCLPGVGPKSAQRMAYHLLQRNRSGGMNLARALTEAMSKIGHCSQCRDFTEEDTCNICNNPRRQNSGLLCVVEMPADIQAIEQTGQFSGRYFVLMGHLSPLDGIGPREIGLDLLQKRLVEESFHEVILATNPTVEGDATANYIAEMCRQHNIKVSRIAHGIPVGGELETVDGTTLTHSFLGRRQID\n\n### Sequence 2:\nMQTSPLLEHLIESLRCLPGVGPKSAQRMAYHLLQRNRNGGMELARALTEAMSKIGHCSHCRTFTEEEVCTICNNPRRQHSGLLCVVEQPSDIQAIEQTGQFSGRYFVLMGHLSPLDGIGPKEIGLHLLQKRLQEESFHEVILATNPTVEGDATANYIAELCMQQNIKVTRIAHGIPVGGELEMVDGTTLTHSFLGRRQL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00896", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMQNQRIRIRLKAFDHRLIDQSTAEIVETAKRTGAQVRGPIPLPTRKERFTVLISPHVNKDARDQYEIRTHKRLVDIVEPTEKTVDALMRLDLAAGVDVQISLG\n\n### Sequence 2:\nMMKKLLIVSVKAYQKYISPLSPPSCRYKPTCSAYMLTAIEKHGTKGILMGIARILRCHPFVAGGVDPVPEDFSLMRNKNTSKNAEKA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00897", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMRVVIQRVKGAILSVRKENIGENEKELEIISEIKNGLICFLGIHKNDTWEDALYIIRKCLNLRLWNNDNKTWDKNVKDLNYELLIVSQFTLFGNTKKGNKPDFHLAKEPNEALIFYNKIIDEFKKQYNDDKIKIGKFGNYMNIDVTNDGPVTIYIDTHDINLNK\n\n### Answer:", "output": "<SEQ_2D>CEEEEEEEEEEEEEEECCCCCCCCCCEEEEEEESSEEEEEEECBTTCCHHHHHHHHHHHHHCCCEEETTEEEEECTTTTTCEEEEEECGGGGCBCSSSSSCBCTTBCCTTHHHHHHHHHHHHHHHHSCTTSEEECCTTSCEEEEEEEEEEEEEEEEGGGCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00898", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMGLLDAAVSRQNVLVTSVDSVLNWARLSSLWPMGFGLACCAIEMMATNASNYDLERFGIFPRSSPRQSDLMIVAGTVTMKMAERVVRLYEQMPEPRYVLSMGSCSNCGGPYWEHGYHVLKGVDRVIPVDVYVPGCPPRPEALIGGLMKIQELIRMEGLGVSRADALKKLAEAESDPQPLIEEARKQKTA\n\n### Sequence 2:\nMVSTMLSGLVLWLTFGWTPALAYSPRTPDRVSETDIQRLLHGVMEQLGIARPRVEYPAHQAMNLVGPQSIEGGAHEGLQHLGPFGNIPNIVAELTGDNTPKDFSEDQGYPDPPNPCPIGKTDDGCLENTPDTAEFSREFQLHQHLFDPEHDYPGLGKWNKKLLYEKMKGGQRRKRRSVNPYLQGQRLDNVVAKKSVPHFSDEDKDPE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00899", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNFQSVISLLHHFWGTRGCLIAQPYDIEKGAGTKNPQTFLRALGPEPWAVAYVEPCRRPTDGRYGENPNRFQHYYQYQVLIKPSPDNIQEIYLDSLRALGIRPEDHDIRFVEDNWEDATVGAWGTGWEVWLDGMEITQFTYFQQCGGIDCRPVSIEITYGLERLTMYLQQVEAITKIHWTDNITYGDVFLQNEIEQSTYNFEASNPELLLTLFSLYEQEATQLTKKGLVLPSLDYVMKCSHTFNLLDARGVISVTERTRYIARIRHLARKVAHLYVEQREKLGFPLLKDLKPAIPVGQVEVAATQTKSASAG\n\n### Sequence 2:\nITFQSIIAKLHEFWSARGCLIAQPYDTEKGAGTMSPHTFLRAIGPEPWSVAYVEPCRRPTDGRYGENPNRFQHYYQYQVLIKPSPDNIQEIYLDSLRSLGIHPEDHDIRFVEDNWESPTLGAWGVGWEVWLDGMEITQFTYFQQCGGIDCRPVSIEITYGLERLAMYLQNVDAITKIQWNDQVNYGEIFLQNEIEQSTYNFEASTPDLLFNLFSLYEQEAKQLIERGLVIPSLDYVLKCSHSFNLLDARGVIAVAERTRYIGRIRNLAREVAQLYLQQRENLGFPL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00900", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nABCA3 SFTPC SFTPB SFTPA1\n</cell>", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as ABCA3, SFTPA1, SFTPC support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "ovary", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["ABCA3", "SFTPC", "SFTPB", "SFTPA1"]}}
{"id": "eval_00901", "category": "Mutation", "instruction": "Describe the effect of mutation Q99220-N380A on protein Protein OS-9 homolog.", "input": "### Wild-type:\nMQAKIIYALSAISALIPLGSSLLAPIEDPIVSNKYLISYIDEDDWSDRILQNQSVMNSGYIVNMGDDLECFIQNASTQLNDVLEDSNEHSNSEKTALLTKTLNQGVKTIFDKLNERCIFYQAGFWIYEYCPGIEFVQFHGRVNTKTGEIVNRDESLVYRLGKPKANVEEREFELLYDDVGYYISEIIGSGDICDVTGAERMVEIQYVCGGSNSGPSTIQWVRETKICVYEAQVTIPELCNLELLAKNEDQKNASPILCRMPAKSKIGSNSIDLITKYEPIFLGSGIYFLRPFNTDERDKL\n\n### Mutant:\nMQAKIIYALSAISALIPLGSSLLAPIEDPIVSNKYLISYIDEDDWSDRILQNQSVMNSGYIVNMGDDLECFIQNASTQLNDVLEDSNEHSNSEKTALLTKTLNQGVKTIFDKLNERCIFYQAGFWIYEYCPGIEFVQFHGRVNTKTGEIVNRDESLVYRLGKPKANVEEREFELLYDDVGYYISEIIGSGDICDVTGAERMVEIQYVCGGSNSGPSTIQWVRETKICVYEAQVTIPELCNLELLAKNEDQKNASPILCRMPAKSKIGSNSIDLITKYEPIFLGSGIYFLRPFNTDERDKL\n\n### Answer:", "output": "Abolishes dimerization. This mutation may affect the association of Yos9 with Hrd3 in vitro.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00902", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLFSSLLCVFVAFSYSGSSVAQKVTQAQSSVSMPVRKAVTLNCLYETSWWSYYIFWYKQLPSKEMIFLIRQGSDEQNAKSGRYSVNFKKAAKSVALTISALQLEDSAKYFCALGE\n\n### Sequence 2:\nMVRVARSVNARKKRRSILAAAKGYRGQRSRLYRKAKEQVLRSLVYAYRDRRKRKSNFRKLWIIRINAAARMEGVTYNRFLQGLKLANIELDRRSLAHLAVHNPDTFSALVGVAKDSLKNAPDPVAAARLRGTSSRT\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00903", "category": "Mutation", "instruction": "Describe the effect of mutation O60869-S87D on protein Endothelial differentiation-related factor 1.", "input": "### Wild-type:\nMAESDWDTVTVLRKKGPTAAQAKSKQAILAAQRRGEDVETSKKWAAGQNKQHSITKNTAKLDRETEELHHDRVTLEVGKVIQQGRQSKGLTQKDLATKINEKPQVIADYESGRAIPNNQVLGKIERAIGLKLRGKDIGKPIEKGPRAK\n\n### Mutant:\nMAESDWDTVTVLRKKGPTAAQAKSKQAILAAQRRGEDVETSKKWAAGQNKQHSITKNTAKLDRETEELHHDRVTLEVGKVIQQGRQDKGLTQKDLATKINEKPQVIADYESGRAIPNNQVLGKIERAIGLKLRGKDIGKPIEKGPRAK\n\n### Answer:", "output": "Loss of interaction with CALM and higher affinity for TBP. Same effect; when associated with D-65 and D-74. The phosphorylation of EDF-1 at site S87 by PKA modulates its interaction with CaM.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00904", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nOSME_ECO57\n\n### Answer:", "output": "Osmotically-inducible putative lipoprotein OsmE, Activator of ntr-like gene protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00905", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nGALC_XENTR\n\n### Answer:", "output": "Belongs to the glycosyl hydrolase 59 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00906", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "GP1BA, PF4, PPBP", "output": "The expression pattern is consistent with megakaryocyte in a context associated with type 2 diabetes, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "ovary", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["GP1BA", "PF4", "PPBP"]}}
{"id": "eval_00907", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMASTRCKLARYLEDLEDVDLKKFKMHLEDYPPQKGCIPLPRGQTEKADHVDLATLMIDFNGEEKAWAMAVWIFAAINRRDLYEKAKRDEPKWGSDNARVSNPTVICQEDS\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHHTCCHHHHHHHHHHHHTCCSCTTCCCCCHHHHHHCCHHHHHHHHHHHHCHHHHHHHHHHHHHHTTCHHHHHHHHHTCCCTTTSSCCTTSCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00908", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMARVTVEDAVKQIGNRFDMILVAARRARQIAVQGKDPMVEEQNDKPTVIALREIEKGLVNAGTLDADERQSVREREAAEIAAVAAIAEGRSL\n\n### Sequence 2:\nMARVTVEDAVDQIGNRFDMILVAARRARQIAVQGKDPMVDEMNDKPTVIALREIELGLVNATTLDADERQTVRDREAAEVAAVSAIAEGRTL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00909", "category": "Mutation", "instruction": "Describe the effect of mutation Q10666-A127S on protein Protein pop-1.", "input": "### Wild-type:\nMMADEELGDEVKVFRRDEDADDDPMISGETSEQQLADDKKEAVMEAELDGAGRNPSIDVLKSAFPKVEPMSPSFPGLMSHFSPGYSAAALPMFMPLFMNPYAAALRSPSLMFPMGAMSPTFPMFPPAPVYGAAIAAAAAKQHFENMAPLNMRAGHPMNQMGMPPYMHPSSMAPQNVDRRAQGGGKAKKDDHVKKPLNAFMWFMKENRKALLEEIGNNEKQSAELNKELGKRWHDLSKEEQAKYFEMAKKDKETHKERYPEWSARENYAVNKKKTKKRRDKSIPSENNDQKKCRARFGVNN\n\n### Mutant:\nMMADEELGDEVKVFRRDEDADDDPMISGETSEQQLADDKKEAVMEAELDGAGRNPSIDVLKSAFPKVEPMSPSFPGLMSHFSPGYSAAALPMFMPLFMNPYAAALRSPSLMFPMGAMSPTFPMFPPSPVYGAAIAAAAAKQHFENMAPLNMRAGHPMNQMGMPPYMHPSSMAPQNVDRRAQGGGKAKKDDHVKKPLNAFMWFMKENRKALLEEIGNNEKQSAELNKELGKRWHDLSKEEQAKYFEMAKKDKETHKERYPEWSARENYAVNKKKTKKRRDKSIPSENNDQKKCRARFGVNN\n\n### Answer:", "output": "Enhances nuclear asymmetry, proficient in nuclear export and gain of par-5 interaction; when associated with A-107; A-109; A-118 and A-120.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00910", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nHBA_TUPGL\n\n### Answer:", "output": "Belongs to the globin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00911", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMPNQNLIALGFDFGMKRIGVAVGQTVTHSANAIAILKAQDGVPDWEKIKMLIETWHANVLVVGIPYNMDGSEQTLTFAARKFARKLQTRFGLPVSMVDERLTTIEAKRQWYEQGLTKRPQHLDNYAAKLILEQWLQEQKNE\n\n### Sequence 2:\nPSKPRIVLGFDFGLRYIGVAVGQEVTHSANPLTTLKAHEGNPDWNQITQLIRQWNPDLLIVGLPLNMDQSEQFLTKAARRFGHRLHGRYGLAVEWVDERLSTVEARERLNIKSSASGRRQGIDQMAAQCILQTWLTEQQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00912", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMLLVIGLGNPGKEYQYTRHNVGFIAIEKIANQYNSSFSTKKKFNCEIAETISYGQKIIFIKPTTYMNLSGKSVISVKTYYNIYPAKSFVIHDDIDLETGRVKFKTGGGNGGHNGLKSIDGVIGNNYNRIRIGVGRPQNNQDLADYVLNHFSKPEYKTVMQAIDRITSNFGLILENKLEEFKNKMA\n\n### Sequence 2:\nMLLVIGLGNPGKEYQYTRHNVGFIAIEKIVNQYNSSFSTKKKFNCEIAETISDGQKIIFIKPTTYMNLSGKSVISVKTYYNIYPAKIFVIHDDIDLETGRIKFKTGGGNGGHNGLKSIDGVIGNNYNRIRIGVGRPQNNQGVADYVLNHFSKPEYETVMQAIDRITSNFGLILENKLEEFKNKIA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00913", "category": "Mutation", "instruction": "Describe the effect of mutation O25338-R338G on protein Endonuclease MutS2.", "input": "### Wild-type:\nMSDAPKRSLNPTLMMNNNNTPPKPLEESLDLKEFIALFKTFFAKERDTIALENDLKQTFTYLNEVDAIGLPTPKSVKESDLIIIKLTKLGTLHLDEIFEIVKRLHYIVVLQNAFKTFTHLKFHERLNAIVLPPFFNDLIALFDDEGKIKQGANATLDALNESLNRLKKESVKIIHHYARSKELAPYLVDTQSHLKHGYECLLLKSGFSGAIKGVVLERSANGYFYLLPESAQKIAQKIAQIGNEIDCCIVEMCQTLSHSLQKHLLFLKFLFKEFDFLDSLQARLNFAKAYNLEFVMPSFT\n\n### Mutant:\nMSDAPKRSLNPTLMMNNNNTPPKPLEESLDLKEFIALFKTFFAKERDTIALENDLKQTFTYLNEVDAIGLPTPKSVKESDLIIIKLTKLGTLHLDEIFEIVKRLHYIVVLQNAFKTFTHLKFHERLNAIVLPPFFNDLIALFDDEGKIKQGANATLDALNESLNRLKKESVKIIHHYARSKELAPYLVDTQSHLKHGYECLLLKSGFSGAIKGVVLERSANGYFYLLPESAQKIAQKIAQIGNEIDCCIVEMCQTLSHSLQKHLLFLKFLFKEFDFLDSLQARLNFAKAYNLEFVMPSFT\n\n### Answer:", "output": "Lack of ATPase activity. Establishment of HpmutS2 enhances the efficiency of exogenous DNA incorporation. HpMutS2 promotes DNA strand exchange reactions in vitro.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00914", "category": "Mutation", "instruction": "Describe the effect of mutation Q86UE8-H518R on protein Serine/threonine-protein kinase tousled-like 2.", "input": "### Wild-type:\nMMEELHSLDPRRQELLEARFTGVGVSKGPLNSESSNQSLCSVGSLSDKEVETPEKKQNDQRNRKRKAEPYETSQGKGTPRGHKISDYFEFAGGSAPGTSPGRSVPPVARSSPQHSLSNPLPRRVEQPLYGLDGSAAKEATEEQSALPTLMSVMLAKPRLDTEQLAQRGAGLCFTFVSAQQNSPSSTGSGNTEHSCSSQKQISIQHRQTQSDLTIEKISALENSKNSDLEKKEGRIDDLLRANCDLRRQIDEQQKMLEKYKERLNRCVTMSKKLLIEKSKQEKMACRDKSMQDRLRLGHFT\n\n### Mutant:\nMMEELHSLDPRRQELLEARFTGVGVSKGPLNSESSNQSLCSVGSLSDKEVETPEKKQNDQRNRKRKAEPYETSQGKGTPRGHKISDYFEFAGGSAPGTSPGRSVPPVARSSPQHSLSNPLPRRVEQPLYGLDGSAAKEATEEQSALPTLMSVMLAKPRLDTEQLAQRGAGLCFTFVSAQQNSPSSTGSGNTEHSCSSQKQISIQHRQTQSDLTIEKISALENSKNSDLEKKEGRIDDLLRANCDLRRQIDEQQKMLEKYKERLNRCVTMSKKLLIEKSKQEKMACRDKSMQDRLRLGHFT\n\n### Answer:", "output": "In MRD57; severely reduced kinase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00915", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nARHG1_RAT\n\n### Answer:", "output": "Cytoplasm. Membrane. Note=Translocated to the membrane by activated GNA13 or LPA stimulation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00916", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMNMA_SHIB3\n\n### Answer:", "output": "tRNA-specific 2-thiouridylase MnmA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00917", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nITGA2B:high PPBP:high GP1BA:high\n</cell>", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as GP1BA, PPBP, ITGA2B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "lymph node", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["ITGA2B", "PPBP", "GP1BA"]}}
{"id": "eval_00918", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nWAT1_ARATH\n\n### Answer:", "output": "Required for secondary wall formation in fibers, especially in short days conditions.  Promotes indole metabolism and transport (e. g.  tryptophan, neoglucobrassicin and auxin (indole-3-acetic acid)).  May prevent salicylic-acid (SA) accumulation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00919", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD3D, IL7R, CD4, CD45RO, S100A4", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD4, IL7R, CD45RO support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "lung", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD3D", "IL7R", "CD4", "CD45RO", "S100A4"]}}
{"id": "eval_00920", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nIRX2 GCG TTR\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as TTR, GCG, IRX2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "blood", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IRX2", "GCG", "TTR"]}}
{"id": "eval_00921", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: Nc1ccc(O)c(C(=O)O)c1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "Nc1ccc(O)c(C(=O)O)c1", "compound_name": null, "pubchem_cid": null, "MW": 153.14, "LogP": 0.67, "TPSA": 83.55, "QED": 0.41, "functional_groups": ["aromatic ring", "hydroxyl", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_00922", "category": "Cell", "instruction": "Based on the marker genes, describe the enriched biological pathway.", "input": "GCG, IRX2, TTR", "output": "The marker profile suggests functional enrichment consistent with the typical biology of pancreatic alpha cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "colon", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_pathway", "top_genes": ["GCG", "IRX2", "TTR"]}}
{"id": "eval_00923", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CCCCNCC(O)c1ccc(O)cc1", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains aromatic ring, amine, hydroxyl, phenol. Its descriptor profile: MW=209.29, LogP=1.82, TPSA=52.49, HBD=3, HBA=3, QED=0.63.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CCCCNCC(O)c1ccc(O)cc1", "compound_name": null, "pubchem_cid": null, "MW": 209.29, "LogP": 1.82, "TPSA": 52.49, "QED": 0.63, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00924", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CN(C)C(=O)CCNc1ccc(Br)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BBBP", "canonical_smiles": "CN(C)C(=O)CCNc1ccc(Br)cc1", "compound_name": null, "pubchem_cid": null, "MW": 271.16, "LogP": 2.34, "TPSA": 32.34, "QED": 0.91, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "halide"]}}
{"id": "eval_00925", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CN(C)C=Nc1c(C#N)c(CC#N)nn1-c1ccccc1", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, amine, nitrile. Its descriptor profile: MW=278.32, LogP=2.03, TPSA=81.0, HBD=0, HBA=4, QED=0.63.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "CN(C)C=Nc1c(C#N)c(CC#N)nn1-c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 278.32, "LogP": 2.03, "TPSA": 81.0, "QED": 0.63, "functional_groups": ["aromatic ring", "amine", "nitrile"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00926", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDTLIKEFDVALRAIAGATRTARANPADRLAPDTEQMSADERRHVAGLMRINHVGEVCAQALYQAQKLTARDGAVRAQMDAAAREEEDHLAWCAERLRELGSRPSLLNPLWYAGAFAIGWMAGRAGDRVSLGFVAETERQVEHHLGGHLDRLPEADGRSRAILEQMRDDEIRHGNAARDAGGIPLPAPVRALMRGASRVMTTAAYRI\n\n### Sequence 2:\nMKVTPDTPGRDDPGPGRRIGLDVGTVRIGVAASDRDARLAMPVETVPRETGMKGPDRGDIDRLIDIITEYDAVEVVVGLPRDLQGNGSASVKQAREIAFRIRRRLTATGREIPVRLGDERLTTVVATQALRASGINERDGRKVIDQAAAVEILQTWLDGRATALNTLPETPHTDEKGNFPR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00927", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAWTPLFLFLLTCCPGSNSQAVVTQEPSLTVSPGGTVTLTCGSSTGAVTSGHYPYWFQQKPGQAPRTLIYDTSNKHSWTPARFSGSLLGGKAALTLLGAQPEDEAEYYCLLSYSGAR\n\n### Sequence 2:\nMAIIPDKQDSTVLERKQQKLKPPSMYKVVLLNDDFTPMEFVVMVVQEYFKKDRETATQIMLKVHREGRGVCGVYTRDIASTKVEQVVTHARQAGHPLQCVMEEA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00928", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nNEAP1_ARATH\n\n### Answer:", "output": "Nucleus inner membrane; Single-pass membrane protein. Nucleus, nucleoplasm. Note=Colocalized with bZIP18 in the nucleoplasm.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00929", "category": "Mutation", "instruction": "Describe the effect of mutation P0AEH1-A23E on protein Regulator of sigma-E protease RseP.", "input": "### Wild-type:\nMLSFLWDLASFIVALGVLITVHAFGHFWVARRCGVRVERFSIGFGKALWRRTDKLGTEYVIALIPLGGYVKMLDERAEPVVPELRHHAFNNKSVGQRAAIIAAGPVANFIFAIFAYWLVFIIGVPGVRPVVGEIAANSIAAEAQIAPGTELKAVDGIETPDWDAVRLQLVDKIGDESTTITVAPFGSDQRRDVKLDLRHWAFEPDKEDPVSSLGIRPRGPQIEPVLENVQPNSAASKAGLQAGDRIVKVDGQPLTQWVTFVMLVRDNPGKSLALEIERQGSPLSLTLIPESKPGNGKAIG\n\n### Mutant:\nMLSFLWDLASFIVALGVLITVHEFGHFWVARRCGVRVERFSIGFGKALWRRTDKLGTEYVIALIPLGGYVKMLDERAEPVVPELRHHAFNNKSVGQRAAIIAAGPVANFIFAIFAYWLVFIIGVPGVRPVVGEIAANSIAAEAQIAPGTELKAVDGIETPDWDAVRLQLVDKIGDESTTITVAPFGSDQRRDVKLDLRHWAFEPDKEDPVSSLGIRPRGPQIEPVLENVQPNSAASKAGLQAGDRIVKVDGQPLTQWVTFVMLVRDNPGKSLALEIERQGSPLSLTLIPESKPGNGKAIG\n\n### Answer:", "output": "High basal sigma-E activity, sigma-E induced. Proteolysis of RseA. Mutations affecting the E23 residue of YaeL enhance its ability to support cell growth. This mutation may enhance the stability or function of the RseA protein, potentially enhancing its ability to inhibit the transcriptional activity of sigma(E) and modulate the stress response pathway in Escherichia coli. Helix-stabilizing residue substitution in the substrate transmembrane segment destabilizes the substrate-RseP interaction.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00930", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMQLTEQQHDKLSKVQLDESWKHSLAEFLVSSRMDELRQFLIEQKNQDKVIYPPSKQIFNALNTTPLSAVKVVILGQDPYHGPNQANGLSFSVQKGIVLPPSLRNIFHELNTDLGIPVPKHGDLTKWADQGVLLLNSVLTVEAGQPTSHQKRGWEQFTDSIIDVLNEQREHVVFILWGAYAQRKGQRIDREKHLVLKAAHPSPLAANRGGFFGCKVFSKTNNYLKQHGIEPIDWQLDA\n\n### Sequence 2:\nMQLTEQQQDKLSKVQLEESWKRSLTSFLLSPYMDSLRDFLFQQKQAQKTIYPPSKQIFNALNITPLDHVKVVILGQDPYHGPNQANGLSFSVQRGVALPPSLRNIFHELHTDLGVPVSRHGDLTKWAEQGVLLLNSVLTVEAGQPTSHHKQGWEEFTDAVIDVLNEQREHIVFILWGAYAQRKGQRINREKHLVLTAAHPSPLAANRGGFFGCKVFSKTNQYLKQHGIEPIDWQLDA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00931", "category": "Mutation", "instruction": "Describe the effect of mutation Q6ZVD8-F783V on protein PH domain leucine-rich repeat-containing protein phosphatase 2.", "input": "### Wild-type:\nMKRNGSRNCLNRRSRFGSRERDWLREDVKRGCVYLYGADTTTATTTTTTSSSSSSSSSSSDLHLVLCTVETPASEICAGEGRESLYLQLHGDLVRRLEPTERPLQIVYDYLSRLGFDDPVRIQEEATNPDLGCMIRFYGEKPCHMDRLDRILLSGIYNVRKGKTQLHKWAERLVVLCGTCLIVSSVKDCQTGKMHILPLVGGKIEEVKRRQYSLAFSSAGAQAQTYHVSFETLAEYQRWQRQASKVVSQRISTVDLSCYSLEEVPEHLFYSQDITYLNLRHNFMQLERPGGLDTLYKFSQ\n\n### Mutant:\nMKRNGSRNCLNRRSRFGSRERDWLREDVKRGCVYLYGADTTTATTTTTTSSSSSSSSSSSDLHLVLCTVETPASEICAGEGRESLYLQLHGDLVRRLEPTERPLQIVYDYLSRLGFDDPVRIQEEATNPDLGCMIRFYGEKPCHMDRLDRILLSGIYNVRKGKTQLHKWAERLVVLCGTCLIVSSVKDCQTGKMHILPLVGGKIEEVKRRQYSLAFSSAGAQAQTYHVSFETLAEYQRWQRQASKVVSQRISTVDLSCYSLEEVPEHLFYSQDITYLNLRHNFMQLERPGGLDTLYKFSQ\n\n### Answer:", "output": "Decreases activity. Mutation of F783 to V impairs the activity of PHLPP2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00932", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "Cc1nnc(N)nc1C=Cc1cccs1", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains no standard functional groups detected. Its descriptor profile: MW=218.29, LogP=1.99, TPSA=64.69, HBD=1, HBA=5, QED=0.84.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "Cc1nnc(N)nc1C=Cc1cccs1", "compound_name": null, "pubchem_cid": null, "MW": 218.29, "LogP": 1.99, "TPSA": 64.69, "QED": 0.84, "functional_groups": [], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00933", "category": "Mutation", "instruction": "Describe the effect of mutation P02699-Y257M on protein Rhodopsin.", "input": "### Wild-type:\nMNGTEGPNFYVPFSNKTGVVRSPFEAPQYYLAEPWQFSMLAAYMFLLIMLGFPINFLTLYVTVQHKKLRTPLNYILLNLAVADLFMVFGGFTTTLYTSLHGYFVFGPTGCNLEGFFATLGGEIALWSLVVLAIERYVVVCKPMSNFRFGENHAIMGVAFTWVMALACAAPPLVGWSRYIPEGMQCSCGIDYYTPHEETNNESFVIYMFVVHFIIPLIVIFFCYGQLVFTVKEAAAQQQESATTQKAEKEVTRMVIIYVIAFLICWLPYAGVAFYIFTHQGSDFGPIFMTIPAFFAKTSAV\n\n### Mutant:\nMNGTEGPNFYVPFSNKTGVVRSPFEAPQYYLAEPWQFSMLAAYMFLLIMLGFPINFLTLYVTVQHKKLRTPLNYILLNLAVADLFMVFGGFTTTLYTSLHGYFVFGPTGCNLEGFFATLGGEIALWSLVVLAIERYVVVCKPMSNFRFGENHAIMGVAFTWVMALACAAPPLVGWSRYIPEGMQCSCGIDYYTPHEETNNESFVIYMFVVHFIIPLIVIFFCYGQLVFTVKEAAAQQQESATTQKAEKEVTRMVIIMVIAFLICWLPYAGVAFYIFTHQGSDFGPIFMTIPAFFAKTSAV\n\n### Answer:", "output": "Causes shift to the inactivated conformation. The mutation in rhodopsin results in a constitutively inactive mutant, decreasing the inherent basal activity of the receptor. This mutation is associated with a pathological outcome.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00934", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\n3MGH_CLOD6\n\n### Answer:", "output": "Belongs to the DNA glycosylase MPG family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00935", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMIQSFTRLNVADNSGAKEIMCIKVLGGSHKRYASVGSVIVASVKKAIPNGKVKRGQVVKAVVVRTKKEIQRKNGSLVRFDDNAAVILDAKKDPVGTRIFGPVSREVRYANFMKIISLAPEVV\n\n### Sequence 2:\nMIQSFTRLNVADNSGAKEIMCIKVLGGSHKRYASVGSVIVASVKKAIPNGKVKRGQVVKAVVVRTKKEIQRKNGSLVRFDDNAAVILDAKKDPVGTRIFGPVSREVRYANFMKIISLAPEL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00936", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNLQHHFLIAMPALQDPIFRRSVVYICEHNTNGAMGIIVNKPLENLKIEGILEKLKITPEPRDESIRLDKPVMLGGPLAEDRGFILHTPPSNFASSIRISDNTVMTTSRDVLETLGTDKQPSDVLVALGYASWEKGQLEQEILDNAWLTAPADLNILFKTPIADRWREAAKLIGVDILTMPGVAGHA\n\n### Sequence 2:\nMNLQHHFLIAMPALQDPIFRRSVVYICEHNTNGAMGIIVNKPLENLKIEGILEKLKITPEPRDESIRLDKPVMLGGPLAEDRGFILHTPPSNFASSIRISDNTVMTTSRDVLETLGTDKQPSDILVALGYASWEKGQLEQEILDNAWLTAPADLNILFKTPIADRWREAAKLIGVDILTMPGVAGHA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00937", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDLKKYIYDIPDFPSPGIIFRDITPLLQNPETFRRTVELLAEKVVDLRPTHVVAIESRGFMFGAPLAYKLGLGFVPVRKEGKLPRESISVSYDLEYGNNTLEIHTDALKPGDRVVIVDDVLATGGTMKATVELCERLGAKVEALLFVIELLALEGRKKLTGKKVISLVQY\n\n### Sequence 2:\nMDFDKYIRVIDDFPKPGISFKDITTLLKDGEAYRAAVDAIVERVRESQPDLIVGPEARGFLLGAPVAYALGIGFVPVRKPGKLPGKTVSETYELEYGSDTLEVHADAIQPGQRIAIVDDLLATGGTTSATARLIEKTGAQVAGMSFLIELGFLEGRKRLEGYEVFSLIKY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00938", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "SATB2, SLC17A7, RBFOX3", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as SLC17A7, RBFOX3, SATB2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "spleen", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SATB2", "SLC17A7", "RBFOX3"]}}
{"id": "eval_00939", "category": "Mutation", "instruction": "Describe the effect of mutation P05166-N536D on protein Propionyl-CoA carboxylase beta chain, mitochondrial.", "input": "### Wild-type:\nMAAALRVAAVGARLSVLASGLRAAVRSLCSQATSVNERIENKRRTALLGGGQRRIDAQHKRGKLTARERISLLLDPGSFVESDMFVEHRCADFGMAADKNKFPGDSVVTGRGRINGRLVYVFSQDFTVFGGSLSGAHAQKICKIMDQAITVGAPVIGLNDSGGARIQEGVESLAGYADIFLRNVTASGVIPQISLIMGPCAGGAVYSPALTDFTFMVKDTSYLFITGPDVVKSVTNEDVTQEELGGAKTHTTMSGVAHRAFENDVDALCNLRDFFNYLPLSSQDPAPVRECHDPSDRLVP\n\n### Mutant:\nMAAALRVAAVGARLSVLASGLRAAVRSLCSQATSVNERIENKRRTALLGGGQRRIDAQHKRGKLTARERISLLLDPGSFVESDMFVEHRCADFGMAADKNKFPGDSVVTGRGRINGRLVYVFSQDFTVFGGSLSGAHAQKICKIMDQAITVGAPVIGLNDSGGARIQEGVESLAGYADIFLRNVTASGVIPQISLIMGPCAGGAVYSPALTDFTFMVKDTSYLFITGPDVVKSVTNEDVTQEELGGAKTHTTMSGVAHRAFENDVDALCNLRDFFNYLPLSSQDPAPVRECHDPSDRLVP\n\n### Answer:", "output": "In PA-2; affects heteromeric and homomeric assembly. The mutation , located at the carboxyl-terminal end of the PCCB gene, was found to inhibit alpha-beta heteromeric and/or the beta-beta homomeric interaction independently of the cultivation temperature, reflecting its primary effect on the assembly.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00940", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "S=c1[nH]nc(C23CC4CC(CC(C4)C2)C3)n1-c1ccc(Cl)cc1", "output": "The molecule contains the following functional groups: aromatic ring, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "S=c1[nH]nc(C23CC4CC(CC(C4)C2)C3)n1-c1ccc(Cl)cc1", "compound_name": null, "pubchem_cid": null, "MW": 345.9, "LogP": 5.05, "TPSA": 33.61, "QED": 0.77, "functional_groups": ["aromatic ring", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "halide"]}}
{"id": "eval_00941", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nC1QA:medium CD163:high C1QB:medium CD68:high MRC1:medium\n</cell>", "output": "This cell is most consistent with macrophage, a tissue macrophage. Key markers such as CD163, MRC1, C1QA support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "liver", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["C1QA", "CD163", "C1QB", "CD68", "MRC1"]}}
{"id": "eval_00942", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nHSCB_BORBR\n\n### Answer:", "output": "Belongs to the HscB family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00943", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRS10_GEOSM\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uS10 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00944", "category": "Mutation", "instruction": "Describe the effect of mutation Q40235-N191D on protein Receptor-like protein Cf-9.", "input": "### Wild-type:\nMDCVKLVFLMLYTFLCQLALSSSLPHLCPEDQALSLLQFKNMFTINPNASDYCYDIRTYVDIQSYPRTLSWNKSTSCCSWDGVHCDETTGQVIALDLRCSQLQGKFHSNSSLFQLSNLKRLDLSFNNFTGSLISPKFGEFSNLTHLDLSHSSFTGLIPSEICHLSKLHVLRICDQYGLSLVPYNFELLLKNLTQLRELNLESVNISSTIPSNFSSHLTTLQLSGTELHGILPERVFHLSNLQSLHLSVNPQLTVRFPTTKWNSSASLMTLYVDSVNIADRIPKSFSHLTSLHELYMGRCN\n\n### Mutant:\nMDCVKLVFLMLYTFLCQLALSSSLPHLCPEDQALSLLQFKNMFTINPNASDYCYDIRTYVDIQSYPRTLSWNKSTSCCSWDGVHCDETTGQVIALDLRCSQLQGKFHSNSSLFQLSNLKRLDLSFNNFTGSLISPKFGEFSNLTHLDLSHSSFTGLIPSEICHLSKLHVLRICDQYGLSLVPYNFELLLKDLTQLRELNLESVNISSTIPSNFSSHLTTLQLSGTELHGILPERVFHLSNLQSLHLSVNPQLTVRFPTTKWNSSASLMTLYVDSVNIADRIPKSFSHLTSLHELYMGRCN\n\n### Answer:", "output": "Abolishes resistance to C.fulvum.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00945", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nFLGI_BURMA\n\n### Answer:", "output": "Assembles around the rod to form the L-ring and probably protects the motor/basal body from shearing forces during rotation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00946", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nTP63 KRT14 KRT5\n</cell>", "output": "This cell is most consistent with basal epithelial cell, a basal epithelial cell. Key markers such as TP63, KRT14, KRT5 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "basal epithelial cell", "tissue": "blood", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TP63", "KRT14", "KRT5"]}}
{"id": "eval_00947", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKVSLFVTCLVDMFQTNVGKATVELLERLGCEVDFPEGQICCGQPAYNSGYVHDAKKAMKRMIETFQDSEYVVSPSGSCTTMFREYPHLFQDDPKWADKAKKLADKTYELTDFIVNVLGVEDVGATLHTKATLHTSCHMTRLLGVRKEPMKLLSHVKGLQFTELPGKHNCCGFGGTFSVKMAQISEQMVDEKVECVEETGAEVLIGADCGCLMNIGGRLGRKDKNVKVMHIAEVLNSR\n\n### Sequence 2:\nMKVSLFVTCLVDMFQTNVGKATVEVLERLGCEVDFPEGQICCGQPAYNSGYVTDAKKAMKRMIAAFEEAEYVVSPSGSCTTMFREYPHLFQDDPKWAAKAQQLADKTYELTDFIVNVLGVEDVGAVLHKKATVHTSCHMTRLLGVSEEPMKLLRHVKGLELTALPGKHQCCGFGGTFSVKMAQISEQMVDEKVACVEDTEAEVLIGADCGCLMNIGGRLDRKDKNVRVMHIAEVLNSR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00948", "category": "Mutation", "instruction": "Describe the effect of mutation Q7TSA3-P41E on protein B- and T-lymphocyte attenuator.", "input": "### Wild-type:\nMKTVPAMLGTPRLFREFFILHLGLWSILCEKATKRNDEECPVQLTITRNSKQSARTGELFKIQCPVKYCVHRPNVTWCKHNGTICVPLEVSPQLYTSWEENQSVPVFVLHFKPIHLSDNGSYSCSTNFNSQVINSHSVTIHVRERTQNSSEHPLITVSDIPDATNASGPSTMEERPGRTWLLYTLLPLGALLLLLACVCLLCFLKRIQGKEKKPSDLAGRDTNLVDIPASSRTNHQALPSGTGIYDNDPWSSMQDESELTISLQSERNNQGIVYASLNHCVIGRNPRQENNMQEAPTEYA\n\n### Mutant:\nMKTVPAMLGTPRLFREFFILHLGLWSILCEKATKRNDEECEVQLTITRNSKQSARTGELFKIQCPVKYCVHRPNVTWCKHNGTICVPLEVSPQLYTSWEENQSVPVFVLHFKPIHLSDNGSYSCSTNFNSQVINSHSVTIHVRERTQNSSEHPLITVSDIPDATNASGPSTMEERPGRTWLLYTLLPLGALLLLLACVCLLCFLKRIQGKEKKPSDLAGRDTNLVDIPASSRTNHQALPSGTGIYDNDPWSSMQDESELTISLQSERNNQGIVYASLNHCVIGRNPRQENNMQEAPTEYA\n\n### Answer:", "output": "In strain: 129/SvEv; requires 2 nucleotide substitutions.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00949", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMPKQAIIPAGTTTPIAPFVPGTMADGIVYVSGTLPFDKDNNVVHVGDASAQTRHVLETIQNVIATAGGTMDDVTFNMIMIKDWADYAKVNAVYAEFFPGTKPARYCIQCGLVKPDALVEIASIAHVGNKA\n\n### Sequence 2:\nMPKQVIIPPGTTTPIAPFVPGTLADGVVYVSGTLPFDKQNNVVHIGDPKAQTRHVLETIKSVIETAGGSMADVTFNSIFITDWTNYAAINEVYAEFFPGDKPARFCIQCGLVKPDALVEIASVAHIG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00950", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nSSKWEACCDRCACTKSIPPQCHCADIRLNSCHSACESCACTHSIPAQCRCFDITDFCYKPCSG\n\n### Sequence 2:\nSSKWEACCDRCACTKSIPPQCHCADIRLNSCHSACESCACTHSIPAQCRCFDITDFCYKPCSG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00951", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAKKSGEGQKIISDNRQARFNYHIIETYEAGVALQGTEVKSIREGKVNLRDGYALIRNNEAWLLNVHISPYQKAGDFFNHEPRRSRRLLLHKQEIRKLIGKVEQQGLTLVPLRMYFKRGWVKVDIALAQGKKLYDKRETIKRRDDKRAMQRAIKQY\n\n### Sequence 2:\nMSNASGGGPRPRRRDGVDPALRSRARRRALQAVYAWQISGGVAKQVIAHFAHEQAYEVADLAYFEDLVEGVLTHCAELDEKLTPYLDRTIEEVDAIERAVLRLGAYELLYRQDVPYRVVINEAIMTAKRFGSKYGHTYVNGVLDRAALALRKVEVLG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00952", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CPA3, TPSAB1, KIT, TPSB2", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as CPA3, TPSAB1, KIT support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CPA3", "TPSAB1", "KIT", "TPSB2"]}}
{"id": "eval_00953", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMRRRQLIRQLLERDKTPLAILFMAAVVGTLVGLAAVAFDKGVAWLQNQRMGALVHTADNYPLLLTVAFLCSAVLAMFGYFLVRKYAPEAGGSGIPEIEGALEDQRPVRWWRVLPVKFFGGLGTLGGGMVLGREGPTVQIGGNIGRMVLDIFRLKGDEARHTLLATGAAAGLAAAFNAPLAGILFIIEEMRPQFRYTLISIKAVFIGVIMSTIMYRIFNHEVALIDVGKLSDAPLNTLWLYLILGIIFGIFGPIFNKWVLGMQDLLHRVHGGNITKWVLMGGAIGGLCGLLGFVAPATSGGGFNLIPIATAGNFSMGMLVFIFVARVITTLLCFSSGAPGGIFAPMLALGTVLGTAFGMVAVELFPQYHLEAGTFAIAGMGALLAASIRAPLTGIILVLEMTDNYQLILPMIITGLGATLLAQFTGGKPLYSAILARTLAKQEAEQK\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHHHHHTCSCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTSSSHHHHHHHHHHHHHHHHHHHHHHHHHHCGGGSSCSHHHHHHHHHTCSCCCHHHHHHHHHHHHHHHHHTTCSCBSHHHHHHHHHHHHHHHHHHTTCCSHHHHHHHHHHHHHHHHHHHHTCHHHHHHHHHHTTSCSSSCCCCCHHHHHHHHHHHHHHHHHTTTTCCSCCCCCCCCCCGGGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTCHHHHHHHHHHHHHHHHHHHHHCGGGSSCCSTHHHHHHTTCSCHHHHHHHHHHHHHHHHHHHTTTCSSBSHHHHHHHHHHHHHHHHHHHHHHCGGGTCCHHHHHHHHHTHHHHHHTCCHHHHHHHHHHHHCCGGGHHHHHHHHHHHHHHHHHTTCCCHHHHHHHHHHHHHHHTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00954", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "CD3D, FOXP3, CD4, IL2RA, CTLA4", "output": "The expression pattern is consistent with regulatory T cell in a context associated with Alzheimer disease, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "ovary", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CD3D", "FOXP3", "CD4", "IL2RA", "CTLA4"]}}
{"id": "eval_00955", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSECA_RHDSA\n\n### Answer:", "output": "Belongs to the SecA family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00956", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "KRT8, CDH1, EPCAM, KRT18", "output": "The marker profile suggests functional enrichment consistent with the typical biology of epithelial cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "epithelial cell", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["KRT8", "CDH1", "EPCAM", "KRT18"]}}
{"id": "eval_00957", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCAPSD_CLVN\n\n### Answer:", "output": "Encapsidates the viral DNA into characteristic twinned ('geminate') particles.  Binds the genomic viral ssDNA and shuttles it into and out of the cell nucleus.  The CP of bipartite geminiviruses is not required for cell-to-cell or systemic movement.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00958", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDNLJ_VIBA3\n\n### Answer:", "output": "Belongs to the NAD-dependent DNA ligase family. LigA subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00959", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nSPRT_SERP5\n\n### Answer:", "output": "Cytoplasm.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00960", "category": "Mutation", "instruction": "Describe the effect of mutation Q99418-D280R on protein Cytohesin-2.", "input": "### Wild-type:\nMEDGVYEPPDLTPEERMELENIRRRKQELLVEIQRLREELSEAMSEVEGLEANEGSKTLQRNRKMAMGRKKFNMDPKKGIQFLVENELLQNTPEEIARFLYKGEGLNKTAIGDYLGEREELNLAVLHAFVDLHEFTDLNLVQALRQFLWSFRLPGEAQKIDRMMEAFAQRYCLCNPGVFQSTDTCYVLSFAVIMLNTSLHNPNVRDKPGLERFVAMNRGINEGGDLPEELLRNLYDSIRNEPFKIPEDDGNDLTHTFFNPDREGWLLKLGGGRVKTWKRDWFILTDNCLYYFEYTTDKEP\n\n### Mutant:\nMEDGVYEPPDLTPEERMELENIRRRKQELLVEIQRLREELSEAMSEVEGLEANEGSKTLQRNRKMAMGRKKFNMDPKKGIQFLVENELLQNTPEEIARFLYKGEGLNKTAIGDYLGEREELNLAVLHAFVDLHEFTDLNLVQALRQFLWSFRLPGEAQKIDRMMEAFAQRYCLCNPGVFQSTDTCYVLSFAVIMLNTSLHNPNVRDKPGLERFVAMNRGINEGGDLPEELLRNLYDSIRNEPFKIPEDDGNDLTHTFFNPDREGWLLKLGGGRVKTWKRRWFILTDNCLYYFEYTTDKEP\n\n### Answer:", "output": "Does not enhance ARL4D GTP-dependent interaction but promotes targeting to the plasma membrane mediated by ARL4C, ARL4C and ARL4D. This mutation may improve the function of ARNO in regulating endocytosis, actin dynamics, and cell adhesion.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00961", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTRMY_PYRFU\n\n### Answer:", "output": "Specifically catalyzes the N1-methylation of pseudouridine at position 54 (Psi54) in tRNAs.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00962", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CCC(C(=O)OCCN(CC)CC)c1ccccc1\nDescriptors: MW=263.38, LogP=3.07, TPSA=29.54, HBD=0, HBA=3", "output": "Descriptors: MW=263.38, LogP=3.07, TPSA=29.54, HBD=0, HBA=3, QED=0.68, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "CCC(C(=O)OCCN(CC)CC)c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 263.38, "LogP": 3.07, "TPSA": 29.54, "QED": 0.68, "functional_groups": ["aromatic ring", "ester", "amine", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": null, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00963", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDWLTILGISVALAMDAFAVALAAGAVISPITGRHLFRLGFHFGLFQALMPIGGWLLGMTVQKWISAYDHWIAFGLLAYVGGRMVHEAFEEDDEEKTPSDPTKGMTMVMLSVATSIDAFAVGLSIAMLGVSVWLPATVIGLVAGVLTVAGMLLGRRLGEKWGKRVEICGGLVLCLIGLKILLEHTLLK\n\n### Sequence 2:\nMIIYLHGFDSNSPGNHEKVLQLQFIDPDVRLVSYSTRHPKHDMQHLLKEVDKMLQLNVDERPLICGVGLGGYWAERIGFLCDIRQVVFNPNLFPYENMEGKIDRPEEYADIATKCVTNFREKNRDRCLVILSRHDEALDSQRSAQALHPYYEIVWDEEQTHKFKNISPHLQRIKAFKTLG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00964", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGLVPRGSHMKLQFNLKAYFKTSADPTPAKDAIAALFEEANSTLLTRGAPEGQGAKVTEWKLGEDRIELTLQSGRYVRVHDAIFRLRKQLAEALGKKYKIGIRGIEVESFIIKVPADHELRMLKVPYIKSMENIEGGIQLELEVGEAEMKNRVPDRILTLLEEKIEAAQYGAKAEHWNLLWQREPMEHPFKEDPTQAMMKEGWLKRGSSRGQWIHGPQSARIFRTFEKIVLEELLEPLGYREMIFPKLVTWEVWMKSGHAKGVYPEIYYVCPPQTRDPDYWEEVADYYKVTHEVPTKLIKEKIAEPIGGMCYAQCPPFWMYVAGETLPNEEIPVKVFDRSGTSHRYESGGIHGIERVDEFHRIEIVWIGTKEEVLKCAEELHDRYMHIFNDILDIEWRKARVTPWFMAQEGLLGLAEENTVGTTDYEACLPYRGPDGEWLEFQNVSINGDKYPKGFNVKLQSGDELWSGCSGVGLERWAAVFLAQKGLDPANWPEEFRNRVGEMPKGIRFL\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCEEEEEEECCCCCCCCCCCCTTHHHHTTIIIIITCCCSSSSCSSBCCCCCCCCCCCCBCCEECSSSCTTTSTTHHHHHTCCCCCCCCCCCCCCCEEEEEEEEEECSSCCCCCCCTTEEEEEECSSEEEEEECCCHHHHHTTHHHHHHHHHHHHHHHCCCCCCCCCEEEEEECCCCCCCCCSCHHHHHHHTTSEEECSSTTCEEECHHHHHHHHHHHHHHHHHTTTTTTCEECBCCSEEEHHHHHHHSGGGTCGGGCCEEECBSCCCHHHHHHHHHHHHHHSSCCHHHHHHHBCSCCEEECSSSSGGGGGGTTTCEECGGGCSEEEEECSSEEECCCCSSCCCTTSCSEEEEEEEEEEECHHHHHHHHHHHHHHHHHHHHHTSCCCEEEEEESCHHHHHTTCCSCCSSSSCEEEEEEECCGGGCTTCCCEEEEEEEEEETHHHHHHTCEETTCCCCEEEEEEEEHHHHHHHHHHHHCSCTTTSCHHHHHHHCSCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00965", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: CC(=O)NC(C)(C)C1CCCC(C(=O)NC(Cc2ccccc2)C(O)C[NH2+]Cc2cccc(C(C)C)c2)C1\nDescriptors: MW=522.75, LogP=3.68, TPSA=95.04, HBD=4, HBA=3", "output": "Descriptors: MW=522.75, LogP=3.68, TPSA=95.04, HBD=4, HBA=3, QED=0.34, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "BACE", "canonical_smiles": "CC(=O)NC(C)(C)C1CCCC(C(=O)NC(Cc2ccccc2)C(O)C[NH2+]Cc2cccc(C(C)C)c2)C1", "compound_name": null, "pubchem_cid": null, "MW": 522.75, "LogP": 3.68, "TPSA": 95.04, "QED": 0.34, "functional_groups": ["aromatic ring", "amide", "hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00966", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVQSVSAIRRVSRLRRHARLRKKVSGTSERPRLVVNRSARHIHVQLVNDVTGTTVAAASSIEADVRGLQGDKKVRSVRVGQLIAERAKAAGINTVVFDRGGYTYGGRIAALADSVRENGLNF\n\n### Sequence 2:\nMEQYLQAFEFVEEMVVLPKYLSWELYHHLAVLLREKYPKTYKNKGYIFNIKVKSILDNRITPTGQIVLVVMFQSDLYVPQVGHVFTERIRVNSVDDRYQWITIEPLTVFLRSNIPYKPNTLVTVQICSIKMDNTLCFGTILD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00967", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMGFFDRLKAGLAKTRERLLKAIPWGGNLEEVLEELEMALLAADVGLSATEEILQEVRASGLKEAVKEKLVGMLEPDKPKPVEPKGRVVLVVGVNGVGKTTTIAKLGRYYQNLGKKVMFCAGDTFRAAGGTQLSEWGKRLSIPVIQGPEGTDPAALAYDAVQAMKARGYDLLFVDTAGRLHTKHNLMEELKKVKRAIAKADPEEPKEVWLVLDAVTGQNGLEQAKKFHEAVGLTGVIVTKLDGTAKGGVLIPIVRTLKVPIKFVGVGEGPDDLQPFDPEAFVEALLED\n\n### Answer:", "output": "<SEQ_3Di>DVCLVLLLLLQVLLQPPQLQVQPPPDDDCPSLVSVLVSLVSLLLDPVLSVVLSVQQVPDPHLVSSLVSQLVLLALVVADDDDFPFAEEEEDEAPPLQRLVVVLLVCLVVVVVVFQEAEEQLQQQDDCSRVVNCVSCVVRVHHYQYDHRPDLSLVSLLVVLVVCVVVVGTHYYYYHHRDDPDPDDRLVSVLSNQVSNCVNPVVPDSAYEYEGEQVVRPVVQVVVVSNCPRRNHQEYEYEDSSPDSNLSSSSVNCNRSVHHHAWYQSDSDSPRTGGDDSSSNSCSSSDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00968", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRM49_BOVIN\n\n### Answer:", "output": "39S ribosomal protein L49, mitochondrial, L49mt, MRP-L49.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00969", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGKNKDVRITITLECTNCPQNKLKRYPGISRYTTQKNRRNTPNRLELKKFCPHCSLHTLHKEIKK\n\n### Sequence 2:\nMAKVKITLVKSLIGRKKDQIATVNALGLKKIGNIVEHEETPQISGMIKKVSYLLKVEEA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00970", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRBFA_BACC1\n\n### Answer:", "output": "Ribosome-binding factor A.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00971", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CC1(O)CC(c2nc(-c3ccc4ccc(-c5ccccc5)nc4c3)c3c(N)nccn23)C1", "output": "The molecule contains the following functional groups: aromatic ring, hydroxyl, pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "ClinTox", "canonical_smiles": "CC1(O)CC(c2nc(-c3ccc4ccc(-c5ccccc5)nc4c3)c3c(N)nccn23)C1", "compound_name": null, "pubchem_cid": null, "MW": 421.5, "LogP": 4.82, "TPSA": 89.33, "QED": 0.44, "functional_groups": ["aromatic ring", "hydroxyl", "pyridine"], "moleculenet_labels": {"FDA_APPROVED": 0, "CT_TOX": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "hydroxyl", "pyridine"]}}
{"id": "eval_00972", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQSKSRKIAILGYRSVGKSSLTIQFVEGQFVDSYDPTIENTFTKLITVNGQEYHLQLVDTAGQDEYSDINGYILVYSVTSIKSFEVIKVIHGKLLDMVGPIMLVGNKKDLHMERVISYEEGKALAESWNAAFLESSAKENQTAVDVFRRIILEAEKLE\n\n### Answer:", "output": "<SEQ_3Di>DAAEAEEEEAFAPPQQLPLLVCCQAVVDGDPDDDLADWDWDWDWDAAPNHIYIYIYTSHSYDFLPDSHLEYEYEGEQLDVRRVVRSVSVVVVVCVVVCQYAYEHEPPVPVVRGDHDPVNRVVVCVVVVYYYDYYYSRDSVRSNCVVSVSVVSSVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00973", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "CLDN5, KDR, PECAM1, CDH5, VWF", "output": "This profile is most consistent with endothelial cell, which is commonly found in kidney under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "endothelial cell", "tissue": "kidney", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CLDN5", "KDR", "PECAM1", "CDH5", "VWF"]}}
{"id": "eval_00974", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMHSSLIKLGFLLLLLQVSLSHAQLSPSFYDKTCPQVFDIVTNTIVNALRSDPRIAASILRLHFHDCFVNGCDASILLDNTTSFRTEKDAFGNANSARGFDVIDKMKAAIEKACPRTVSCADMLAIAAKESIVLAGGPSWMVPNGRRDSLRGFMDLANDNLPGPSSTLKQLKDRFKNVGLDRSSDLVALSGGHTFGKSQCQFIMDRLYNFGETGLPDPTLDKSYLATLRKQCPRNGNQSVLVDFDLRTPTLFDNKYYVNLKENKGLIQSDQELFSSPDAADTLPLVRAYADGQGTFFDAFVKAIIRMSSLSPLTGKQGEIRLNCRVVNSKSKIMDVVDDALEFASFM\n\n### Sequence 2:\nLITLG--CLMLHASLSAAQLTPTFYDRSCPNVTNIVRETIVNELRSDPRIAASILRLHFHDCFVNGCDASILLDNTTSFRTEKDAFGNANSARGFPVIDRMKAAVERACPRTVSCADMLTIAAQQSVTLAGGPSWRVPLGRRDSLQAFLELANANLPAPFFTLPQLKASFRNVGLDRPSDLVALSGGHTFGKNQCQFILDRLYNFSNTGLPDPTLNTTYLQTLRGLCPLNGNRSALVDFDLRTPTVFDNKYYVNLKERKGLIQSDQELFSSPNATDTIPLVRAYADGTQTFFNAFVEAMNRMGNITPTTGTQGQIRLNCRVVNSNSLLHDVV-DIVDFVSSM\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00975", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSAITSNLTHHFLIAMPNMADPNFVRSLTYIAEHNEQGALGIIVNRPIDMTLAALFERIDVPLEADGLAGQPVYFGGPVQTDRGFVLHRPAGEWHSTLVVNDEVGLTSSRDILQAVGSSGEPPEVLVTLGYAGWTAGQLEQEIADNSWLTVPADLAIVFGLPPEERLAAAMQMLGIDFANLSESAGHA\n\n### Sequence 2:\nNLTDNFLIAMPTLEDPYFSNALVYICEHNENGALGIIVNRPIDMNLASLLEKIDIKLEAENLADMPVYFGGPVQLDRGFVLHRPIGQWQSTLAINSDVGLTSSRDVLSSVGSAGLPAEILVTLGYAGWDAGQLEEELAQNSWLTVPAKASILFDLPPEERLPAAMQKLGISFTQLSDVAGHA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00976", "category": "Mutation", "instruction": "Describe the effect of mutation P35555-V984I on protein Fibrillin-1 [Cleaved into: Asprosin].", "input": "### Wild-type:\nMRRGRLLEIALGFTVLLASYTSHGADANLEAGNVKETRASRAKRRGGGGHDALKGPNVCGSRYNAYCCPGWKTLPGGNQCIVPICRHSCGDGFCSRPNMCTCPSGQIAPSCGSRSIQHCNIRCMNGGSCSDDHCLCQKGYIGTHCGQPVCESGCLNGGRCVAPNRCACTYGFTGPQCERDYRTGPCFTVISNQMCQGQLSGIVCTKTLCCATVGRAWGHPCEMCPAQPHPCRRGFIPNIRTGACQDVDECQAIPGLCQGGNCINTVGSFECKCPAGHKLNEVSQKCEDIDECSTIPGICE\n\n### Mutant:\nMRRGRLLEIALGFTVLLASYTSHGADANLEAGNVKETRASRAKRRGGGGHDALKGPNVCGSRYNAYCCPGWKTLPGGNQCIVPICRHSCGDGFCSRPNMCTCPSGQIAPSCGSRSIQHCNIRCMNGGSCSDDHCLCQKGYIGTHCGQPVCESGCLNGGRCVAPNRCACTYGFTGPQCERDYRTGPCFTVISNQMCQGQLSGIVCTKTLCCATVGRAWGHPCEMCPAQPHPCRRGFIPNIRTGACQDVDECQAIPGLCQGGNCINTVGSFECKCPAGHKLNEVSQKCEDIDECSTIPGICE\n\n### Answer:", "output": "In MFS. The mutation leads to the substitution of valine by isoleucine (), both uncharged amino acids, which only differ in a single methyl group. This defect was identified in a patient with a classic phenotype of Marfan Syndrome (MfS), suggesting that conservative substitutions in this region may lead to a less severe phenotype of the disease.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00977", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCAP_HYDVD\n\n### Answer:", "output": "Adenylyl cyclase-associated protein, CAP.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00978", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: COc1cccc(OC)c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)(C)[C@@H]2C(=O)O\nDescriptors: MW=380.42, LogP=0.95, TPSA=105.17, HBD=2, HBA=6", "output": "Descriptors: MW=380.42, LogP=0.95, TPSA=105.17, HBD=2, HBA=6, QED=0.73, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "BBBP", "canonical_smiles": "COc1cccc(OC)c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)(C)[C@@H]2C(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 380.42, "LogP": 0.95, "TPSA": 105.17, "QED": 0.73, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_00979", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nHHAAPLPELLSNNGKHALMVDGAPYIILGSQTNNSSNYPDALKDVWPSMEKMGANTLSIPVAWEQIEPVEGQFDFSFVDVLLKEARQRKVRLVLLWFATWKNNAPHYAPAWVKLDNARFPRVVKEDGDTLNSLSPLGQNTLAADKKAFVELMKYLAKRDKDHTVIMVQVQNEVGTYGAVRDYSPMAQAVFNAAVPDDLIQKLQLKPGTWSQVFGRDADEFFHAYQIARYCDEVTVAGKAIKNLPMYVNVALRNPFNPGLPGQYSSGGGTDNVLHIWKAAAPNIDLIAPDIYFRDYKTVSKVLELYTRPDNALFVAEIGNDQPFARYLFPTLGKGGIGFSPFGMDDTDYTNYPLGAKVYNDETIEQFAQVYRLVNPMMREWARLSYQGQVWGVAEPLDSTTETQKIWNAEAEEKEQHKKDRASALTQQLDLGLWDAEVTYGRPMFWVTPPEGNTPAAGGALIAQLDDNEYLVTAYKARVEFKPSQELAGKKFMIERVEEGRFEKGKWVMERVWNGDQTDWGLNFTDRPHLLRVKMASYSVQ\n\n### Answer:", "output": "<SEQ_3Di>DFDFDAWDWDFDPLAIATDASHFGFFFAEAEAPQPQQALVNVVQFVVLCVLLVTQEYEYEDFQCQQPLDPPRGDCVSVVNVVVVSVVVNHAYEYEYAALDDQQAGPRHHCVAQVDCVQFNFAAAPVGDGDRGTALLRVSRLVRSLVSLQVVLLVCLVPVSSCSYQEYEQHAAHFAAPHQHGPPPSLVVQQQDKDDVLLCVLVVFDIDHLCVTPPPCSSLLSLLLSVLLSSLVSLVSSVVSPNHAYEYEHQEAQPVCRPDDNNGGGNACYLVNVSSNCSRRVSHQAYEYADLDLAQVRVVVVLVSRSTSRHAYERQEYALALSTQLVVFQCVQSLHRYDYYYHSGDPPDDPPVSHDPDDDSVSSVSSSLVVVQQVLCRRVSSVCSRPHHKHKDFQNGAADDVVNVVVQVDDVVVVVVQVVSLVSQWDWDDPAQKIKIKGADDDLADCDGGDADVGGRKMKMWGDPDNFKIKIKIERIFIAMDTPDDDPQWDKDWPWKFWFGADPNDTDGDGTDDDSCPPHGHGHYSHHTIMMTGMDIDHPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00980", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "P2RY12, CSF1R, CX3CR1, TMEM119", "output": "This profile is most consistent with microglia, which is commonly found in tumor under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "tumor", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["P2RY12", "CSF1R", "CX3CR1", "TMEM119"]}}
{"id": "eval_00981", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRS16_RHIWR\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bS16 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00982", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPRHA_PENBI\n\n### Answer:", "output": "Belongs to the PhyH family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00983", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAVPKKKTSKSRRNMRRSHLALGKVNVIVDSQTGEYKLPHHVSLVDGTYNNRLVVTKKIKTEEEVA\n\n### Sequence 2:\nMTDLIHDATLDATGLNCPEPVMMLHKHVRELAAGGVLKVIATDPSTRRDIPKFCVFLGHELLGQQEEAGTYLYWIRKKAD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00984", "category": "Mutation", "instruction": "Describe the effect of mutation Q8IVT2-S397D on protein Mitotic interactor and substrate of PLK1.", "input": "### Wild-type:\nMDRVTRYPILGIPQAHRGTGLVLDGDTSYTYHLVCMGPEASGWGQDEPQTWPTDHRAQQGVQRQGVSYSVHAYTGQPSPRGLHSENREDEGWQVYRLGARDAHQGRPTWALRPEDGEDKEMKTYRLDAGDADPRRLCDLERERWAVIQGQAVRKSSTVATLQGTPDHGDPRTPGPPRSTPLEENVVDREQIDFLAARQQFLSLEQANKGAPHSSPARGTPAGTTPGASQAPKAFNKPHLANGHVVPIKPQVKGVVREENKVRAVPTWASVQVVDDPGSLASVESPGTPKETPIEREIRLA\n\n### Mutant:\nMDRVTRYPILGIPQAHRGTGLVLDGDTSYTYHLVCMGPEASGWGQDEPQTWPTDHRAQQGVQRQGVSYSVHAYTGQPSPRGLHSENREDEGWQVYRLGARDAHQGRPTWALRPEDGEDKEMKTYRLDAGDADPRRLCDLERERWAVIQGQAVRKSSTVATLQGTPDHGDPRTPGPPRSTPLEENVVDREQIDFLAARQQFLSLEQANKGAPHSSPARGTPAGTTPGASQAPKAFNKPHLANGHVVPIKPQVKGVVREENKVRAVPTWASVQVVDDPGSLASVESPGTPKETPIEREIRLA\n\n### Answer:", "output": "No effect on cortical localization; when associated with D-394; D-395; D-471; D-582 and D-586. Phosphorylation of MISP by Plk1 is required for correct mitotic spindle positioning.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00985", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRLUE_XANCP\n\n### Answer:", "output": "Ribosomal large subunit pseudouridine synthase E, rRNA pseudouridylate synthase E, rRNA-uridine isomerase E.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00986", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nIF1A_AERPE\n\n### Answer:", "output": "Seems to be required for maximal rate of protein biosynthesis.  Enhances ribosome dissociation into subunits and stabilizes the binding of the initiator Met-tRNA(I) to 40 S ribosomal subunits (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00987", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSQEKKEELQSEAQVTKEETPQANEAAAEAEAIVNEFDLLQEELNSLKDKYARVHADFENIKKRLEREKYSAVEYANEKFAKDMIPVMDALHMALSSSSSIIDSAEHLEKLKEGIELTLKQLSTALEKHGITMVSHDAPFDPNIHNAIQSVDSDTVESGQIVQTFQTGYKYKDRPLREAMVVVAN\n\n### Sequence 2:\nMATCSIKDWQGNATGEVDLDLPVASAATASHVVYLAFKRQMVNSRQGTASTLTRGEVRGGGRKPWKQKGTGRARAGSIRSPLWRKGGVIFGPKPRDFEIKMNRKERRLALRTALQSRVEDLIVVDEFEGQLAAPKTRELVQAFERWGVDMASQSILLILRERQTNTYLSARNLPNVKVITAGNLNVRDLLATDWIVVTGPALELIKETYGAVA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_00988", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNQTRIDFRQAKFLISAPDIAHLDKHLPGDVGVEIAFAGRSNAGKSSALNALTEQKSLARTSKTPGRTQLINVFELDAQRRLVDLPGYGFAKVPLEMKHKWQNALGEYLQKRACLSGVVVLMDIRHPLKDLDRQMIEWSVASEIPVLALLTKADKLGQSEKMKTVNAVRKELAEFGDWVSVEPFSALKGTGKPKVLAILDAWCHPDWLVEELEQAED\n\n### Sequence 2:\nMTESRIDFRRAKFLISAPDIAHLDQYLPGDVGVEIAFAGRSNAGKSSALNALTEQKSLARTSKTPGRTQLINVFELDAQRRLVDLPGYGFAQVPLALKNKWQQALGEYLQKRACLSGVVVLMDIRHPLKDLDMQMIEWAVASEIPVLALLTKSDKLAQSAKMKTVNEVRLALSEFGDWVQVEPFSSLKGTGKPKVLAILNEWCHPQWLTDELDAANN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00989", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NPD8-A63F on protein Ubiquitin-conjugating enzyme E2 T.", "input": "### Wild-type:\nMQRASRLKRELHMLATEPPPGITCWQDKDQMDDLRAQILGGANTPYEKGVFKLEVIIPERYPAEPPQIRFLTPIYHPNIDSAGRICLDVLKLPPKGAWRPSLNIATVLTSIQLLMSEPNPDDPLMADISSEFKYNKPAFLKNARQWTEKHARQKQKADEEEMLDNLPEAGDSRVHNSTQKRKASQLVGIEKKFHPDV\n\n### Mutant:\nMQRASRLKRELHMLATEPPPGITCWQDKDQMDDLRAQILGGANTPYEKGVFKLEVIIPERYPFEPPQIRFLTPIYHPNIDSAGRICLDVLKLPPKGAWRPSLNIATVLTSIQLLMSEPNPDDPLMADISSEFKYNKPAFLKNARQWTEKHARQKQKADEEEMLDNLPEAGDSRVHNSTQKRKASQLVGIEKKFHPDV\n\n### Answer:", "output": "Increased binding to FANCL. Mutational analysis reveals that the F63 residue of FANCL is required for substrate binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00990", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nHEMH_PROM5\n\n### Answer:", "output": "Belongs to the ferrochelatase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_00991", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGSHMPNLCVSATFNPPVITMLGSALREETVKLLEQRIPTDPVKFLFYPNPDHWRMELSQHFCDDLHKSAVFLTIIEGLEGEGWNLRASNSIRDSESGKDTTKLFFARR\n\n### Answer:", "output": "<SEQ_3Di>DVDAKKKKWKWKDVVTKIKIAIPPDDVVLQVLLQCVLVCVGWHWDFDVDPTMTMTGDDNDGDDPVSVVVNVVSNQVSVVVVQWHWDDKDKDADPVVRMIMMMTMTDGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_00992", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "C=CC=C", "output": "No standard functional groups from the preset library were detected in this molecule.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "ESOL", "canonical_smiles": "C=CC=C", "compound_name": null, "pubchem_cid": null, "MW": 54.09, "LogP": 1.36, "TPSA": 0.0, "QED": 0.4, "functional_groups": [], "moleculenet_labels": {"label": -1.87}, "split": "train", "generation_method": "template_from_smarts", "fgs": []}}
{"id": "eval_00993", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CC(C)C(Sc1nc(-c2ccccc2)c(-c2ccccc2)[nH]1)C(=O)NN", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, imidazole.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "CC(C)C(Sc1nc(-c2ccccc2)c(-c2ccccc2)[nH]1)C(=O)NN", "compound_name": null, "pubchem_cid": null, "MW": 366.49, "LogP": 3.85, "TPSA": 83.8, "QED": 0.27, "functional_groups": ["aromatic ring", "amide", "amine", "imidazole"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "imidazole"]}}
{"id": "eval_00994", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMEQNEFGLIEKIVYLNRVAKVVKGGRRFSFSALVVVGDGKGSVGFGLGKAQEVPEALRKATERAKKSMIEVPLIDGTLPYEILGRFGAGHVLLKPASKGTGIIAGGAVRAVMEAVGVTDVLTKAIGTNNPHNVLRATVAGLASLRSAEEVGQLRGKKLEAPRK\n\n### Sequence 2:\nMFRTEDSILPPLEWLSDNNSSVPPAIYDWLMELGSMTLRFERYCTRVHVEPRHEYFIARNELKEEAEHLPESSLYWLREVVLMGDNQPWLLGRTVIPQETLAHHSNALMNLGTLPLGRYLFSNGELTRDYIHIGRKDSLWARRSRLRLSGKPLLLTELFLADSPLYTEDPS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_00995", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nVSQNDIIKALASPLINDGMVVSDFADHVITREQNAPTGLPVEPVGVAIPHTDSKYVRQNAISVGILAEPVNFEDAGGEPDPVPVRVVFMLALGNWFDITNVLWWIMDVIQDADFMQQLLVMNDDEIYQSIYTRISEAAGMAGIHFRRHYVRHLPLEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHHHHHHHHHHTTSBCTTHHHHHHHHHHHSCCEECCSSSEEECCCCCGGGBSSCEEEEEEEEEEEEECBTTCCSSCEEEEEEEEEECSSHHHHHHHHHHHHHHTTCHHHHHHHTTSCHHHHHHHHHHHHHTCGGGTTCBCCGGGEEEECCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_00996", "category": "Mutation", "instruction": "Describe the effect of mutation Q8X0Z0-L170P on protein Phytoene desaturase.", "input": "### Wild-type:\nMSDIKKSVIVIGAGVGGVSTAARLAKAGFKVTILEKNDFTGGRCSLIHNDGHRFDQGPSLLLLPRFFHEIFQDLGTSLTAEGVELLKCEPNYNIWFGDGSSFEMSTDLTKMKKAIEAVEGIDGFERYLGFLQESHRHYEVSVESVLRRNFPSILSLARPEVLFNLFNIHLLESIWTRASKYFWTERLRRVFTFGSMYMGMSPFDAPGTYSLLQYTELAEGILYPRGGFHKVVEALVNVGQRLGVEYRLSTGVKSISIDQATGKANGVVLSDGTHLPSDIVISNADLVYTYNNLLPKTSYA\n\n### Mutant:\nMSDIKKSVIVIGAGVGGVSTAARLAKAGFKVTILEKNDFTGGRCSLIHNDGHRFDQGPSLLLLPRFFHEIFQDLGTSLTAEGVELLKCEPNYNIWFGDGSSFEMSTDLTKMKKAIEAVEGIDGFERYLGFLQESHRHYEVSVESVLRRNFPSILSLARPEVLFNLFNIHPLESIWTRASKYFWTERLRRVFTFGSMYMGMSPFDAPGTYSLLQYTELAEGILYPRGGFHKVVEALVNVGQRLGVEYRLSTGVKSISIDQATGKANGVVLSDGTHLPSDIVISNADLVYTYNNLLPKTSYA\n\n### Answer:", "output": "Enhances the ability to catalyze the fifth desaturation step.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00997", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CNC1=Nc2ncccc2C(c2cccs2)=NC1c1cccs1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains amine, pyridine. Its descriptor profile: MW=338.46, LogP=4.05, TPSA=49.64, HBD=1, HBA=6, QED=0.77.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "CNC1=Nc2ncccc2C(c2cccs2)=NC1c1cccs1", "compound_name": null, "pubchem_cid": null, "MW": 338.46, "LogP": 4.05, "TPSA": 49.64, "QED": 0.77, "functional_groups": ["amine", "pyridine"], "moleculenet_labels": {"label": 2.63}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_00998", "category": "Mutation", "instruction": "Describe the effect of mutation Q54QG0-N654D on protein Bifunctional glycosyltransferase pgtA.", "input": "### Wild-type:\nMNDSPIISVVLPFLIKDNDDKSLNYQGINNLIISIDSIIEQTFKEWELILVDDGSNNEILEQLLSKRYSTDNRIKFIINKENKGIVKSLNDAILNHCSPTSKYIARMDSDDISHPTRLQSQLKYLQSNETIDILGCPIKMFNNNKLIEILNNNNNNNNINNNVKELINIINNEESFKFIQHPDKDILMWSMFFNCCIVHPSVIFKRSIFTIEHCYEENNQFPFIEDYLFWLKSLIMKGLNISNIQSSTPLLYLRKHNNSISFKNIEKQKDSTANASCYYLNILFKRFNIDSEIIQNSSLS\n\n### Mutant:\nMNDSPIISVVLPFLIKDNDDKSLNYQGINNLIISIDSIIEQTFKEWELILVDDGSNNEILEQLLSKRYSTDNRIKFIINKENKGIVKSLNDAILNHCSPTSKYIARMDSDDISHPTRLQSQLKYLQSNETIDILGCPIKMFNNNKLIEILNNNNNNNNINNNVKELINIINNEESFKFIQHPDKDILMWSMFFNCCIVHPSVIFKRSIFTIEHCYEENNQFPFIEDYLFWLKSLIMKGLNISNIQSSTPLLYLRKHNNSISFKNIEKQKDSTANASCYYLNILFKRFNIDSEIIQNSSLS\n\n### Answer:", "output": "Decrease of Fuc-transferase activity and 80% increase in Gal-transferase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_00999", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nYNO6_YEAST\n\n### Answer:", "output": "Endoplasmic reticulum membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01000", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAVPKKKTSPSRRGMRRSHQALTGEAYTECSNCGELKRPHHVCGHCGHYDGREVAAAGNSGRGLKGVVRV\n\n### Sequence 2:\nMDNREKSRLLWKCRRGMLELDLVLQKFIANEIDRLTENQLKAFDNLLTHNDPNLYAWLMGHEEPEKELLEIVSFIRNCD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01001", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMFLTRSEYDRGVSTFSPEGRLFQVEYSLEAIKLGSTAIGIATKEGVVLGVEKRATSPLLESDSIEKIVEIDRHIGCAMSGLTADARSMIEHARTAAVTHNLYYDEDINVESLTQSVCDLALRFGEGASGEERLMSRPFGVALLIAGHDADDGYQLFHAEPSGTFYRYNAKAIGSGSEGAQAELLNEWHSSLTLKEAELLVLKILKQVMEEKLDENNAQLSCITKQDGFKIYDNEKTAELIKELKEKEAAESPEEADVEMS\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCTTCBCTTSCBHHHHHHHHHHHTSCCEEEEEETTEEEEEEECCCSSSSBCGGGCCCEEEEETTEEEEEEESGGGHHHHHHHHHHHHHHHHHHHSSCCCHHHHHHHHHHHHTCCSSSCTTSCCCCSSCCCEEEEEEEEETTTEEEEEEECTTSCEEEBSEEEESTTHHHHHHHHHHHCCTTCCHHHHHHHHHHHHHHHCSSCCCTTTEEEEEEETTTEEEECCHHHHHHHHHHHHHHHHTCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01002", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMISRGSKVRILRKESYWFNQVGTIATIDQSGIRYPAVVRFENVNYSGTNTNNFALEELIEVSEVETKSKEA\n\n### Sequence 2:\nMIPLVKTFLVVSSGKVSSRKLKRSWGGAKRLNVPVPVLSGLNSPSSMILLHRSRYCSSSLFFPSIVILYRSVMKLSP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01003", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nHBB HBA1 GATA1 KLF1\n</cell>", "output": "This cell is most consistent with erythroid progenitor, an erythroid progenitor cell. Key markers such as GATA1, HBA1, KLF1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "lung", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["HBB", "HBA1", "GATA1", "KLF1"]}}
{"id": "eval_01004", "category": "Mutation", "instruction": "Describe the effect of mutation Q02763-C233Y on protein Angiopoietin-1 receptor.", "input": "### Wild-type:\nMDSLASLVLCGVSLLLSGTVEGAMDLILINSLPLVSDAETSLTCIASGWRPHEPITIGRDFEALMNQHQDPLEVTQDVTREWAKKVVWKREKASKINGAYFCEGRVRGEAIRIRTMKMRQQASFLPATLTMTVDKGDNVNISFKKVLIKEEDAVIYKNGSFIHSVPRHEVPDILEVHLPHAQPQDAGVYSARYIGGNLFTSAFTRLIVRRCEAQKWGPECNHLCTACMNNGVCHEDTGECICPPGFMGRTCEKACELHTFGRTCKERCSGQEGCKSYVFCLPDPYGCSCATGWKGLQCNE\n\n### Mutant:\nMDSLASLVLCGVSLLLSGTVEGAMDLILINSLPLVSDAETSLTCIASGWRPHEPITIGRDFEALMNQHQDPLEVTQDVTREWAKKVVWKREKASKINGAYFCEGRVRGEAIRIRTMKMRQQASFLPATLTMTVDKGDNVNISFKKVLIKEEDAVIYKNGSFIHSVPRHEVPDILEVHLPHAQPQDAGVYSARYIGGNLFTSAFTRLIVRRCEAQKWGPECNHLCTACMNNGVYHEDTGECICPPGFMGRTCEKACELHTFGRTCKERCSGQEGCKSYVFCLPDPYGCSCATGWKGLQCNE\n\n### Answer:", "output": "In GLC3E; enhanced proteasomal degradation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01005", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMGKNDLVKTLRMNYLFDFYQTLLTEKQRNYMELFYLRDYSLSEIAETFDVSRQAVYDNIRRTGDLVEDYEAKLNLYEKFEQRRKIYDDMKQTLNDSKKLEQYINQLEELE\n\n### Sequence 2:\nMGQNDLVKTLRMNYLFDFYQSLLTNKQRNYLELFYLEDYSLSEIADTFNVSRQAVYDNIRRTGDLVEDYEKKLELYQKFEQRREIYDEMKQHLSNPEQIQRYIQQLEDLE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01006", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: FC(F)(F)c1nc2c(ccc3[nH]c4c(-c5ccccc5)nccc4c32)[nH]1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "FC(F)(F)c1nc2c(ccc3[nH]c4c(-c5ccccc5)nccc4c32)[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 352.32, "LogP": 5.28, "TPSA": 57.36, "QED": 0.43, "functional_groups": ["aromatic ring", "halide", "pyridine", "indole", "imidazole"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_01007", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVVPNTVTWGPNVAIVMIICNLIAFAIGKQVIQIPDAGPAPGTFLGLGLPALLGVTSLGHAIGVGAILGLANIGVL\n\n### Answer:", "output": "<SEQ_3Di>DDLQADDDDPVLVVLLVVLLVVLVVCLLPDPPFLVPPAQSCPPPRGGPVSNVVSNVVSNVSSVVVRSVCSSVVND</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01008", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSYH_STAA9\n\n### Answer:", "output": "Histidine--tRNA ligase, Histidyl-tRNA synthetase, HisRS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01009", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "GFAP, SLC1A3, S100B", "output": "This cell is most consistent with astrocyte, an astrocyte. Key markers such as S100B, SLC1A3, GFAP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "astrocyte", "tissue": "colon", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["GFAP", "SLC1A3", "S100B"]}}
{"id": "eval_01010", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: CCCc1c(OCc2ccc(-c3nn[nH]n3)cc2)ccc(C(C)=O)c1O\nDescriptors: MW=352.39, LogP=3.31, TPSA=100.99, HBD=2, HBA=6", "output": "Descriptors: MW=352.39, LogP=3.31, TPSA=100.99, HBD=2, HBA=6, QED=0.63, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "CCCc1c(OCc2ccc(-c3nn[nH]n3)cc2)ccc(C(C)=O)c1O", "compound_name": null, "pubchem_cid": null, "MW": 352.39, "LogP": 3.31, "TPSA": 100.99, "QED": 0.63, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "ketone", "phenol"], "moleculenet_labels": {"label": 3.1}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01011", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMPRRAPAFPLSDIKAQMLFANNIKAQQASKRSFKEGAIETYEGLLSVDPRFLSFKNELSRYLTDHFPANVDEYGRVYGNGVRTNFFGMRHMNGFPMIPATWPLASNLKKRADADLADGPVSERDNLLFRAAVRLMFSDLEPVPLKIRKGSSTCIPYFSNDMGTKIEIAERALEKAEEAGNLMLQGKFDDAYQLHQMGGAYYVVYRAQSTDAITLDPKTGKFVSKDRMVADFEYAVTGGEQGSLFAASKDASRLKEQYGIDVPDGFFCERRRTAMGGPFALNAPIMAVAQPVRNKIYSKYAYTFHHTTRLNKEEKVKEWSLCVATDVSDHDTFWPGWLRDLICDELLNMGYAPWWVKLFETSLKLPVYVGAPAPEQGHTLLGDPSNPDLEVGLSSGQGATDLMGTLLMSITYLVMQLDHTAPHLNSRIKDMPSACRFLDSYWQGHEEIRQISKSDDAMLGWTKGRALVGGHRLFEMLKEGKVNPSPYMKISYEHGGAFLGDILLYDSRREPGSAIFVGNINSMLNNQFSPEYGVQSGVRDRSKRKRPFPGLAWASMKDTYGACPIYSDVLEAIERCWWNAFGESYRAYREDMLKRDTLELSRYVASMARQAGLAELTPIDLEVLADPNKLQYKWTQADVSANIHEVLMHGVSVEKTERFLRSVMPR\n\n### Answer:", "output": "<SEQ_2D>CCCCCCEEETTSHHHHTTSCSSHHHHHHHHCCCCCCCCEEETTEETTCHHHHHHHHHHHHHHHHHSCCCBCTTSCBCTTCCCSCSGGGCEEEEEECBSCCCCCSBCHHHHHHTTCCSSCSSHHHHHHHHHHHHHHHSSCCCCCEEEETTCCCCTTTCCCCHHHHHHHHHHHHHHHHHHHHHHHTTCHHHHHHHHCCCSCEEEEEEEECCCCEEECTTTCCEEECCCEEECHHHHHTTTSSSCEEECBCBCTHHHHHHCCCCCTTCEEEEEEEEEECCHHHHHHHHHHHHHHHHHHHHTTHHHHCCCSHHHHHHHHTTCSEEEEEEESSHHHHCCHHHHHHHHHHHHHHTCCHHHHHHHHHHTTCCEEECCSSTTCCCEEESCTTSCCCBCSCCTTSTTHHHHHHHHHHHHHHHHHHHHTCGGGGGGCSSHHHHHHHHHHHHTTCSSEEEEEETTEEEEEECSSTTHHHHHHHHHHHHHTSSCSCSSSCEEECSSCEETTEEEECCTTCCGGGCEEEECHHHHHHHHHSCSSCCCTTCSCGGGSSSSCGGGHHHHHHHHHTTSTTHHHHHHHHHHHHHHHHSSCHHHHHHHHHHHHHHHHTTTCSSSCTTCCSTTCCHHHHHHHHCGGGGTTTCCGGGSCHHHHHHHEEEECHHHHHHHHHHHSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01012", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMHHHHHHSSGVDLGTQDNLYFQSMSLRLERDGAVARLLIDRADRRNAFSLDMWQRLPELLAEASGDDALRVLVVKSANGGAFCAGADIAELLANKDDAAFHAANQQAINRAQYELARFRLPTVAMVEGDCIGGGCGIALACDMRIAAPAARFGITPAKLGLVYPLHDVKLLVDLVGPGQARRLMFTGGLIDANEAHRIGLVELLGESEDALVGQLATVSSFSTQAIKSFVRRVLDGQVADDADSLRVFASAFEGADFREGTGAFLEKRPPVF\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCSEEEEEETTEEEEEECCGGGTTCBCHHHHHHHHHHHHHHHTCTTCCEEEEEETTSSCSBCSBCHHHHHHTTTCHHHHHHHHHHHHHHHHHHHTCSSCEEEEECSEEETHHHHHHHHSSEEEECTTCEEECCGGGTTCCCCHHHHHHHHHHHHHHHHHHHHHHCCCEEHHHHHHHTSCSEECSCSHHHHHHHHHSCHHHHHHHHHHHHHHHTTCCSCCHHHHHHHHHGGGSHHHHHHHHHHHTTSCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01013", "category": "Mutation", "instruction": "Describe the effect of mutation Q7TSG3-S284A on protein F-box only protein 5.", "input": "### Wild-type:\nMSRRTCSDLRRPSSCPCRLGARTTVDGCKEESPVLSVTMKCFNCNPDLSELEVVKPEDSGIEASYSPVCLEPSCNDCVRNHERLSFIDSPIVGHDNKENQRVQNTLDSSNETEELEASRLYEDSGYSSFTQSDRDDGILILENFRNSPQARLLPSQSPDQHPNKTLLPVLHFERVVCSTLKKNGKRNPKVDREMLKEVIASGNFRLQNIIGKKMGLEHLDILAELSRRGFVHLLANILTKLSGMDLVNLSKVSRIWKKILENNKGAFQLYSKTMQRVIESSKLSLHATTRGYVVGRAALT\n\n### Mutant:\nMSRRTCSDLRRPSSCPCRLGARTTVDGCKEESPVLSVTMKCFNCNPDLSELEVVKPEDSGIEASYSPVCLEPSCNDCVRNHERLSFIDSPIVGHDNKENQRVQNTLDSSNETEELEASRLYEDSGYSSFTQSDRDDGILILENFRNSPQARLLPSQSPDQHPNKTLLPVLHFERVVCSTLKKNGKRNPKVDREMLKEVIASGNFRLQNIIGKKMGLEHLDILAELSRRGFVHLLANILTKLSGMDLVNLSKVSRIWKKILENNKGAFQLYSKTMQRVIESSKLALHATTRGYVVGRAALT\n\n### Answer:", "output": "Decreases phosphorylation by RPS6KA2. p90Rsk2 associates with and phosphorylates Emi1 upstream of the binding region for Cdc20, thus stabilizing their interaction.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01014", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nTPSAB1 TPSB2 CPA3 KIT\n</cell>", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as TPSAB1, KIT, TPSB2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "breast", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TPSAB1", "TPSB2", "CPA3", "KIT"]}}
{"id": "eval_01015", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKRTYQPSKIRRKRSLGFRARMATAAGREIIRRRRAKGRKKLAA\n\n### Sequence 2:\nMKRTYQPSKLVRKRRHGFRARMATAGGRKVIARRRAHGRKRLSA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01016", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY1421_MYCTU\n\n### Answer:", "output": "Nucleotide-binding protein Rv1421.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01017", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nTTTLAFKFRHGVIVAADSRATAGAYIASQTVKKVIEINPYLLGTMAGGAADCSFWERLLARQCRIYELRNKERISVAAASKLLANMVYQYKGMGLSMGTMICGWDKRGPGLYYVDSEGNRISGATFSVGSGSVYAYGVMDRGYSYDLEVEQAYDLARRAIYQATYRDAYSGGAVNLYHVREDGWIRVSSDNVADLHEKYSGSTP\n\n### Answer:", "output": "<SEQ_2D>CCEEEEEETTEEEEEECCCEEETTEEEESCCCCEEEEETTEEEECCBCHHHHHHHHHHHHHHHHHHHHHHTSCCCHHHHHHHHHHHHHTTTTSCCCBEEEEEEECSSSEEEEEEESSSCEEECSEEEESTTHHHHHHHHHTTCCTTCCHHHHHHHHHHHHHHHHHHBTTCCSEEEEEEECSSCEEEEEEEEHHHHHHHHTCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01018", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SYT0-H40A on protein Annexin D1.", "input": "### Wild-type:\nMATLKVSDSVPAPSDDAEQLRTAFEGWGTNEDLIISILAHRSAEQRKVIRQAYHETYGEDLLKTLDKELSNDFERAILLWTLEPGERDALLANEATKRWTSSNQVLMEVACTRTSTQLLHARQAYHARYKKSLEEDVAHHTTGDFRKLLVSLVTSYRYEGDEVNMTLAKQEAKLVHEKIKDKHYNDEDVIRILSTRSKAQINATFNRYQDDHGEEILKSLEEGDDDDKFLALLRSTIQCLTRPELYFVDVLRSAINKTGTDEGALTRIVTTRAEIDLKVIGEEYQRRNSIPLEKAITKDT\n\n### Mutant:\nMATLKVSDSVPAPSDDAEQLRTAFEGWGTNEDLIISILAARSAEQRKVIRQAYHETYGEDLLKTLDKELSNDFERAILLWTLEPGERDALLANEATKRWTSSNQVLMEVACTRTSTQLLHARQAYHARYKKSLEEDVAHHTTGDFRKLLVSLVTSYRYEGDEVNMTLAKQEAKLVHEKIKDKHYNDEDVIRILSTRSKAQINATFNRYQDDHGEEILKSLEEGDDDDKFLALLRSTIQCLTRPELYFVDVLRSAINKTGTDEGALTRIVTTRAEIDLKVIGEEYQRRNSIPLEKAITKDT\n\n### Answer:", "output": "Loss of peroxidase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01019", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nYCF3_NICTO\n\n### Answer:", "output": "Essential for the assembly of the photosystem I (PSI) complex.  May act as a chaperone-like factor to guide the assembly of the PSI subunits.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01020", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: Cc1cc(NCCO)nc2c1ccc1ccccc12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "Cc1cc(NCCO)nc2c1ccc1ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 252.32, "LogP": 3.1, "TPSA": 45.15, "QED": 0.7, "functional_groups": ["aromatic ring", "hydroxyl", "pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01021", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPAPA5_MYCUA\n\n### Answer:", "output": "Belongs to the acyltransferase PapA5 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01022", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nFADR_SHEB5\n\n### Answer:", "output": "Multifunctional regulator of fatty acid metabolism.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01023", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMQIILLEKVANLGNLGDIVKVKDGYARNFLIPNRKARRATKDAIAEFEVRRAELEKVAAEKLAAAQAVGEKLNGQTFEITQKSGVDGRLFGSVTNGDVAELLKKAGYEIEKAQVRMPEGPLKMIGEHGVQVALHTDVVVDVTVNVIGDHA\n\n### Sequence 2:\nMQIILLEKVANLGNLGDIVKVKDGYARNFLIPNRKARRATKDAIAEFEVRRAELEKVAAEKLAAAQAVGEKLNGQTFEITQKSGVDGRLFGSVTNGDVAELLKKAGYEIEKAQVRMPEGPLKMIGEHGVQVALHTDVLVDVTVNVIGDHA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01024", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAVAMSMDANCPKNIFLLCRNTGIGFDDLRLHCIFCTKQLTTTELQAFALRELNVVWRRGAPYGACARCLLVEGIARRLKYWEYSYYVSGVEEETKQSIDTQQIRCYMCHKPLVKEEKDRHRNEKRRLHKISGHWRGSCQYCWSRCTVRIPR\n\n### Sequence 2:\nMDDQRPKNIFLLCRDSGISFDDLRLHCIFCAKVLTTAELSAFALRELNVVWRTGAPYGACARCLLLQGIVRRLKHWDYSLYVEGVEEETKQSIDTQQVRCYMCHKPLVKEEKDRHRNERRRLHCIAGYWRGSCQYCWLRCTVRIPQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01025", "category": "Mutation", "instruction": "Describe the effect of mutation Q2FV59-W38C on protein 4,4'-diapophytoene synthase.", "input": "### Wild-type:\nMTMMDMNFKYCHKIMKKHSKSFSYAFDLLPEDQRKAVWAIYAVCRKIDDSIDVYGDIQFLNQIKEDIQSIEKYPYEHHHFQSDRRIMMALQHVAQHKNIAFQSFYNLIDTVYKDQHFTMFETDAELFGYCYGVAGTVGEVLTPILSDHETHQTYDVARRLGESLQLINILRDVGEDFDNERIYFSKQRLKQYEVDIAEVYQNGVNNHYIDLWEYYAAIAEKDFQDVMDQIKVFSIEAQPIIELAARIYIEILDEVRQANYTLHERVFVDKRKKAKLFHEINSKYHRI\n\n### Mutant:\nMTMMDMNFKYCHKIMKKHSKSFSYAFDLLPEDQRKAVCAIYAVCRKIDDSIDVYGDIQFLNQIKEDIQSIEKYPYEHHHFQSDRRIMMALQHVAQHKNIAFQSFYNLIDTVYKDQHFTMFETDAELFGYCYGVAGTVGEVLTPILSDHETHQTYDVARRLGESLQLINILRDVGEDFDNERIYFSKQRLKQYEVDIAEVYQNGVNNHYIDLWEYYAAIAEKDFQDVMDQIKVFSIEAQPIIELAARIYIEILDEVRQANYTLHERVFVDKRKKAKLFHEINSKYHRI\n\n### Answer:", "output": "Decrease in C(30) carotene synthase activity. C(40) carotene synthase activity acquired. confer the same phenotype but are present in the random library at much lower frequencies. The mutation in the CrtM protein has been found to have functional importance. Further mutagenesis at this amino acid residue has led to the discovery of synthase specificity control at the second step of the two-step reaction.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01026", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "TTR, GCG, IRX2", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as TTR, GCG, IRX2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "ovary", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TTR", "GCG", "IRX2"]}}
{"id": "eval_01027", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPQITLWQRPLVTIKIGGQLKEALLDTGADDTVLEENSLPGRWKPKMIGGIGGFIKVRQYDQILIEICGHKAIGTVLVGPTPVNIIGRNLLTQIGCTLNF\n\n### Answer:", "output": "<SEQ_3Di>DDDDVPDFDKWWKDFPHDIDIAGEDAVAQAWEDEDDDDDDQWDWDWDADVVGIWTWTKDFQTWMQTNNRIDTTIHTYTDDPGTYDYPVTCVVVPHDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01028", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMPLYVIDKPITLHILTQLRDKYTDQINFRKNLVRLGRILGYEISNTLDYEIVEVETPLGVKTKGVDITDLNNIVIINILRAAVPLVEGLLKAFPKARQGVIGASRVEVDGKEVPKDMDVYIYYKKIPDIRAKVDNVIIADPMIATASTMLKVLEEVVKANPKRIYIVSIISSEYGVNKILSKYPFIYLFTVAIDPELNNKGYILPGLGDAGDRAFG\n\n### Answer:", "output": "<SEQ_2D>CCEEECCCHHHHHHHHHHHCTTCCHHHHHHHHHHHHHHHHHHHHTTSCCEEEEEECTTSCEEEEEECGGGGSEEEEEEETTTHHHHHHHHHHCTTCEEEEEEEEECCCCCSSCCSCCCEEEEEEECCCCCTTTCEEEEECSEESSSHHHHHHHHHHGGGCCSEEEEECSEEEHHHHHHHHHHCTTSEEEESEEESEECTTSCEESSCSCHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01029", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nATPE_DESAH\n\n### Answer:", "output": "Belongs to the ATPase epsilon chain family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01030", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRLP5_ARATH\n\n### Answer:", "output": "Belongs to the RLP family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01031", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: CCOC(=O)N1CC(C)=C(C)CN1C(=O)Nc1ccc(Cl)c(Cl)c1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "CCOC(=O)N1CC(C)=C(C)CN1C(=O)Nc1ccc(Cl)c(Cl)c1", "compound_name": null, "pubchem_cid": null, "MW": 372.25, "LogP": 4.55, "TPSA": 61.88, "QED": 0.78, "functional_groups": ["aromatic ring", "ester", "amide", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01032", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGKNKDVRITITLECTNCPQNKLKRYPGISRYTTQKNRRNTPNRLELKKFCPHCSLHTLHKEIKK\n\n### Sequence 2:\nMAKNKEIRINVTLECTNCANNSQKSSRGISRYITQKNRRNTPNRLELNKFCPYCVGHTVHKEIKK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01033", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRECA_MYCMY\n\n### Answer:", "output": "Can catalyze the hydrolysis of ATP in the presence of single-stranded DNA, the ATP-dependent uptake of single-stranded DNA by duplex DNA, and the ATP-dependent hybridization of homologous single-stranded DNAs.  It interacts with LexA causing its activation and leading to its autocatalytic cleavage.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01034", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMARYTGPSCRLCRRENTELFLKGERCYTDKCAIKRRNYPPGQHGQGRSKTSDYGVQLREKQKVRRIYGILENQFRGYFERADRLKGVTGENLLFLLERRLDNIVYRLGFASSRIEARQLVRHGHFTLNGKKVTIPSIQVKTGDTVELREKSRKVASINESLEAVVRRGIPQWLELDKGAFKGSVKTLPVREDITMPIQEQLIVELYSK\n\n### Sequence 2:\nMKLVLIRHGESEWNRLNLFTGWTDVPLTPRGESEAQEGGRVLQEAGFDFDLCYTSFLKRAIRTLNFVLQALDREWLPVHKSWKLNERHYGDLQGLNKTETAQKYGEQQVRVWRRSFDVAPPPLTVGDARCPHTQASYRGVCASGRTPVLPFTESLKDTVARVVPYFEEEIKPQMISGQRVLIVAHGNSLRALMKHIESLDETQIMEVNLPTGVPLVYEFEADFTLCGKRFLGNEADVAARAQAVADQGKSN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01035", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "O=Cc1ccc(F)cc1", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring, aldehyde, halide. Its descriptor profile: MW=124.11, LogP=1.64, TPSA=17.07, HBD=0, HBA=1, QED=0.52.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=Cc1ccc(F)cc1", "compound_name": null, "pubchem_cid": null, "MW": 124.11, "LogP": 1.64, "TPSA": 17.07, "QED": 0.52, "functional_groups": ["aromatic ring", "aldehyde", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01036", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: Cc1cccc(C)c1N(CCCc1cccnc1)C(=O)C(C)N\nDescriptors: MW=311.43, LogP=3.01, TPSA=59.22, HBD=1, HBA=3", "output": "Descriptors: MW=311.43, LogP=3.01, TPSA=59.22, HBD=1, HBA=3, QED=0.89, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "Cc1cccc(C)c1N(CCCc1cccnc1)C(=O)C(C)N", "compound_name": null, "pubchem_cid": null, "MW": 311.43, "LogP": 3.01, "TPSA": 59.22, "QED": 0.89, "functional_groups": ["aromatic ring", "amide", "amine", "pyridine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 1.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01037", "category": "Mutation", "instruction": "Describe the effect of mutation P0DPR0-A176W on protein Main hemagglutinin component type C.", "input": "### Wild-type:\nMSQTNANDLRNNEVFFISPSNNTNKVLDKISQSEVKLWNKLSGANQKWRLIYDTNKQAYKIKVMDNTSLILTWNAPLSSVSVKTDTNGDNQYWYLLQNYISRNVIIRNYMNPNLVLQYNIDDTLMVSTQTSSSNQFFKFSNCIYEALNNRNCKLQTQLNSDRFLSKNLNSQIIVLAQWFDSSRQKWIIEYNETKSAYTLKCQENNRYLTWIQNSNNYVETYQSTDSLIQYWNINYLDNDASKYILYNLQDTNRVLDVYNSQIANGTHVIVDSYHGNTNQQWIINLI\n\n### Mutant:\nMSQTNANDLRNNEVFFISPSNNTNKVLDKISQSEVKLWNKLSGANQKWRLIYDTNKQAYKIKVMDNTSLILTWNAPLSSVSVKTDTNGDNQYWYLLQNYISRNVIIRNYMNPNLVLQYNIDDTLMVSTQTSSSNQFFKFSNCIYEALNNRNCKLQTQLNSDRFLSKNLNSQIIVLWQWFDSSRQKWIIEYNETKSAYTLKCQENNRYLTWIQNSNNYVETYQSTDSLIQYWNINYLDNDASKYILYNLQDTNRVLDVYNSQIANGTHVIVDSYHGNTNQQWIINLI\n\n### Answer:", "output": "Significantly increased binding to bovine mucin, increased binding to porcine mucin. Binding of N-acetyl-beta-neuraminic acid, significantly increased binding to N-acetyl-D-galactosamine. Significantly increased binding to both bovine and porcine mucin; when associated with A-278-279-A. The double mutant at site I has avidity for N-acetylneuraminic acid but no avidity for galactose. The mutant (site I knockout) had a K(d) value of 31.5mM for N-acetylneuraminic acid (Neu5Ac) and 61.3mM for galactose (Gal), while the K(d) value for N-acetylgalactosamine (GalNAc) was too high to be determined.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01038", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nKLPEAYAIFDPLVDVLPVIPVLFLALAFVWQAAVGFR\n\n### Answer:", "output": "<SEQ_3Di>DDDPVCPVCPVVVVPPVCVVVVVVVVVVVVVVVVPVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01039", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "NKG7, FCGR3A, KLRD1, GNLY, GZMB", "output": "The marker profile suggests functional enrichment consistent with the typical biology of natural killer cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "natural killer cell", "tissue": "intestine", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["NKG7", "FCGR3A", "KLRD1", "GNLY", "GZMB"]}}
{"id": "eval_01040", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nQUEF_ECOSE\n\n### Answer:", "output": "NADPH-dependent 7-cyano-7-deazaguanine reductase, 7-cyano-7-carbaguanine reductase, NADPH-dependent nitrile oxidoreductase, PreQ(0) reductase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01041", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nEFMT1_HUMAN\n\n### Answer:", "output": "EEF1A lysine methyltransferase 1, N(6)-adenine-specific DNA methyltransferase 2, Protein-lysine N-methyltransferase N6AMT2, eEF1A-KMT.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01042", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAL1L1_XENLA\n\n### Answer:", "output": "Cytosolic 10-formyltetrahydrofolate dehydrogenase, 10-FTHFDH, FDH, Aldehyde dehydrogenase family 1 member L1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01043", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nMOAA_LISMF\n\n### Answer:", "output": "Belongs to the radical SAM superfamily. MoaA family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01044", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y6R1-G721T on protein Electrogenic sodium bicarbonate cotransporter 1.", "input": "### Wild-type:\nMEDEAVLDRGASFLKHVCDEEEVEGHHTIYIGVHVPKSYRRRRRHKRKTGHKEKKEKERISENYSDKSDIENADESSSSILKPLISPAAERIRFILGEEDDSPAPPQLFTELDELLAVDGQEMEWKETARWIKFEEKVEQGGERWSKPHVATLSLHSLFELRTCMEKGSIMLDREASSLPQLVEMIVDHQIETGLLKPELKDKVTYTLLRKHRHQTKKSNLRSLADIGKTVSSASRMFTNPDNGSPAMTHRNLTSSSLNDISDKPEKDQLKNKFMKKLPRDAEASNVLVGEVDFLDTPFI\n\n### Mutant:\nMEDEAVLDRGASFLKHVCDEEEVEGHHTIYIGVHVPKSYRRRRRHKRKTGHKEKKEKERISENYSDKSDIENADESSSSILKPLISPAAERIRFILGEEDDSPAPPQLFTELDELLAVDGQEMEWKETARWIKFEEKVEQGGERWSKPHVATLSLHSLFELRTCMEKGSIMLDREASSLPQLVEMIVDHQIETGLLKPELKDKVTYTLLRKHRHQTKKSNLRSLADIGKTVSSASRMFTNPDNGSPAMTHRNLTSSSLNDISDKPEKDQLKNKFMKKLPRDAEASNVLVGEVDFLDTPFI\n\n### Answer:", "output": "Moderate increase of the sodium-dependent ion transport activity. Mutation of T721 to G resulted in a gain of Na(+)-dependent base transport mediated by NBC1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01045", "category": "Mutation", "instruction": "Describe the effect of mutation B8N2C8-N137K on protein Sterol 14-alpha demethylase.", "input": "### Wild-type:\nMIFSRSMASFTLVSAYAAAGLLAIIVLNLLRQLLFRNKTDPPLVFHWIPFLGSTVTYGMDPYAFFFSCRQKYGDIFTFILLGRKITVYLGIQGNEFILNGKLKDVNAEEIYSPLTTPVFGSDIVYDCPNSKLMEQKNFIKFGLTQAALESHVPLIEKEVLDYLKTSPNFKGTSGRVEITDAMAEITIFTAGRALQGEEVRKKLTAEFADLYHDLDRGFTPINFMLPWAPLPRNRKRDAAHARMREIYMDIINERRKNPDRETSDMIWNLMHCTYKNGQPLPDKEIAHMMITLLMAGQHSS\n\n### Mutant:\nMIFSRSMASFTLVSAYAAAGLLAIIVLNLLRQLLFRNKTDPPLVFHWIPFLGSTVTYGMDPYAFFFSCRQKYGDIFTFILLGRKITVYLGIQGNEFILNGKLKDVNAEEIYSPLTTPVFGSDIVYDCPNSKLMEQKKFIKFGLTQAALESHVPLIEKEVLDYLKTSPNFKGTSGRVEITDAMAEITIFTAGRALQGEEVRKKLTAEFADLYHDLDRGFTPINFMLPWAPLPRNRKRDAAHARMREIYMDIINERRKNPDRETSDMIWNLMHCTYKNGQPLPDKEIAHMMITLLMAGQHSS\n\n### Answer:", "output": "Leads to susceptibility to voriconazole.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01046", "category": "Mutation", "instruction": "Describe the effect of mutation O43707-S262P on protein Alpha-actinin-4.", "input": "### Wild-type:\nMVDYHAANQSYQYGPSSAGNGAGGGGSMGDYMAQEDDWDRDLLLDPAWEKQQRKTFTAWCNSHLRKAGTQIENIDEDFRDGLKLMLLLEVISGERLPKPERGKMRVHKINNVNKALDFIASKGVKLVSIGAEEIVDGNAKMTLGMIWTIILRFAIQDISVEETSAKEGLLLWCQRKTAPYKNVNVQNFHISWKDGLAFNALIHRHRPELIEYDKLRKDDPVTNLNNAFEVAEKYLDIPKMLDAEDIVNTARPDEKAIMTYVSSFYHAFSGAQKAETAANRICKVLAVNQENEHLMEDYEK\n\n### Mutant:\nMVDYHAANQSYQYGPSSAGNGAGGGGSMGDYMAQEDDWDRDLLLDPAWEKQQRKTFTAWCNSHLRKAGTQIENIDEDFRDGLKLMLLLEVISGERLPKPERGKMRVHKINNVNKALDFIASKGVKLVSIGAEEIVDGNAKMTLGMIWTIILRFAIQDISVEETSAKEGLLLWCQRKTAPYKNVNVQNFHISWKDGLAFNALIHRHRPELIEYDKLRKDDPVTNLNNAFEVAEKYLDIPKMLDAEDIVNTARPDEKAIMTYVPSFYHAFSGAQKAETAANRICKVLAVNQENEHLMEDYEK\n\n### Answer:", "output": "In FSGS1; no effect on protein abundance; no effect on homodimerization; loss of localization to the nucleus; prevents nuclear localization of the wild-type protein; decreased interaction with PPARG and RARA; loss of transcriptional coactivator activity; dominant negative effect on nuclear receptors-mediated transcription; increased actin-binding affinity. The crystal structure of the ABD of alpha-actinin-4 mutant described here displays the canonical compact conformation. Furthermore, the sedimentation coefficients by analytical ultracentrifugation of wild-type and FSGS mutant ABDs (, , and ) are nearly identical (2. 50+/-0. 03 S) and are in good agreement with the theoretical value calculated from the crystal structure (2. 382 S), implying that the compact conformation is retained in solution. The absence of a structural change suggests that the compact ABD conformation observed in the majority of the structures is highly stable and is preserved in solution, even in FSGS mutant ABDs. These mutations alter the amino acid sequence and contribute to the development of FSGS.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01047", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFEV_DANRE\n\n### Answer:", "output": "Protein FEV, Protein Pet-1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01048", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMHHHHHHSSGVDLGTENLYFQSNANPPPPETSNPNKPKRQTNQLQYLLRVVLKTLWKHQFAWPFQQPVDAVKLNLPDYYKIIKTPMDMGTIKKRLENNYYWNAQECIQDFNTMFTNCYIYNKPGDDIVLMAEALEKLFLQKINELPTEE\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCTTSCCCCCHHHHHHHHTHHHHHHHSTTCGGGSSCCCTTTTTCTTHHHHCSSCCCHHHHHHHHHTTCCSSHHHHHHHHHHHHHHHHHHSCTTCHHHHHHHHHHHHHHHHHTTCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01049", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNAIPLEHGLALASVLFALGLVGLMVRRNILFVLMSLEVMMNAAALAFVVAGSRWGQPDGQVMFILVLSLAAAEASIGLAILLQLYRRFHTLDIDAASEMRG\n\n### Sequence 2:\nMQFSTTPTLEGQTIVEYCGVVTGEAILGANIFRDFFAGIRDIVGGRSGAYEKELRKAREIAFEELGSQARALGADAVVGIDIDYETVGQNGSMLMVSVSGTAVKTRR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01050", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nYQHA_SHIBS\n\n### Answer:", "output": "Cell membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01051", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGLYR1_DROPS\n\n### Answer:", "output": "Nucleosome-destabilizing factor that is recruited to genes during transcriptional activation and colocalizes with a subset of trimethylated 'Lys-36' histone H3 (H3K36me3)-enriched regions.  Binds DNA (in vitro).  Facilitates Pol II transcription through nucleosomes.  Facilitates male-specific lethal (MSL) histone acetyltransferase complex targeting to active genes on the X chromosome.  Stimulates the acetylation of 'Lys-56' of nucleosomal histone H3 (H3K56ac) by nej (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01052", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "Cc1cc(Cl)cc2c1SC(=C1Sc3c(C)cc(Cl)cc3C1=O)C2=O", "output": "The molecule contains the following functional groups: aromatic ring, ketone, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "Cc1cc(Cl)cc2c1SC(=C1Sc3c(C)cc(Cl)cc3C1=O)C2=O", "compound_name": null, "pubchem_cid": null, "MW": 393.32, "LogP": 6.1, "TPSA": 34.14, "QED": 0.51, "functional_groups": ["aromatic ring", "ketone", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "ketone", "halide"]}}
{"id": "eval_01053", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "CCCc1cc(CC2CS(=O)(=O)CC([NH2+]Cc3cccc(C(C)(C)C)c3)C2O)cc(F)c1N", "output": "The SMILES represents an unnamed molecule with MW=477.67, LogP=2.74, TPSA=97.0. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "BACE", "canonical_smiles": "CCCc1cc(CC2CS(=O)(=O)CC([NH2+]Cc3cccc(C(C)(C)C)c3)C2O)cc(F)c1N", "compound_name": null, "pubchem_cid": null, "MW": 477.67, "LogP": 2.74, "TPSA": 97.0, "QED": 0.53, "functional_groups": ["aromatic ring", "hydroxyl", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01054", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMNLALMKMWFALGSMGLMFLAVASIYLSRFKCQNRFLKIAISSFAYMCMLISGIIVFVVVFSGPVNE\n\n### Sequence 2:\nMAKANEHFFYVLKCNDNSYYGGYTTDVTRREAEHNAGIRCKYTKTRRPVKVIHFEKFETRSEATKAEAAFKKLSRKNKDSYLIEREEDSE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01055", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nEFTU_TOBAC\n\n### Answer:", "output": "Elongation factor Tu, chloroplastic, EF-Tu.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01056", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMALTDSKKKATSKRKRDEPEQEQTHTLTSQLEMLTDEEGDDDNAGDVSDDGEYDEFPEIDAESDSEEEDEEEDEEDEEDEPDISDEEDSDSDESLHVFPKAKTVVSDITGQQKRVYPEIEPDYDSDSSTEDSPNRVGNVPMHWYDDLPHIGYDINGKRVLRPAKGDELDKFLATIEDPDSWTSAFDKKMQMDKPLTSEELDIIQRLYAGENPDATYDPYEPMVEWFTGKGKEEIMPLSAAPEPKRRWVRSKWEKQKIMKIVRAIRQGRIVPNKPKTSSKQFEFYSIWSSEPSSSQPPPLPAPKRPLPTNAESYNPPEEYLPTEEEKQEWLKQDPEDRERDFLPQKFSSLRHVPAYDQFIKERFNRQLDLYLAPRIQRVKLNIDPNSLIPKLPSPSSLKPFPNYRSLLFTHPKGRARCVSVSPDGAWAVSGDEDGVVSLWEVNVGCEIRRWKFEGKIGSLEWCPRADACYFAVGIEETIHFLIPPNLDPTVLALTQTLLAPSTLPPAPATPSAVKWSSSSLSSWSVEQPILSLNLPPSSGLPTQISWHKKGDYIATVSSGGAQNGVWIHQFTRRHSQAPFKKIKGAVQQVLFHPIKPHFFVATQQYVRLYNLAEQKLIKTLQPGIRWISSMDVHPSGDHVIVGGYDRKLCWFDLELSEKPYKVLRYHSRAIRSLHFHPTYPLFASSSDDGSIQIFHARVYNDLMTDPLIVPLKILRGHQITDGLGILQVKWTPKHPWLLSAAADGTVAVWCS\n\n### Sequence 2:\nMVELESMSSSSEWFLNNNSNIHNNTHNHNHNNNNNNNNINSNNHGHSANSSAHNHNNNNNNNTNNNSNNSNNNNNNNNNNNNTQNTNNGTFLTPLPVLTNSSTGLGKSIDWLYTNTSNTDSHNINSSSNKVTKLSNGFSLLNQDGTQSPSCFLNLSSTGNHEDPVQVPMNNNPQDLLFQFNISPQLQSQNNNSSNNNNNNNNNGSNTPLNGSNGSNNNYSSPISPVPQYQHSAGSSPIQEYPYYSPSSSPHSNESTSPITVIKDSGSIINPINNNNNNNNNNNNNNNNNNNNNNNNNNKNNLNNNTFQFSSLKSTTGIAPPLQTDYHLLYNFVQQNFSGKPEDSLLNQTLNNNNNNNNNNNNNNNNTTIQTSPQPLVLPQIQNPFLPYNISTPISVPNNILSSAQSSTNSSPNNNNNNNNNNNNIFSTVNNNNNLVNNNSNNMELDVDKPFSQEKFYLHLKSIVPSFNENFEASSESAQSESSQESDYDALNERNSGVKKRGRDEDQIDTSGQVVLSREHVLKLSSKEIEEYVSRLKMHHILTQAEEKELKKQRRLVKNREYASQSRSRRKIYVENIETKLQKTNQDCASIKSQLNSVKEENKALKKQLYSLTNTLKSNPSLAEAFGKIFSPIGNNKTSSATLFVFFFLFTFTFLFQSSTVTFNSDRVSSIQRNLLSLEETQATEWNIKRAILEETKQYVDSLLTSLYTSKLQSLSETPLDTSNYIINNNNNESTSETTTNNKVVSTSSDDISNVLSNLSVSDKEVPQKCKDSSDLKCDSPPLSPLN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01057", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAKKVVGTIELMIPAQQASPSPPVGPALGQHGVNIMEFVKSFNAATAEMKPGTVVPVVITVYADRSFTFILKTPPASYLLKEAAGIKKGASDPKREKVGKVTKEQIREIAEIKMKDLNTEDIEAAMRIIAGTARSMGIEVEGLEA\n\n### Sequence 2:\nMAKKVAAQIKLQLPAGKATPAPPVGPALGQHGVNIMEFCKRFNAETAD-KAGMILPVVITVYEDKSFTFIIKTPPASFLLKKAAGIESGSSEPKRKIVGKVTRKQIEEIAKIKMPDLNANDLEAAMRIIEGTARSMGIEV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01058", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTRMD_YERP3\n\n### Answer:", "output": "Belongs to the RNA methyltransferase TrmD family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01059", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY3974_ARATH\n\n### Answer:", "output": "Putative BTB/POZ domain-containing protein At3g29740.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01060", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGLELFLDLVSQPSRAVYIFAKKNGIPLELRTVDLVKGQHKSKEFLQINSLGKLPTLKDGDFILTESSAILIYLSCKYQTPDHWYPSDLQARARVHEYLGWHADCIRGTFGIPLWVQVLGPLIGVQVPEEKVERNRTAMDQALQWLEDKFLGDRPFLAGQQVTLADLMALEELMQPVALGYELFEGRPRLAAWRGRVEAFLGAELCQEAHSIILSILEQAAKKTLPTPSPEAYQAMLLRIARIP\n\n### Answer:", "output": "<SEQ_2D>CCEEEEECTTSHHHHHHHHHHHHTTCCEEEEECCTTTTGGGSHHHHTTCTTCCSCEEEETTEEECTHHHHHHHHHHHTTCCTTTSCCSHHHHHHHHHHHHHHHHHTTTTSSHHHHHHTHHHHTTCCCCHHHHHHHHHHHHHHHHHHHTTTTTTSSBTTBSSCCHHHHHHHHHTHHHHHHTCCTTTTCHHHHHHHHHHHHHHCHHHHHHHSTHHHHHHHHHHTTCCCCCCHHHHHHHHHHHTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01061", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CN1CCN(CCCN2c3ccccc3Sc3ccc(C(F)(F)F)cc32)CC1", "output": "The SMILES represents an unnamed molecule with MW=407.51, LogP=4.95, TPSA=9.72. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "BBBP", "canonical_smiles": "CN1CCN(CCCN2c3ccccc3Sc3ccc(C(F)(F)F)cc32)CC1", "compound_name": null, "pubchem_cid": null, "MW": 407.51, "LogP": 4.95, "TPSA": 9.72, "QED": 0.71, "functional_groups": ["aromatic ring", "amine", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01062", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSGSTGERSFADIITSIRYWVIHSITIPSLFIAGWLFVSTGLAYDVFGSPRPNEYFTESRQGIPLITGRFDPLEQLDEFSRSF\n\n### Sequence 2:\nMAKKGTRVVVTLECTEARTSTDPKRSNGVSRYTTEKNRRNTTERLELKKFNPHLNRMTIHKEIK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01063", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: OC(c1ccccc1)(c1ccccc1)C1CN2CCC1CC2\nDescriptors: MW=293.41, LogP=3.26, TPSA=23.47, HBD=1, HBA=2", "output": "Descriptors: MW=293.41, LogP=3.26, TPSA=23.47, HBD=1, HBA=2, QED=0.94, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "OC(c1ccccc1)(c1ccccc1)C1CN2CCC1CC2", "compound_name": null, "pubchem_cid": null, "MW": 293.41, "LogP": 3.26, "TPSA": 23.47, "QED": 0.94, "functional_groups": ["aromatic ring", "amine", "hydroxyl"], "moleculenet_labels": {"NR-AR": 1.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01064", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVDREQLVQKARLAEQAERYDDMAAAMKNVTELNEPLSNEERNLLSVAYKNVVGARRSSWRVISSIEQKTSADGNEKKIEMVRAYREKIEKELEAVCQDVLSLLDNYLIKNCSETQYESKVFYLKMKGDYYRYLAEVATGEKRATVVESSEKAYSEAHEISKEHMQPTHPIRLGLALNYSVFYYEIQNAPEQACHLAKTAFDDAIAELDTLNEDSYKDSTLIMQLLRDNLTLWT\n\n### Answer:", "output": "<SEQ_3Di>DDLVVLVVVLVVCVVVVVLVSNLVSLLVVLVVLDAADPVSLVSNVVSLCVVLVVLVVVLVVLVVVCVVPVPPPDPVVVVVSVVVNVVSLVVNVVSLVSLVVSLVPRQCVPDDPPPLVSNLSSLLSQLVSLLSVLVPDDDVSNVVSLVSNVVSLVVNQVSCVPPHQLLDLSNLVSLQVVLVCCCPRVVHLVVSLVSLVVSLVSCVVPVVPDDPPDRVSSVVSSVVSVVVNVVSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01065", "category": "Mutation", "instruction": "Describe the effect of mutation Q96QA5-I245N on protein Gasdermin-A.", "input": "### Wild-type:\nMTMFENVTRALARQLNPRGDLTPLDSLIDFKRFHPFCLVLRKRKSTLFWGARYVRTDYTLLDVLEPGSSPSDPTDTGNFGFKNMLDTRVEGDVDVPKTVKVKGTAGLSQNSTLEVQTLSVAPKALETVQERKLAADHPFLKEMQDQGENLYVVMEVVETVQEVTLERAGKAEACFSLPFFAPLGLQGSINHKEAVTIPKGCVLAFRVRQLMVKGKDEWDIPHICNDNMQTFPPGEKSGEEKVILIQASDVGDVHEGFRTLKEEVQRETQQVEKLSRVGQSSLLSSLSKLLGKKKELQDLE\n\n### Mutant:\nMTMFENVTRALARQLNPRGDLTPLDSLIDFKRFHPFCLVLRKRKSTLFWGARYVRTDYTLLDVLEPGSSPSDPTDTGNFGFKNMLDTRVEGDVDVPKTVKVKGTAGLSQNSTLEVQTLSVAPKALETVQERKLAADHPFLKEMQDQGENLYVVMEVVETVQEVTLERAGKAEACFSLPFFAPLGLQGSINHKEAVTIPKGCVLAFRVRQLMVKGKDEWDIPHICNDNMQTFPPGEKSGEEKVILNQASDVGDVHEGFRTLKEEVQRETQQVEKLSRVGQSSLLSSLSKLLGKKKELQDLE\n\n### Answer:", "output": "Abolished cleavage by S.pyogenes effector protein SpeB, preventing pyroptosis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01066", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nMRC1:low C1QB:low CD163:low\n</cell>", "output": "This cell is most consistent with macrophage, a tissue macrophage. Key markers such as C1QB, MRC1, CD163 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "blood", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MRC1", "C1QB", "CD163"]}}
{"id": "eval_01067", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nTNNI3:high MYH7:medium TNNT2:medium MYH6:high\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as MYH6, MYH7, TNNI3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["TNNI3", "MYH7", "TNNT2", "MYH6"]}}
{"id": "eval_01068", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nIL5RA EPX PRG2 CLC\n</cell>", "output": "This cell is most consistent with eosinophil, an eosinophil. Key markers such as EPX, IL5RA, PRG2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "eosinophil", "tissue": "ovary", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IL5RA", "EPX", "PRG2", "CLC"]}}
{"id": "eval_01069", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMDKLLIAADNALRTLFARPRAAQPSPARGLPEGDLSPAQRREAGALMRVNHVGEVCAQALYMGQAAVTRDPALRARLMEAAREETDHLAWTAERLQALGSRPSLLNPLWFAGAFAIGWTAAQVSDAASLGFVVETENQVARHLQGHLERMPPQDAASLAVIERMQADEQRHADDARAAGASDLPAPARVLMAAAARVMTATAHHI\n\n### Sequence 2:\nMDKILTAADAALRTLFAQPSAAERSPAAGIAEGELTEAQRRLSGALMRVNHVGEVCAQALYNAQAMVTRDMVLREHLLEAAREETDHLAWTRDRLNALGDRPSLLNPLWFAGAFAIGVVAAKISDAASLGFVVETEEQVSAHLESHLGLLPEEDRASLAVVARMKDDEERHAAAAREAGALQLPPPARALMRLAAKVMTTTAHRI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01070", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nMYH11 CNN1 ACTA2 TAGLN\n</cell>", "output": "This cell is most consistent with smooth muscle cell, a smooth muscle cell. Key markers such as CNN1, MYH11, TAGLN support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "smooth muscle cell", "tissue": "bone marrow", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MYH11", "CNN1", "ACTA2", "TAGLN"]}}
{"id": "eval_01071", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nASLYEKLGGAAAVDLAVEKFYGKVLADERVNRFFVNTDMAKQKQHQKDFMTYAFGGTDRFPGRSMRAAHQDLVENAGLTDVHFDAIAENLVLTLQELNVSQDLIDEVVTIVGSVQHRNDVLNR\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHTSHHHHHHHHHHHHHHHHHCTTTGGGGTTSCHHHHHHHHHHHHHHHTTCGGGSCHHHHHHHHHHHHHHHCCCHHHHHHHHHHHHHHHHHTTCCHHHHHHHHHHHTCHHHHHHHTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01072", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLKMGVVLFVFLVLFPLATLQLDADQPVERYAENKQLVSPYERRQIILHALGQRQCCDWQWCDGACDCCA\n\n### Sequence 2:\nMAQKKGGGSTRNGRDSKPKMLGVKAFGGELISAGSIIVRQRGTKFHPGTNVGVGKDHTLFALVDGHVSFGTKGAMSKHTVSVTPA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01073", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVTMRDLLECGVHFGHQTRRWNPKMKKFIFGERKGIYIIDLQKTIRYFRYTYNIVRDAAAEGKTVLFVGTKKQAGVAIKEYAEKCGMPYVNHRWLGGMMTNFGTIKQSIRKLEVIEAMEEDGSINLLTKKEALMLRRKKEKLLATLGGIRNMKSLPDMVFVIDTVKEKIAVAEANKLRMPVVAPIDTNCDPDIVDFPIPGNDDAIRSVQLFCQEMAEAINEGKALRDQDEAEQVEPVSQEEKDEVVAEAMSEADFEEQ\n\n### Sequence 2:\nMRLVQLSRHSIAFPSPEGALREPNGLLALGGDLSPARLLMAYQHGIFPWFSPGDPILWWSPDPRAVLWPEKFHLSRSMKRFHNASPYRVTLNYAFDRVIDGCANHRDEGTWITRGIEEAYRRLHELGHAHSIEVWRDQELVGGMYGVSQGALFCGESMFSRQENASKTALLVFCAEFIRHGGKLIDCQVLNSHTASLGAIEIPRRDYLDHLAGLRQQPLASRFWVPRTLFLPRK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01074", "category": "Mutation", "instruction": "Describe the effect of mutation P02701-T58I on protein Avidin.", "input": "### Wild-type:\nMVHATSPLLLLLLLSLALVAPGLSARKCSLTGKWTNDLGSNMTIGAVNSRGEFTGTYTTAVTATSNEIKESPLHGTQNTINKRTQPTFGFTVNWKFSESTTVFTGQCFIDRNGKEVLKTMWLLRSSVNDIGDDWKATRVGINIFTRLRTQKE\n\n### Mutant:\nMVHATSPLLLLLLLSLALVAPGLSARKCSLTGKWTNDLGSNMTIGAVNSRGEFTGTYITAVTATSNEIKESPLHGTQNTINKRTQPTFGFTVNWKFSESTTVFTGQCFIDRNGKEVLKTMWLLRSSVNDIGDDWKATRVGINIFTRLRTQKE\n\n### Answer:", "output": "Decrease in about 50% of the molecules. This mutation may result in unaltered biotin-binding affinity or other non-functional changes in the protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01075", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTRMB_AROAE\n\n### Answer:", "output": "tRNA (guanine-N(7)-)-methyltransferase, tRNA (guanine(46)-N(7))-methyltransferase, tRNA(m7G46)-methyltransferase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01076", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "C=C1CC(Cn2ccc(=O)[nH]c2=O)(c2ccc(Cl)cc2)OC1=O", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, ester, ether, halide. Its descriptor profile: MW=332.74, LogP=1.59, TPSA=81.16, HBD=1, HBA=4, QED=0.68.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "C=C1CC(Cn2ccc(=O)[nH]c2=O)(c2ccc(Cl)cc2)OC1=O", "compound_name": null, "pubchem_cid": null, "MW": 332.74, "LogP": 1.59, "TPSA": 81.16, "QED": 0.68, "functional_groups": ["aromatic ring", "ester", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01077", "category": "Mutation", "instruction": "Describe the effect of mutation X5I9Y2-Y60A on protein Alpha-conotoxin GI.", "input": "### Wild-type:\nMGMRMMFTVFLLVVLATTVVSFPSERASDGRDDTAKDEGSDMDKLVEKKECCNPACGRHYSCGR\n\n### Mutant:\nMGMRMMFTVFLLVVLATTVVSFPSERASDGRDDTAKDEGSDMDKLVEKKECCNPACGRHASCGR\n\n### Answer:", "output": "65-fold decrease in inhibitory potency towards mouse muscle-type acetylcholine receptor. No increase in inhibitory potency towards alpha-3-beta-2/CHRNA3-CHRNB2, alpha-3-beta-4/CHRNA3-CHRNB4, alpha-4-beta-4/CHRNA4-CHRNB4, alpha-7/CHRNA7, and alpha-9-alpha-10/CHRNA9-CHRNA10 nAChRs.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01078", "category": "Mutation", "instruction": "Describe the effect of mutation P00898-L289P on protein Anthranilate synthase component 1.", "input": "### Wild-type:\nMQTPKPTLELLTCDAAYRENPTALFHQVCGDRPATLLLESADIDSKDDLKSLLLVDSALRITALGDTVTIQALSDNGASLLPLLDTALPAGVENDVLPAGRVLRFPPVSPLLDEDARLCSLSVFDAFRLLQGVVNIPTQEREAMFFGGLFAYDLVAGFEALPHLEAGNNCPDYCFYLAETLMVIDHQKKSTRIQASLFTASDREKQRLNARLAYLSQQLTQPAPPLPVTPVPDMRCECNQSDDAFGAVVRQLQKAIRAGEIFQVVPSRRFSLPCPSPLAAYYVLKKSNLSPYMFFMQDND\n\n### Mutant:\nMQTPKPTLELLTCDAAYRENPTALFHQVCGDRPATLLLESADIDSKDDLKSLLLVDSALRITALGDTVTIQALSDNGASLLPLLDTALPAGVENDVLPAGRVLRFPPVSPLLDEDARLCSLSVFDAFRLLQGVVNIPTQEREAMFFGGLFAYDLVAGFEALPHLEAGNNCPDYCFYLAETLMVIDHQKKSTRIQASLFTASDREKQRLNARLAYLSQQLTQPAPPLPVTPVPDMRCECNQSDDAFGAVVRQLQKAIRAGEIFQVVPSRRFSLPCPSPLAAYYVLKKSNPSPYMFFMQDND\n\n### Answer:", "output": "Increase in feedback control by tryptophan. The mutant form of the complex with the substitution displayed a marked increase in affinity for tryptophan with little or no change in substrate affinity or catalytic capacity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01079", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "CCOC(=O)c1ccccc1C(=C1C=C(Br)C(=O)C(Br)=C1)c1cc(Br)c(O)c(Br)c1", "output": "The SMILES represents an unnamed molecule with MW=661.97, LogP=7.04, TPSA=63.6. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "Tox21", "canonical_smiles": "CCOC(=O)c1ccccc1C(=C1C=C(Br)C(=O)C(Br)=C1)c1cc(Br)c(O)c(Br)c1", "compound_name": null, "pubchem_cid": null, "MW": 661.97, "LogP": 7.04, "TPSA": 63.6, "QED": 0.36, "functional_groups": ["aromatic ring", "ester", "hydroxyl", "ether", "ketone", "halide", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": 1.0, "SR-ATAD5": null, "SR-HSE": null, "SR-MMP": 1.0, "SR-p53": null}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01080", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLSDPTSTLMFAEAAEAADVVARQFSRNHATMETLAASLRAAPPPFVVTCARGSSDHAATYGKYLLETQLGLVVASASPSVGSVYAAPLQLRGALFIVISQSGKSPDLLRNAEAAKAAGARVIALVNVEDSPLAQLADTVIPLHAGAEKSVAATKSYLASLAALLQLAAYWKQDSSLRAALDLLPDALREAWQCDWSAVTEGLVEATNLFVLGRGLGLGAAQEAALKFKETCSLHAEAYSSAEVKHGPMALVDRGFPVLAFAQPDETGAGTRAVVEEFTARGAQVWMAGAGGNLPVAAAPHPLCAPLLTVQSFYRAINALALRRGFNPDLPPHLNKV\n\n### Answer:", "output": "<SEQ_3Di>DPDLCPALLNVLLLCQLVLLVVLCVVCVVVLLVVLVVCVVQPAQAEEEEEDDLQVLLVLLLQLLCCVQVVGHYYYDDQPPPDDDPDQDQRARYAYEYEEAQQAPPSSLVVLVSNVVNHYQYEYAYQHCPHSRNVSHPYYRHLSSPDRLFPDDRSSSLSSNSVSLSNSCSNSVPVQSVVLSVVLSVQLNVLLPWDPVVVLVVCLPAAAEEEWFARLQQSLSQNLQVLCCKLPVHNYYYDYLVVCVPPVVVDQAQRYAYEFEAELVGGTNHDPVSVVVSVVRVHHYFYDYNPTPTDWGDRSDSSSRNSSVSSNSSSSSSSSSVSNQFRSRDTPVVTGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01081", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nESALERRSSSVVMNNVMKKPDLSDPKLRAKLAKGMGRNYYGEPAWPNDLLYMFPVCILGTFAAIVGLAVMQPTPTGEPANPFATPLEILPEWYFFPTFNLLRVIPNKLLGVLSMAAVPAGLITVPFIENVNKFQNPFRRPVATSVFLLGTVVAIWLGIGATLPIDKAISLGFW\n\n### Sequence 2:\nSILKKPDLTDPKLRAKLAKGMGHNYYGEPAWPNDLLYVFPVVIIGTFACSIGLAILEPSSIGEKSNPFATPLEILPEWYFFPTFNLLRVIPNKLLGVLSMAAVPAGLLTVPFIENVNKFQNPFRRPIATTVFLIGTVVSIWLGIGATMPINNAITLGLF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01082", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGATC_METCA\n\n### Answer:", "output": "Aspartyl/glutamyl-tRNA(Asn/Gln) amidotransferase subunit C, Asp/Glu-ADT subunit C.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01083", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAEQHGAPEQAAAGKSHGDLGGSYKVILYELENFQGKRCELSAECPSLTDSLLEKVGSIQVESGPWLAFESRAFRGEQFVLEKGDYPRWDAWSNSRDSDSLLSLRPLNIDSPHHKLHLFENPAFSGRKMEIVDDDVPSLWAHGFQDRVASVRAINGTWVGYEFPGYRGRQYVFERGEYRHWNEWDASQPQLQSVRRIRDQKWHKRGRFPSS\n\n### Sequence 2:\nMAEQHSTPEQAAAGKSHGGLGGSYKVIVYEMENFQGKRCELTAECPNLTESLLEKVGSIQVESGPWLAFERRAFRGEQYVLEKGDYPRWDAWSNSHHSDSLLSLRPLHIDGPDHKLHLFENPAFGGRKMEIVDDDVPSLWAHGFQDRVASVRAINGTWVGYEFPGYRGRQYVFERGEYRHWNEWDANQPQLQSVRRIRDQKWHKRGVFLSS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01084", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDAPA_LIMRJ\n\n### Answer:", "output": "Catalyzes the condensation of (S)-aspartate-beta-semialdehyde [(S)-ASA] and pyruvate to 4-hydroxy-tetrahydrodipicolinate (HTPA).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01085", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMHICILSGSTLGGAEYVAEHLNDVLETQGFSTALFHGPNLSDIENEKIWLVVTSTHGAGELPDNLKPLFDELANSQKDFSDVRFAVVGLGSSDYDTFCYAADKVEQTLQAKSAVKICETLKIDVLNVDDQESYAEEWLPSFIEGLK\n\n### Sequence 2:\nICIITGSTLGGAEYVAEHIAEILEQQDYPVRLEHGPNFEEVIDEKCWLVVTSTHGAGELPDNIKPLFEKLAFHPKQLADLRFAVIGLGNSDYDTFCHAVDHVEQLLLSKDALQLCESLRMDMLTITDPEHTAEQWLPQFLSQL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01086", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMALDTPLSRAPAIAPEPKGILDPATGLPIGATDPTFLSINDELADRGFLLTTADDLITWARTGSLMWMTFGLACCAVEMMQMSMPRYDAERFGFAPRASPRQSDVMIVAGTLTNKMAPALRKVYDQMPEPRYVISMGSCANGGGYYHYSYSVVRGCDRVVPVDIYVPGCPPSAEALLYGVLLLQKKIRRTGTIER\n\n### Sequence 2:\nMALTPTFSRAPDIAPAPKGIIDPATGRPIGANDPTFLSINDELADRGFLVTSADELINWARTGSLMWMTFGLACCAVEMMQMSMPRYDCERFGFAPRGSPRQSDVMIVAGTLTNKMAPALRKVYDQMPEPRYVISMGSCANGGGYYHYSYSVVRGCDRVVPVDIYVPGCPPSAEALLYGVLLLQRKIRRIGTIER\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01087", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAKVNLIESPYSLLQLKGIGPKKIEVLQQLNIHTVEDLVLYLPTRYEDNTVIDLNQAEDQSNVTIEGQVYTAPVVAFFGRNNSKLTVHLMVNNIAVKCIFFNQPYLKKKIELNQTITVKGKWNRVKQEITGNRVFFNSQGTQTQENADVQLEPVYRIKEGIKQKQIRDQIRQALNDVTIHEWLTDELREKYKLETLDFTLNTLHHPKSKEDLLRARRTYAFTELFLFELRMQWLNRLEKSSDEAIEIDYGLDQVKSFIDRLPFELTEAQKSSVNEIFRDLKAPIRMHRLLQGDVGSGKTVVAAICMYALKTAGYQSALMVPTEILAEQHAESLMALFGDSMNVALLTGSVKGKKRKILLEQLENGTIDCLIGTHALIQDDVIFHNVGLVITDEQHRFGVNQRQLLREKGAMTNVLFMTATPIPRTLAISVFGEMDVSSIKQLPKGRKPIITTWAKHEQYDKVLMQMTSELKKGRQAYVICPLIESSEHLEDVQNVVALYESLQQYYGVSRVGLLHGKLSADEKDEVMQKFSNHEIDVLVSTTVVEVGVNVPNATFMMIYDADRFGLSTLHQLRGRVGRSDQQSYCVLIASPKTETGIERMTIMTQTTDGFELSERDLEMRGPGDFFGVKQSGLPDFLVANLVEDYRMLEVARDEAAELIQSGVFFENTYQHLRHFVEENLLHRSFD\n\n### Sequence 2:\nMAKVNLIESPYSLLQLKGIGPKKIEVLQQLNIHTVEDLVLYLPTRYEDNTVIDLNQAEDQSNVTIEGQVYTAPVVAFFGRNKSKLTVHLMVNNIAVKCIFFNQPYLKKKIELNQTITVKGKWNRVKQEITGNRVFFNSQGTQTQENADVQLEPVYRIKEGIKQKQIRDQIRQALNDVTIHEWLTDELREKYKLETLDFTLNTLHHPKSKEDLLRARRTYAFTELFLFELRMQWLNRLEKSSDEAIEIDYDLDQVKSFIDRLPFELTEAQKSSVNEIFRDLKAPIRMHRLLQGDVGSGKTVVAAICMYALKTAGYQSALMVPTEILAEQHAESLMALFGDSMNVALLTGSVKGKKRKILLEQLENGTIDCLIGTHALIQDDVIFHNVGLVITDEQHRFGVNQRQLLREKGAMTNVLFMTATPIPRTLAISVFGEMDVSSIKQLPKGRKPIITTWAKHEQYDKVLMQMTSELKKGRQAYVICPLIESSEHLEDVQNVVALYESLQQYYGVSRVGLLHGKLSADEKDEVMQKFSNHEIDVLVSTTVVEVGVNVPNATFMMIYDADRFGLSTLHQLRGRVGRSDQQSYCVLIASPKTETGIERMTIMTQTTDGFELSERDLEMRGPGDFFGVKQSGLPDFLVANLVEDYRMLEVARDEAAELIQSGVFFENTYQHLRHFVEENLLHRSFD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01088", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: Cc1cnc(-c2ccc(C)c(NC(=O)c3ccc(OCc4ccccn4)cc3)c2)[nH]1\nDescriptors: MW=398.47, LogP=4.92, TPSA=79.9, HBD=2, HBA=4", "output": "Descriptors: MW=398.47, LogP=4.92, TPSA=79.9, HBD=2, HBA=4, QED=0.48, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "Cc1cnc(-c2ccc(C)c(NC(=O)c3ccc(OCc4ccccn4)cc3)c2)[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 398.47, "LogP": 4.92, "TPSA": 79.9, "QED": 0.48, "functional_groups": ["aromatic ring", "amide", "ether", "pyridine", "imidazole"], "moleculenet_labels": {"label": 3.19}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01089", "category": "Mutation", "instruction": "Describe the effect of mutation P20586-G45A on protein p-hydroxybenzoate hydroxylase.", "input": "### Wild-type:\nMKTQVAIIGAGPSGLLLGQLLHKAGIDNVILERQTPDYVLGRIRGGVLEQGMVDLLREAGVDRRMARDGLVHEGVEIAFAGQRRRIDLKRLSGGKTVTVYGQTEVTRDLMEAREACGATTVYQAAEVRLHDLQGERPYVTFERDGERLRLDCDYIAGCDGFHGISRQSIPAERLKVFERVYPFGWLGLLADTPPVSHELIYANHPRGFALCSQRSATRSRYYVQVPLSEKVEDWSDERFWTELKARLPSEVAEKLVTGPSLEKSIAPLRSFVVEPMQHGRLFLAGDAAHIVPPTGAKGLN\n\n### Mutant:\nMKTQVAIIGAGPSGLLLGQLLHKAGIDNVILERQTPDYVLGRIRAGVLEQGMVDLLREAGVDRRMARDGLVHEGVEIAFAGQRRRIDLKRLSGGKTVTVYGQTEVTRDLMEAREACGATTVYQAAEVRLHDLQGERPYVTFERDGERLRLDCDYIAGCDGFHGISRQSIPAERLKVFERVYPFGWLGLLADTPPVSHELIYANHPRGFALCSQRSATRSRYYVQVPLSEKVEDWSDERFWTELKARLPSEVAEKLVTGPSLEKSIAPLRSFVVEPMQHGRLFLAGDAAHIVPPTGAKGLN\n\n### Answer:", "output": "The positions of the substrate and the flavin are not altered. The crystal structure of the enzyme is a symmetric dimer, with both monomers identical to wild-type PHBH. H72 is maintained in its original position within the active site, and no changes are observed in the structure. The two structures correlate with evidence from kinetic studies for two forms of PHBH. One form of the enzyme dominates turnover and hydroxylates, while the other contributes little to turnover and fails to hydroxylate. PHBH favors alternative conformations over the in conformation. The effect of this mutation on the structure and function of PHBH illustrates the unimportance of the si side loop in the conformational state of PHBH and, consequently, the function of the enzyme. This work demonstrates some general principles of how enzymes use conformational movements to restrict both access and egress of substrates and product, while allowing access to the solvent at a critical stage in catalysis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01090", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSPEB_SALSV\n\n### Answer:", "output": "Belongs to the arginase family. Agmatinase subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01091", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMVSPRVRLAALALSVCAILCLGMHASAFPPKPDNPGDNASPEQMARYKAAVRHYINLITRQRYGKRALTPENWIYRDPAEERVTYGLDDYAMW\n\n### Sequence 2:\nMAQQSPYSAAMAEQRHQEWLRFVDLLKNAYQNDLHLPLLNLMLTPDEREALGTRVRIVEELLRGEMSQRELKNELGAGIATITRGSNSLKAAPVELRQWLEDVLLKSD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01092", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPURE_VIBPA\n\n### Answer:", "output": "N5-carboxyaminoimidazole ribonucleotide mutase, N5-CAIR mutase, 5-(carboxyamino)imidazole ribonucleotide mutase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01093", "category": "Mutation", "instruction": "Describe the effect of mutation Q8IV16-C68Y on protein Glycosylphosphatidylinositol-anchored high density lipoprotein-binding protein 1.", "input": "### Wild-type:\nMKALGAVLLALLLFGRPGRGQTQQEEEEEDEDHGPDDYDEEDEDEVEEEETNRLPGGRSRVLLRCYTCKSLPRDERCNLTQNCSHGQTCTTLIAHGNTESGLLTTHSTWCTDSCQPITKTVEGTQVTMTCCQSSLCNVPPWQSSRVQDPTGKGAGGPRGSSETVGAALLLNLLAGLGAMGARRP\n\n### Mutant:\nMKALGAVLLALLLFGRPGRGQTQQEEEEEDEDHGPDDYDEEDEDEVEEEETNRLPGGRSRVLLRCYTYKSLPRDERCNLTQNCSHGQTCTTLIAHGNTESGLLTTHSTWCTDSCQPITKTVEGTQVTMTCCQSSLCNVPPWQSSRVQDPTGKGAGGPRGSSETVGAALLLNLLAGLGAMGARRP\n\n### Answer:", "output": "In HLPP1D; does not interact with LPL; promotes formation of dimers and oligomers severely reducing number of monomers. Missense mutation in lipoprotein lipase (LPL) is a critical determinant of childhood-onset chylomicronaemia syndrome. The mutation in GPIHBP1 interferes with lipoprotein lipase (LPL) binding, leading to familial chylomicronemia.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01094", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nFCGR3B S100A9 CXCR2 S100A8\n</cell>", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as FCGR3B, S100A9, S100A8 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "heart", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["FCGR3B", "S100A9", "CXCR2", "S100A8"]}}
{"id": "eval_01095", "category": "Mutation", "instruction": "Describe the effect of mutation P00451-Y155H on protein Coagulation factor VIII.", "input": "### Wild-type:\nMQIELSTCFFLCLLRFCFSATRRYYLGAVELSWDYMQSDLGELPVDARFPPRVPKSFPFNTSVVYKKTLFVEFTDHLFNIAKPRPPWMGLLGPTIQAEVYDTVVITLKNMASHPVSLHAVGVSYWKASEGAEYDDQTSQREKEDDKVFPGGSHTYVWQVLKENGPMASDPLCLTYSYLSHVDLVKDLNSGLIGALLVCREGSLAKEKTQTLHKFILLFAVFDEGKSWHSETKNSLMQDRDAASARAWPKMHTVNGYVNRSLPGLIGCHRKSVYWHVIGMGTTPEVHSIFLEGHTFLVRNH\n\n### Mutant:\nMQIELSTCFFLCLLRFCFSATRRYYLGAVELSWDYMQSDLGELPVDARFPPRVPKSFPFNTSVVYKKTLFVEFTDHLFNIAKPRPPWMGLLGPTIQAEVYDTVVITLKNMASHPVSLHAVGVSYWKASEGAEYDDQTSQREKEDDKVFPGGSHTHVWQVLKENGPMASDPLCLTYSYLSHVDLVKDLNSGLIGALLVCREGSLAKEKTQTLHKFILLFAVFDEGKSWHSETKNSLMQDRDAASARAWPKMHTVNGYVNRSLPGLIGCHRKSVYWHVIGMGTTPEVHSIFLEGHTFLVRNH\n\n### Answer:", "output": "In HEMA; moderate. The missense mutation was investigated for the nature of the new amino acid, the location of the substituted amino acid within crystallographic and theoretical models, and the degree of conservation of the native residue in factor VIII (FVIII) protein and FVIII-related protein family aligned sequences.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01096", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nAROA_PROMT\n\n### Answer:", "output": "Belongs to the EPSP synthase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01097", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nKDSB_PSEAE\n\n### Answer:", "output": "3-deoxy-manno-octulosonate cytidylyltransferase, CMP-2-keto-3-deoxyoctulosonic acid synthase, CKS, CMP-KDO synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01098", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "KLF1, GATA1, HBB", "output": "The expression pattern is consistent with a healthy erythroid progenitor without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "colon", "disease": "normal", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["KLF1", "GATA1", "HBB"]}}
{"id": "eval_01099", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNUOK_NEIMB\n\n### Answer:", "output": "NADH-quinone oxidoreductase subunit K, NADH dehydrogenase I subunit K, NDH-1 subunit K.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01100", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSRAGTWLKMLGAGIVICVGGPAFVQSIRPTDEELFKRYNPELQRRSLEEGDRRAQEFDDYVNRLKQWSKSDKSIWYAAQEQQEQKRSEAEALRNQAKDEARAQREEMRKELLGGK\n\n### Sequence 2:\nMSRAGTWMKMLGVGIVICVGGPAFVQYIRPTDEELFKRYNPDLQKRSLEEGDRRAREFDEYVTRLKQWSKSDKSIWYAAQEQQEQKRVEAEVQRNQARDDAKVQREEMRKELLGEK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01101", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPSSISVKLVAAESLYKRDVFRSPDPFAVLTIDGYQTKSTSAAKKTLNPYWNETFKFDDINENSILTIQVFDQKKFKKKDQGFLGVVNVRVGDVLGHLDEDTATSSGRPREETITRDLKKSNDGMAVSGRLIVVLSKLPSSSPHSQAPSGHTASSSTNTSSTTRTNGHSTSSTRNHSTSHPSRGTAQAVESTLQSGTTAATNTATTSHRSTNSTSSATRQYSSFEDQYGRLPPGWERRTDNFGRTYYVDHNTRTTTWKRPTLDQTEAERGNQLNANTELERRQHRGRTLPGGSSDNSSVTVQVGGGSNIPPVNGAAAAAFAATGGTTSGLGELPSGWEQRFTPEGRAYFVDHNTRTTTWVDPRRQQYIRTYGPTNTTIQQQPVSQLGPLPSGWEMRLTNTARVYFVDHNTKTTTWDDPRLPSSLDQNVPQYKRDFRRKVIYFRSQPALRILPGQCHIKVRRKNIFEDAYQEIMRQTPEDLKKRLMIKFDGEEGLDYGGVSREFFFLLSHEMFNPFYCLFEYSAYDNYTIQINPNSGINPEHLNYFKFIGRVVGLGVFHRRFLDAFFVGALYKMMLRKKVVLQDMEGVDAEVYNSLNWMLENSIDGVLDLTFSADDERFGEVVTVDLKPDGRNIEVTDGNKKEYVELYTQWRIVDRVQEQFKAFMDGFNELIPEDLVTVFDERELELLIGGIAEIDIEDWKKHTDYRGYQESDEVIQWFWKCVSEWDNEQRARLLQFTTGTSRIPVNGFKDLQGSDGPRRFTIEKAGEVQQLPKSHTCFNRVDLPQYVDYDSMKQKLTLAVEETIGFGQE\n\n### Sequence 2:\nNLRVTIIAADGLYKRDVFRFPDPFAVATVGGEQTHTTSVIKKTLNPYWNEMFDLR-VNEDSILAIQIFDQKKFKKKDQGFLGVINVRIGDVI-----DLQMGG----DEMLTRDLKKSNDNLVVHGKLIINLSTNLSTPNTNQANGLHRSNLQSSTSSGLVPQVTAPHASPGPSQLDPTASNPSLNPQRVPSTTRPSSTVAPVNGAA-------APGASRTNLSSFEDSQGRLPAGWERREDNLGRTYYVDHNTRTTTWTRPSSNYNEATQRTQREANMQLERRAHQSRMLPEDRTGASSPNLQENQQAQTPPAGGSANAVSMMATGATTAGTGELPPGWEQRTTPEGRPYFVDHNTRTTTWVDPRRQQYIRMYGQNANGNNTTIQQQPVSQLGPLPSGWEMRLTNTARVYFVDHNTKTTTWDDPRLPSSLDQGVPQYKRDFRRKLIYFRSQPALRIMSGQCHVKVRRNNIFEDSYAEIMRQSASDLKKRLMIKFDGEDGLDYGGLSREFFFLLSHEMFNPFYCLFEYSAHDNYTLQINPHSGVNPEHLNYFKFIGRVVGLAIFHRRFLDSFFIGAFYKMMLRKKVSLQDMEGVDEDLHRNLTWTLDNDIEGIIELTFAVDDEKFGERRTIDLKPGGRDIPVTNENKHEYVELVTEWKIVKRVEEQFNAFMSGFNELIPADLVNVFDERELELLIGGIADIDVDDWKKHTDYRGYQEQDEVIQNFWKIVRTWDAEQKSRLLQFTTGTSRIPVNGFKDLQGSDGPRRFTIEKSGDPAALPKSHTCFNRLDLPPYKTHETLEHKLSIAVEETLGFGQE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01102", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL3_DECAR\n\n### Answer:", "output": "50S ribosomal protein L3.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01103", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLNKNKRWYLIALYDIYQGLLTTKQCEYFNLHYFKDLSFSEIAELKEISKSAISDCLNKVCDQLLKYEQALLIYEKNKKRNDLYTLINDSELVKKLKDI\n\n### Sequence 2:\nMDIQLINIGFGNIVSANRVIAIVSPESAPIKRIISDARERGQLIDATYGRRTRAVIITDSSHVVLSAIQPETVAHRFVVNKEVQANSN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01104", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "N[C@@H](C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(Cl)CS[C@H]12)c1ccccc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, hydroxyl, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BBBP", "canonical_smiles": "N[C@@H](C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(Cl)CS[C@H]12)c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 367.81, "LogP": 0.62, "TPSA": 112.73, "QED": 0.67, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "hydroxyl", "halide"]}}
{"id": "eval_01105", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nQSVLTQPPSVSVAPGQTARISCSGDNIGSYYVHWYQQKPGQAPVLVIYEDSERPSGIPERFSGSNSGNTATLTISGTQAEDEADYYCSSYDDPNFQVFGGGTKLTVLGQPKAAPSVTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADSSPVKAGVETTTPSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTHEGSTVEKTVAPTECS\n\n### Answer:", "output": "<SEQ_2D>CCCCBCCSEEEECTTSCEEECCBCTTGGGSCCEEEEECTTSCCEEEEBTTTBCCTTCCTTEEEEEETTEEEEEESSCCSTTCSEEEEEECTTTTCCEECCCEEEEECCSCCBCCEEEEECCCHHHHHTTCEEEEEEEEEEBSCCEEEEEEETTEECCTTEEECCCEECTTSCEEEEEEEEECHHHHTTSSCEEEEEEETTEEEEEEECSCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01106", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGPDA_RICTY\n\n### Answer:", "output": "Glycerol-3-phosphate dehydrogenase [NAD(P)+], NAD(P)H-dependent glycerol-3-phosphate dehydrogenase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01107", "category": "Mutation", "instruction": "Describe the effect of mutation Q8IWQ3-T260A on protein Serine/threonine-protein kinase BRSK2.", "input": "### Wild-type:\nMTSTGKDGGAQHAQYVGPYRLEKTLGKGQTGLVKLGVHCVTCQKVAIKIVNREKLSESVLMKVEREIAILKLIEHPHVLKLHDVYENKKYLYLVLEHVSGGELFDYLVKKGRLTPKEARKFFRQIISALDFCHSHSICHRDLKPENLLLDEKNNIRIADFGMASLQVGDSLLETSCGSPHYACPEVIRGEKYDGRKADVWSCGVILFALLVGALPFDDDNLRQLLEKVKRGVFHMPHFIPPDCQSLLRGMIEVDAARRLTLEHIQKHIWYIGGKNEPEPEQPIPRKVQIRSLPSLEDIDP\n\n### Mutant:\nMTSTGKDGGAQHAQYVGPYRLEKTLGKGQTGLVKLGVHCVTCQKVAIKIVNREKLSESVLMKVEREIAILKLIEHPHVLKLHDVYENKKYLYLVLEHVSGGELFDYLVKKGRLTPKEARKFFRQIISALDFCHSHSICHRDLKPENLLLDEKNNIRIADFGMASLQVGDSLLETSCGSPHYACPEVIRGEKYDGRKADVWSCGVILFALLVGALPFDDDNLRQLLEKVKRGVFHMPHFIPPDCQSLLRGMIEVDAARRLALEHIQKHIWYIGGKNEPEPEQPIPRKVQIRSLPSLEDIDP\n\n### Answer:", "output": "Decreased phosphorylation. Nearly abolishes stimulation of insulin secretion. Phosphorylation of BRSK2 at T260 by PKA increases its kinase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01108", "category": "Mutation", "instruction": "Describe the effect of mutation P06213-S762R on protein Insulin receptor.", "input": "### Wild-type:\nMATGGRRGAAAAPLLVAVAALLLGAAGHLYPGEVCPGMDIRNNLTRLHELENCSVIEGHLQILLMFKTRPEDFRDLSFPKLIMITDYLLLFRVYGLESLKDLFPNLTVIRGSRLFFNYALVIFEMVHLKELGLYNLMNITRGSVRIEKNNELCYLATIDWSRILDSVEDNYIVLNKDDNEECGDICPGTAKGKTNCPATVINGQFVERCWTHSHCQKVCPTICKSHGCTAEGLCCHSECLGNCSQPDDPTKCVACRNFYLDGRCVETCPPPYYHFQDWRCVNFSFCQDLHHKCKNSRRQG\n\n### Mutant:\nMATGGRRGAAAAPLLVAVAALLLGAAGHLYPGEVCPGMDIRNNLTRLHELENCSVIEGHLQILLMFKTRPEDFRDLSFPKLIMITDYLLLFRVYGLESLKDLFPNLTVIRGSRLFFNYALVIFEMVHLKELGLYNLMNITRGSVRIEKNNELCYLATIDWSRILDSVEDNYIVLNKDDNEECGDICPGTAKGKTNCPATVINGQFVERCWTHSHCQKVCPTICKSHGCTAEGLCCHSECLGNCSQPDDPTKCVACRNFYLDGRCVETCPPPYYHFQDWRCVNFSFCQDLHHKCKNSRRQG\n\n### Answer:", "output": "Increases insulin binding to cells with this mutation is severely reduced, but can be decreased by treatment with trypsin.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01109", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTTCA_PHOPR\n\n### Answer:", "output": "Catalyzes the ATP-dependent 2-thiolation of cytidine in position 32 of tRNA, to form 2-thiocytidine (s(2)C32).  The sulfur atoms are provided by the cysteine/cysteine desulfurase (IscS) system.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01110", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "IL2RA, CD3D, IKZF2, FOXP3, CD4, CTLA4", "output": "This profile is most consistent with regulatory T cell, which is commonly found in heart under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "heart", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["IL2RA", "CD3D", "IKZF2", "FOXP3", "CD4", "CTLA4"]}}
{"id": "eval_01111", "category": "Mutation", "instruction": "Describe the effect of mutation Q6QNK2-T369M on protein Adhesion G-protein coupled receptor D1.", "input": "### Wild-type:\nMEKLLRLCCWYSWLLLFYYNFQVRGVYSRSQDHPGFQVLASASHYWPLENVDGIHELQDTTGDIVEGKVNKGIYLKEEKGVTLLYYGRYNSSCISKPEQCGPEGVTFSFFWKTQGEQSRPIPSAYGGQVISNGFKVCSSGGRGSVELYTRDNSMTWEASFSPPGPYWTHVLFTWKSKEGLKVYVNGTLSTSDPSGKVSRDYGESNVNLVIGSEQDQAKCYENGAFDEFIIWERALTPDEIAMYFTAAIGKHALLSSTLPSLFMTSTASPVMPTDAYHPIITNLTEERKTFQSPGVILSYL\n\n### Mutant:\nMEKLLRLCCWYSWLLLFYYNFQVRGVYSRSQDHPGFQVLASASHYWPLENVDGIHELQDTTGDIVEGKVNKGIYLKEEKGVTLLYYGRYNSSCISKPEQCGPEGVTFSFFWKTQGEQSRPIPSAYGGQVISNGFKVCSSGGRGSVELYTRDNSMTWEASFSPPGPYWTHVLFTWKSKEGLKVYVNGTLSTSDPSGKVSRDYGESNVNLVIGSEQDQAKCYENGAFDEFIIWERALTPDEIAMYFTAAIGKHALLSSTLPSLFMTSTASPVMPTDAYHPIITNLTEERKTFQSPGVILSYL\n\n### Answer:", "output": "Does not affect subcellular location; does not change G-protein coupled receptor activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01112", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "O=C(Cc1c(C(=O)O)[nH]c2ccccc12)Nc1cccc(C(F)(F)F)c1", "output": "The molecule contains the following functional groups: aromatic ring, amide, hydroxyl, halide, indole.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "O=C(Cc1c(C(=O)O)[nH]c2ccccc12)Nc1cccc(C(F)(F)F)c1", "compound_name": null, "pubchem_cid": null, "MW": 362.31, "LogP": 4.07, "TPSA": 82.19, "QED": 0.65, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide", "indole"], "moleculenet_labels": {"label": 1.33}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "hydroxyl", "halide", "indole"]}}
{"id": "eval_01113", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "PDCD1, TOX, CD8A, LAG3, HAVCR2", "output": "The marker profile suggests functional enrichment consistent with the typical biology of exhausted CD8 T cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "spleen", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["PDCD1", "TOX", "CD8A", "LAG3", "HAVCR2"]}}
{"id": "eval_01114", "category": "Mutation", "instruction": "Describe the effect of mutation P70169-N157D on protein Double C2-like domain-containing protein beta.", "input": "### Wild-type:\nMTLRRRGEKATISIQEHMAIDVCPGPIRPIKQISDYFPRFPRGLPPTAAPRAPAPPDAPARSPAASASPRSPSDGARDDDEDVDQLFGAYGASPGPSPGPSPARPPAKPPEDEPDVDGYESDDCTALGTLDFSLLYDQENNALHCTISKAKGLKPMNHNGLADPYVKLHLLPGASKANKLRTKTLRNTLNPSWNETLTYYGITDEDMVRKTLRISVCDEDKFRHNEFIGETRVPLKKLKPNHTKTFSICLEKQLPVDKAEDKSLEERGRILISLKYSSQKQGLLVGIVRCAHLAAMDANG\n\n### Mutant:\nMTLRRRGEKATISIQEHMAIDVCPGPIRPIKQISDYFPRFPRGLPPTAAPRAPAPPDAPARSPAASASPRSPSDGARDDDEDVDQLFGAYGASPGPSPGPSPARPPAKPPEDEPDVDGYESDDCTALGTLDFSLLYDQENNALHCTISKAKGLKPMDHNGLADPYVKLHLLPGASKANKLRTKTLRNTLNPSWNETLTYYGITDEDMVRKTLRISVCDEDKFRHNEFIGETRVPLKKLKPNHTKTFSICLEKQLPVDKAEDKSLEERGRILISLKYSSQKQGLLVGIVRCAHLAAMDANG\n\n### Answer:", "output": "Decrease of interaction with STX4 and insulin-dependent translocation to the cell membrane; when associated with N-163, N-297 and N-303.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01115", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "Cc1c(O)nc(O)nc1C(=O)O", "output": "This molecule is small and polar, with hydrophilic character. It contains hydroxyl, phenol. Its descriptor profile: MW=170.12, LogP=-0.11, TPSA=103.54, HBD=3, HBA=5, QED=0.54.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "Cc1c(O)nc(O)nc1C(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 170.12, "LogP": -0.11, "TPSA": 103.54, "QED": 0.54, "functional_groups": ["hydroxyl", "phenol"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01116", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: CC(C)[N+]1(C)C2CCC1CC(OC(=O)C(CO)c1ccccc1)C2.[Br-]\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and beta-secretase (BACE1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "CC(C)[N+]1(C)C2CCC1CC(OC(=O)C(CO)c1ccccc1)C2.[Br-]", "compound_name": null, "pubchem_cid": null, "MW": 412.37, "LogP": -0.14, "TPSA": 46.53, "QED": 0.54, "functional_groups": ["aromatic ring", "carboxylic acid", "ester", "hydroxyl", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": false}}
{"id": "eval_01117", "category": "Mutation", "instruction": "Describe the effect of mutation Q6SZW1-Q150A on protein NAD.", "input": "### Wild-type:\nMVLTLLLSAYKLCRFFAMSGPRPGAERLAVPGPDGGGGTGPWWAAGGRGPREVSPGAGTEVQDALERALPELQQALSALKQAGGARAVGAGLAEVFQLVEEAWLLPAVGREVAQGLCDAIRLDGGLDLLLRLLQAPELETRVQAARLLEQILVAENRDRVARIGLGVILNLAKEREPVELARSVAGILEHMFKHSEETCQRLVAAGGLDAVLYWCRRTDPALLRHCALALGNCALHGGQAVQRRMVEKRAAEWLFPLAFSKEDELLRLHACLAVAVLATNKEVEREVERSGTLALVEPLV\n\n### Mutant:\nMVLTLLLSAYKLCRFFAMSGPRPGAERLAVPGPDGGGGTGPWWAAGGRGPREVSPGAGTEVQDALERALPELQQALSALKQAGGARAVGAGLAEVFQLVEEAWLLPAVGREVAQGLCDAIRLDGGLDLLLRLLQAPELETRVQAARLLEAILVAENRDRVARIGLGVILNLAKEREPVELARSVAGILEHMFKHSEETCQRLVAAGGLDAVLYWCRRTDPALLRHCALALGNCALHGGQAVQRRMVEKRAAEWLFPLAFSKEDELLRLHACLAVAVLATNKEVEREVERSGTLALVEPLV\n\n### Answer:", "output": "In WQH to A mutant: Increased NAD(+)-binding to ARM repeats, leading to decreased NAD(+) hydrolase activity; when associated with A-103 and A-190.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01118", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMATLSLTVNSGDPPLGALLAVEHVKDDVSISVEEGKENILHVSENVIFTDVNSILRYLARVATTAGLYGSNLMEHTEIDHWLEFSATKLSSCDSFTSTINELNHCLSLRTYLVGNSLSLADLCVWATLKGNAAWQEQLKQKKAPVHVKRWFGFLEAQLEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCEEEEEETTSCCHHHHHHHHHHTTTSCEEEEEESSCEEECSSSCEECSHHHHHHHHHHHCTTTTTTCSSHHHHHHHHHHHHHHHHTTTSCTTHHHHHHHHHHHTSSCSSSSTTSCCHHHHHHHHHHHHCHHHHHHHHTTCSCHHHHHHHHHHHHHCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01119", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: ALK\nLigand SMILES: c1cc2c3c(c4ccc5ccccc5c4cc3c1)CC2\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "ESOL", "canonical_smiles": "c1cc2c3c(c4ccc5ccccc5c4cc3c1)CC2", "compound_name": null, "pubchem_cid": null, "MW": 254.33, "LogP": 5.24, "TPSA": 0.0, "QED": 0.3, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"label": -7.85}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01120", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "COc1ccc2c(c1)/C(=C/CCN(C)C)c1ccccc1O2", "output": "The molecule contains the following functional groups: aromatic ring, amine, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BBBP", "canonical_smiles": "COc1ccc2c(c1)/C(=C/CCN(C)C)c1ccccc1O2", "compound_name": null, "pubchem_cid": null, "MW": 295.38, "LogP": 4.18, "TPSA": 21.7, "QED": 0.72, "functional_groups": ["aromatic ring", "amine", "ether"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amine", "ether"]}}
{"id": "eval_01121", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD4, IL2RA, IKZF2, CD3D", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as IL2RA, IKZF2, CD4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD4", "IL2RA", "IKZF2", "CD3D"]}}
{"id": "eval_01122", "category": "Mutation", "instruction": "Describe the effect of mutation Q6NKS4-A174N on protein Light-harvesting complex-like protein 3 isotype 2, chloroplastic.", "input": "### Wild-type:\nMSISMALFSPPISSSLQNPNLIPKISTSLLSTKRFSLISVPRASSDNGTTSPVVEIPKPASVAVEEVPVKSPAESSSASENGAVGGEATDSSTETVIKYQNAKWVNGTWDLKQFEKDGKTDWDSVIVSEAKRRKWLEDNPETTSNDELVVFDTSIIPWWAWMKRYHLPEAELLAGRAAMIGFFMAYFVDSLTGVGLVDQMGNFFCKTLLFVAVAGVLFIRKNEDLDKLKDLFDETTLYDKQWQAAWKEPDSSTVSSKK\n\n### Mutant:\nMSISMALFSPPISSSLQNPNLIPKISTSLLSTKRFSLISVPRASSDNGTTSPVVEIPKPASVAVEEVPVKSPAESSSASENGAVGGEATDSSTETVIKYQNAKWVNGTWDLKQFEKDGKTDWDSVIVSEAKRRKWLEDNPETTSNDELVVFDTSIIPWWAWMKRYHLPEAELLNGRAAMIGFFMAYFVDSLTGVGLVDQMGNFFCKTLLFVAVAGVLFIRKNEDLDKLKDLFDETTLYDKQWQAAWKEPDSSTVSSKK\n\n### Answer:", "output": "Increases interaction with GGR; when associated with A-171 and A-189. The LHC motif of LIL3 is involved in the complex formation of LIL3 and GGR, which might contribute to the GGR reaction.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01123", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCATC_SCHMA\n\n### Answer:", "output": "Thiol protease.  Has a role as a digestive enzyme.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01124", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMANSRTVLILCGDYMEDYEVMVPFQALQAFGITVHTVCPGKKAGDSCPTAVHDFCGHQTYFESRGHNFTLNATFDEVDLSKYDGLVIPGGRAPEYLALTASVVELVKEFSRSGKPIASICHGQLILAAADTVNGRKCTAYATVGPSLVAAGAKWVEPITPDVCVVDGSLITAATYEGHPEFIQLFVKALGGKITGANKRILFLCGDYMEDYEVKVPFQSLQALGCQVDAVCPEKKAGDRCPTAIHDFEGDQTYSEKPGHTFALTTNFDDLVSSSYDALVIPGGRAPEYLALNEHVLNIVKEFMNSEKPVASICHGQQILAAAGVLKGRKCTAYPAVKLNVVLGGGTWLEPDPIDRCFTDGNLVTGAAWPGHPEFVSQLMALLGIQVSF\n\n### Sequence 2:\nMSDTTLELAKRIIACASVTPEDAGCMDILIERLKPLGFSIEFINRNGVTNLWARRGTTAPLFVFAGHTDVVPTGPLDKWTSPPFAPEIRDGVLYGRGTADMKSSVAASVTAVEAFVAANPQHPGSLAFLLTSDEEGDANDGTIAVVEALKARGETLDFCIIGEPTSVDTLGDMVKNGRRGSLSGVLTVKGIQCHIAYPEKGRNPIHEAAPALAELAATEWDQGNEYYQPTTWQISNIHGGTGATNVVPGSVDIKFNFRFSTASTPEGLQQRLSAILEKHKLDYEIKWTLGARPFLTGRGPLADAATTAIREICGIETELSTTGGTSDGRFIAEICRQMLEIGPVNATSHKIDECIAVDALPKLSAIYRRILEQLMTA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01125", "category": "Mutation", "instruction": "Describe the effect of mutation Q15038-A153K on protein DAZ-associated protein 2.", "input": "### Wild-type:\nMNSKGQYPTQPTYPVQPPGNPVYPQTLHLPQAPPYTDAPPAYSELYRPSFVHPGAATVPTMSAAFPGASLYLPMAQSVAVGPLGSTIPMAYYPVGPIYPPGSTVLVEGGYDAGARFGAGATAGNIPPPPPGCPPNAAQLAVMQGANVLVTQRAGNFFMGGSDGGYTIW\n\n### Mutant:\nMNSKGQYPTQPTYPVQPPGNPVYPQTLHLPQAPPYTDAPPAYSELYRPSFVHPGAATVPTMSAAFPGASLYLPMAQSVAVGPLGSTIPMAYYPVGPIYPPGSTVLVEGGYDAGARFGAGATAGNIPPPPPGCPPNAAQLAVMQGANVLVTQRKGNFFMGGSDGGYTIW\n\n### Answer:", "output": "Enables ubiquitin-mediated degradation. Does not affect nuclear localization. Mutation of the K153 site in PRTB was susceptible to its degradation, proteolysis was suggested to be mediated by ubiquitination of K153.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01126", "category": "Mutation", "instruction": "Describe the effect of mutation P22185-Y253F on protein Ubiquinol oxidase 1, mitochondrial.", "input": "### Wild-type:\nMMSSRLVGTALCRQLSHVPVPQYLPALRPTADTASSLLHGCSAAAPAQRAGLWPPSWFSPPRHASTLSAPAQDGGKEKAAGTAGKVPPGEDGGAEKEAVVSYWAVPPSKVSKEDGSEWRWTCFRPWETYQADLSIDLHKHHVPTTILDKLALRTVKALRWPTDIFFQRRYACRAMMLETVAAVPGMVGGVLLHLKSLRRFEHSGGWIRALLEEAENERMHLMTFMEVAQPRWYERALVLAVQGVFFNAYFLGYLLSPKFAHRVVGYLEEEAIHSYTEFLKDIDSGAIQDCPAPAIALDYW\n\n### Mutant:\nMMSSRLVGTALCRQLSHVPVPQYLPALRPTADTASSLLHGCSAAAPAQRAGLWPPSWFSPPRHASTLSAPAQDGGKEKAAGTAGKVPPGEDGGAEKEAVVSYWAVPPSKVSKEDGSEWRWTCFRPWETYQADLSIDLHKHHVPTTILDKLALRTVKALRWPTDIFFQRRYACRAMMLETVAAVPGMVGGVLLHLKSLRRFEHSGGWIRALLEEAENERMHLMTFMEVAQPRWYERALVLAVQGVFFNAYFLGFLLSPKFAHRVVGYLEEEAIHSYTEFLKDIDSGAIQDCPAPAIALDYW\n\n### Answer:", "output": "No effect. Mutation of Y-253 to phenylalanine in the AOX protein maintains AOX activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01127", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSTLFEALYDYEARTEDDLSFHKGEKFQILNSSEGDWWEARSLTTGETGYIPSIYLAPV\n\n### Answer:", "output": "<SEQ_3Di>DFKKFFCAFDDDDDPQDDGDGGGWIWDFDDDPPPQWTWIATPVPRDTGTDGNVRIDTD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01128", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "S100A4, IL7R, CD3D, CD45RO, CD4", "output": "The marker profile suggests functional enrichment consistent with the typical biology of memory CD4+ T cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "ovary", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["S100A4", "IL7R", "CD3D", "CD45RO", "CD4"]}}
{"id": "eval_01129", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nTAGLN CNN1 ACTA2 MYH11\n</cell>", "output": "This cell is most consistent with smooth muscle cell, a smooth muscle cell. Key markers such as MYH11, CNN1, TAGLN support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "smooth muscle cell", "tissue": "skin", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TAGLN", "CNN1", "ACTA2", "MYH11"]}}
{"id": "eval_01130", "category": "Mutation", "instruction": "Describe the effect of mutation Q8TC94-L380V on protein Actin-like protein 9.", "input": "### Wild-type:\nMDASRPKSSESQSSLEAPRPGPNPSPNVVNKPLQRDSPGMVADRLPPKTGAVVIDMGTGTCKVGFAGQASPTYTVATILGCQPKKPATSGQSGLQTFIGEAARVLPELTLVQPLRSGIVVDWDAAELIWRHLLEHDLRVATHDHPLLFSDPPFSPATNREKLVEVAFESLRSPAMYVASQSVLSVYAHGRVSGLVVDTGHGVTYTVPVFQGYNLLHATERLDLAGNNLTAFLAEMLLQAGLPLGQQDLDLVENIKHHYCYVASDFQKEQARPEQEYKRTLKLPDGRTVTLGKELFQCPEL\n\n### Mutant:\nMDASRPKSSESQSSLEAPRPGPNPSPNVVNKPLQRDSPGMVADRLPPKTGAVVIDMGTGTCKVGFAGQASPTYTVATILGCQPKKPATSGQSGLQTFIGEAARVLPELTLVQPLRSGIVVDWDAAELIWRHLLEHDLRVATHDHPLLFSDPPFSPATNREKLVEVAFESLRSPAMYVASQSVLSVYAHGRVSGLVVDTGHGVTYTVPVFQGYNLLHATERLDLAGNNLTAFLAEMLLQAGLPLGQQDLDLVENIKHHYCYVASDFQKEQARPEQEYKRTLKLPDGRTVTLGKELFQCPEL\n\n### Answer:", "output": "Reduces interaction with ACTL7A. The mutant sperm with the mutation in ACTL9 led to abnormal ultrastructure of the perinuclear theca (PT), contributing to successful normal calcium oscillations in oocytes and subsequent total fertilization success (TFS).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01131", "category": "Mutation", "instruction": "Describe the effect of mutation Q8N653-R170Q on protein Leucine-zipper-like transcriptional regulator 1.", "input": "### Wild-type:\nMAGPGSTGGQIGAAALAGGARSKVAPSVDFDHSCSDSVEYLTLNFGPFETVHRWRRLPPCDEFVGARRSKHTVVAYKDAIYVFGGDNGKTMLNDLLRFDVKDCSWCRAFTTGTPPAPRYHHSAVVYGSSMFVFGGYTGDIYSNSNLKNKNDLFEYKFATGQWTEWKIEGRLPVARSAHGATVYSDKLWIFAGYDGNARLNDMWTIGLQDRELTCWEEVAQSGEIPPSCCNFPVAVCRDKMFVFSGQSGAKITNNLFQFEFKDKTWTRIPTEHLLRGSPPPPQRRYGHTMVAFDRHLYVFG\n\n### Mutant:\nMAGPGSTGGQIGAAALAGGARSKVAPSVDFDHSCSDSVEYLTLNFGPFETVHRWRRLPPCDEFVGARRSKHTVVAYKDAIYVFGGDNGKTMLNDLLRFDVKDCSWCRAFTTGTPPAPRYHHSAVVYGSSMFVFGGYTGDIYSNSNLKNKNDLFEYKFATGQWTEWKIEGQLPVARSAHGATVYSDKLWIFAGYDGNARLNDMWTIGLQDRELTCWEEVAQSGEIPPSCCNFPVAVCRDKMFVFSGQSGAKITNNLFQFEFKDKTWTRIPTEHLLRGSPPPPQRRYGHTMVAFDRHLYVFG\n\n### Answer:", "output": "In SWNTS2; increased RAS-MAPK signaling.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01132", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSTU1_CRYNB\n\n### Answer:", "output": "Protein STU1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01133", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "IAPP, NKX6-1, MAFA, INS", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as INS, IAPP, NKX6-1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "heart", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IAPP", "NKX6-1", "MAFA", "INS"]}}
{"id": "eval_01134", "category": "Mutation", "instruction": "Describe the effect of mutation P13688-D457T on protein Carcinoembryonic antigen-related cell adhesion molecule 1.", "input": "### Wild-type:\nMGHLSAPLHRVRVPWQGLLLTASLLTFWNPPTTAQLTTESMPFNVAEGKEVLLLVHNLPQQLFGYSWYKGERVDGNRQIVGYAIGTQQATPGPANSGRETIYPNASLLIQNVTQNDTGFYTLQVIKSDLVNEEATGQFHVYPELPKPSISSNNSNPVEDKDAVAFTCEPETQDTTYLWWINNQSLPVSPRLQLSNGNRTLTLLSVTRNDTGPYECEIQNPVSANRSDPVTLNVTYGPDTPTISPSDTYYRPGANLSLSCYAASNPPAQYSWLINGTFQQSTQELFIPNITVNNSGSYTCH\n\n### Mutant:\nMGHLSAPLHRVRVPWQGLLLTASLLTFWNPPTTAQLTTESMPFNVAEGKEVLLLVHNLPQQLFGYSWYKGERVDGNRQIVGYAIGTQQATPGPANSGRETIYPNASLLIQNVTQNDTGFYTLQVIKSDLVNEEATGQFHVYPELPKPSISSNNSNPVEDKDAVAFTCEPETQDTTYLWWINNQSLPVSPRLQLSNGNRTLTLLSVTRNDTGPYECEIQNPVSANRSDPVTLNVTYGPDTPTISPSDTYYRPGANLSLSCYAASNPPAQYSWLINGTFQQSTQELFIPNITVNNSGSYTCH\n\n### Answer:", "output": "Increases the binding to ANXA2. The functional changes caused by the mutation in CEACAM1 are not mentioned in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01135", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSYFB_FRATT\n\n### Answer:", "output": "Belongs to the phenylalanyl-tRNA synthetase beta subunit family. Type 1 subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01136", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMFNSVLDTIGNTPLIRLSKASELTGCDIYGKAEFLNPGQSVKDRAALYIIRDAEKRGLLRPGGVIVEGTAGNTGIGLTMVAKALGYRTAIVIPETQSQEKKDALRLLGAELIEVPAAPYRNPNNYVRLSGRLAEQLAKTEPNGAIWANQFDNTVNRQAHIETTAQEIWRDTSDQIDGFVAAVGSGGTLAGTAIGLKERNHNIKIALADPHGAALHAFYTTGELKAEGDSITEGIGQGRITANLEGFTPDFSYQIPDAEALDILFALVEEEGLCLGGSSGINIAGAIRLAKDLGPGHTIVTVLCDYGNRYQSKLFNPAFLRGKSLPVPRWLEKKTEIDIPFEGLEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CBSSGGGGSSCCCEEECHHHHHHHTSEEEEEEGGGSTTSBTHHHHHHHHHHHHHHTTCCCTTCEEEEECSSHHHHHHHHHHHHHTCEEEEEEETTSCHHHHHHHHHHTCEEEEECCCCTTSTTSHHHHHHHHHHHHHTTCTTCEEECCTTTCTHHHHHHHHTHHHHHHHHTTTCCCEEEECCSSSHHHHHHHHHHHHHCTTCEEEEEEETTCSHHHHHHHSSCCCCSCCSCCSCCCCSCCGGGTTCCCSEEEEECHHHHHHHHHHHHHHHSCCBCHHHHHHHHHHHHHHHHHCSSCEEEEEECBBGGGGHHHHTCHHHHHHTTCCCCTTCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01137", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y223-E576G on protein Bifunctional UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase.", "input": "### Wild-type:\nMEKNGNNRKLRVCVATCNRADYSKLAPIMFGIKTEPEFFELDVVVLGSHLIDDYGNTYRMIEQDDFDINTRLHTIVRGEDEAAMVESVGLALVKLPDVLNRLKPDIMIVHGDRFDALALATSAALMNIRILHIEGGEVSGTIDDSIRHAITKLAHYHVCCTRSAEQHLISMCEDHDRILLAGCPSYDKLLSAKNKDYMSIIRMWLGDDVKSKDYIVALQHPVTTDIKHSIKMFELTLDALISFNKRTLVLFPNIDAGSKEMVRVMRKKGIEHHPNFRAVKHVPFDQFIQLVAHAGCMIGN\n\n### Mutant:\nMEKNGNNRKLRVCVATCNRADYSKLAPIMFGIKTEPEFFELDVVVLGSHLIDDYGNTYRMIEQDDFDINTRLHTIVRGEDEAAMVESVGLALVKLPDVLNRLKPDIMIVHGDRFDALALATSAALMNIRILHIEGGEVSGTIDDSIRHAITKLAHYHVCCTRSAEQHLISMCEDHDRILLAGCPSYDKLLSAKNKDYMSIIRMWLGDDVKSKDYIVALQHPVTTDIKHSIKMFELTLDALISFNKRTLVLFPNIDAGSKEMVRVMRKKGIEHHPNFRAVKHVPFDQFIQLVAHAGCMIGN\n\n### Answer:", "output": "Highly increases epimerase activity; highly increases kinase activity. The mutation in the GNE gene causes more than 80% augmentation in both activities of GNE. Structural changes were observed in gel filtration analysis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01138", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nILQESVLNKYRTAGQIAQTALKYVTSLINDSYHSKTTQRQLTVPELCLLTDSFILTRLEQYYKNKVNERGIAIPTTIDIDQISGGWCPEIDDTQNLLNWNKGKDSTFASSVTGTLRPGDLVKITLGVHIDGYTSEVSHTMVIYPVDETKPILQPTGPLLGGKADAVAAAHIAMETVVALLACALTPEKLPASLGGTSSGITGQLIRTIVDTIARSYNCGVVPGSRVRRIRRFLAGQNEGIVAEREYKGVVWTESHQEADLLSAIPSDDFVVQSGEVYLIDLKMASLEHCTKKGLVTLETVDSYTGKSHKAGELIARPGAYVRDFAQTHILKLKTSRQLLTKIDKQGVYPFKLSHLSSNFPFVHENEEELQSLKKDLKSFRLGMSEISNNYLCVESPIQIARWVPWDHILKATNPLPLPKLGVSAIKLKSLMNSTKESISLPVARECNTIVLCDSSVSTTDRPELLRLTGGSKTCQPSWIHSQHELNPQDSIVQGIFQLATLAKDLLLKETQPMKQK\n\n### Answer:", "output": "<SEQ_3Di>DDDVVVVVLPLVQLLLVLQLVVVVCCQPCVVFLDPDPVDHAFQQNSQQVSQLSVVVPPCPPPPPVFPDWGWFAGKAKDWFWWDGQHPAHPVFLVVLCVLLVQQPDPCCCPSVRPDDFFTFIKIKTWIHTPLFIFIWIFTAGRHAWDDVDPVSDRDFADEDQCLLQVVLAPVLVCLVCLLVFCLQDQDLAQCLQPHRPARDWLVSNVVVNVVSCVLSVWDWEQQWWKWFADGNDRCQQVPVLPDDDTLIDGDALACVVVPPVCSPDRGPRRDHFFTKIKGKTKTWRDDPDPFAATWYKAFADADSPPDDDSHDHHGHGFKKFFDPPDDDDDPDDLLVVVPVLSVVHDGRIDGVLRSDPLSNCPPDDPCSVVVVVVVVVSVVVNVVVCVVVVRMDGITIITTWGDGHCLLPVDDDCPSHDDSPAYVVNVNVSLVPDDPGGGTGMTMIMWMKHFAAPVHDDVDHTGIEISQPDDVRRPHDHRDYPSDPDDPPVCVVSVVVSVVCVVPVDHHYSDDDHDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01139", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nSYGB_SHIF8\n\n### Answer:", "output": "Cytoplasm.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01140", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSIAQARKLVEQLKMEANIDRIKVSKAAADLMAYCEAHAKEDPLLTPVPASENPFRE\n\n### Answer:", "output": "<SEQ_3Di>DVVVVVVVVVVCVVVVPDDDDDVVVVVVVVVVVCVVCVVPDCVNVPDPLVPPPVRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01141", "category": "Mutation", "instruction": "Describe the effect of mutation Q8VZ20-S182A on protein Trihelix transcription factor ASR3.", "input": "### Wild-type:\nMALEQLGLGVSAVDGGENSSAPSNDGGDDGVKTARLPRWTRQEILVLIQGKRVAENRVRRGRAAGMALGSGQMEPKWASVSSYCKRHGVNRGPVQCRKRWSNLAGDYKKIKEWESQIKEETESYWVMRNDVRREKKLPGFFDKEVYDIVDGGVIPPAVPVLSLGLAPASDEGLLSDLDRRESPEKLNSTPVAKSVTDVIDKEKQEACVADQGRVKEKQPEAANVEGGSTSQEERKRKRTSFGEKEEEEEEGETKKMQNQLIEILERNGQLLAAQLEVQNLNLKLDREQRKDHGDSLVAVL\n\n### Mutant:\nMALEQLGLGVSAVDGGENSSAPSNDGGDDGVKTARLPRWTRQEILVLIQGKRVAENRVRRGRAAGMALGSGQMEPKWASVSSYCKRHGVNRGPVQCRKRWSNLAGDYKKIKEWESQIKEETESYWVMRNDVRREKKLPGFFDKEVYDIVDGGVIPPAVPVLSLGLAPASDEGLLSDLDRREAPEKLNSTPVAKSVTDVIDKEKQEACVADQGRVKEKQPEAANVEGGSTSQEERKRKRTSFGEKEEEEEEGETKKMQNQLIEILERNGQLLAAQLEVQNLNLKLDREQRKDHGDSLVAVL\n\n### Answer:", "output": "Normal flg22-induced phosphorylation by MPK4.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01142", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSDTEAPVEVQEDFEVVEEFTPVVLATPIPEEVQQAQTEIKLFNKWSFEEVEVKDASLVDYVQVRQPIFVAHTAGRYANKRFRKAQCPIIERLTNSLMMNGRNNGKKLKAVRIIKHTLDIINVLTDQNPIQVVVDAITNTGPREDTTRVGGGGAARRQAVDVSPLRRVNQAIALLTIGAREAAFRNIKTIAETLAEELINAAKGSSTSYAIKKKDELERVAKSNR\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCSSCCCHHHHHHHHSSCCSSSCCCSSCCCSSTTTBTTBCCCSCCSSSCCCCCCSSSTTGGGGSCHHHHHHHTTSCSGGGSSCHHHHHHHHHHHHHHHHHTTCSCHHHHHHHHHHTTSCSEEEEECCSSSSCCEEEEECCHHHHHHHHHHHHHHHHHHHHTTSSSCHHHHHHHHHHHHHTTCSCSHHHHHHHHHHHHHHTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01143", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD1C CLEC10A FCER1A HLA-DQA1\n</cell>", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as CLEC10A, FCER1A, HLA-DQA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "blood", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD1C", "CLEC10A", "FCER1A", "HLA-DQA1"]}}
{"id": "eval_01144", "category": "Mutation", "instruction": "Describe the effect of mutation P9WPF1-C205F on protein Alpha-pyrone synthesis polyketide synthase-like Pks18.", "input": "### Wild-type:\nMNVSAESGAPRRAGQRHEVGLAQLPPAPPTTVAVIEGLATGTPRRVVNQSDAADRVAELFLDPGQRERIPRVYQKSRITTRRMAVDPLDAKFDVFRREPATIRDRMHLFYEHAVPLAVDVSKRALAGLPYRAAEIGLLVLATSTGFIAPGVDVAIVKELGLSPSISRVVVNFMGCAAAMNALGTATNYVRAHPAMKALVVCIELCSVNAVFADDINDVVIHSLFGDGCAALVIGASQVQEKLEPGKVVVRSSFSQLLDNTEDGIVLGVNHNGITCELSENLPGYIFSGVAPVVTEMLWDN\n\n### Mutant:\nMNVSAESGAPRRAGQRHEVGLAQLPPAPPTTVAVIEGLATGTPRRVVNQSDAADRVAELFLDPGQRERIPRVYQKSRITTRRMAVDPLDAKFDVFRREPATIRDRMHLFYEHAVPLAVDVSKRALAGLPYRAAEIGLLVLATSTGFIAPGVDVAIVKELGLSPSISRVVVNFMGCAAAMNALGTATNYVRAHPAMKALVVCIELFSVNAVFADDINDVVIHSLFGDGCAALVIGASQVQEKLEPGKVVVRSSFSQLLDNTEDGIVLGVNHNGITCELSENLPGYIFSGVAPVVTEMLWDN\n\n### Answer:", "output": "Efficiently catalyzed the synthesis of the triketide pyrone of the C6 starter unit and shows weak polyketide synthase activity with the C12 starter molecule. This mutation may result in functional changes in the enzyme, potentially affecting its substrate specificity or catalytic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01145", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAKSPARRCTAKVRRVLSRSVLILCWSLLGAAPAHADDSRLGWPLRPPPAVVRQFDAASPNWNPGHRGVDLAGRPGQPVYAAGSATVVFAGLLAGRPVVSLAHPGGLRTSYEPVVAQVRVGQPVSAPTVIGALAAGHPGCQAAACLHWGAMWGPASGANYVDPLGLLKSTPIRLKPLSSEGRTLHYRQAEPVFVNEAAAGALAGAGHRKSPKQGVFRGAAQGGDIVARQPPGRWVCPSSAGGPIGWHRQ\n\n### Sequence 2:\nMTLEITAAMVKDLREKTNVGMMDCKKALQETGGDLEKAVDLLRQKGLAKAMKRAGKEASEGMVHAYIHAGGRIGVLIEVNCETDFAAKSEDFVEFVKNVAMQVAATNPLGIVPEDISQDVVERERAIYLAQAQESGKPQNILEKMVEGKMRKFFEESTLLQQSYVKDPDKTIQDYLNELTASIGEKIIIRRFARFQLGSE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01146", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UQL6-S259A on protein Histone deacetylase 5.", "input": "### Wild-type:\nMNSPNESDGMSGREPSLEILPRTSLHSIPVTVEVKPVLPRAMPSSMGGGGGGSPSPVELRGALVGSVDPTLREQQLQQELLALKQQQQLQKQLLFAEFQKQHDHLTRQHEVQLQKHLKQQQEMLAAKQQQEMLAAKRQQELEQQRQREQQRQEELEKQRLEQQLLILRNKEKSKESAIASTEVKLRLQEFLLSKSKEPTPGGLNHSLPQHPKCWGAHHASLDQSSPPQSGPPGTPPSYKLPLPGPYDSRDDFPLRKTASEPNLKVRSRLKQKVAERRSSPLLRRKDGTVISTFKKRAVEI\n\n### Mutant:\nMNSPNESDGMSGREPSLEILPRTSLHSIPVTVEVKPVLPRAMPSSMGGGGGGSPSPVELRGALVGSVDPTLREQQLQQELLALKQQQQLQKQLLFAEFQKQHDHLTRQHEVQLQKHLKQQQEMLAAKQQQEMLAAKRQQELEQQRQREQQRQEELEKQRLEQQLLILRNKEKSKESAIASTEVKLRLQEFLLSKSKEPTPGGLNHSLPQHPKCWGAHHASLDQSSPPQSGPPGTPPSYKLPLPGPYDSRDDFPLRKTAAEPNLKVRSRLKQKVAERRSSPLLRRKDGTVISTFKKRAVEI\n\n### Answer:", "output": "Reduces AMPK- and caMK-dependent phosphorylation and the subsequent nuclear export. Abolishes nuclear export; when associated with A-498. Does not affect phosphorylation by PKN1 and PKN2. Little is known of other mechanisms for regulating the subcellular distribution of class IIa HDACs.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01147", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTHSA_THEK1\n\n### Answer:", "output": "Thermosome subunit alpha, Chaperonin subunit alpha, Thermosome subunit 1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01148", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C(CCCCl)c1ccc(F)cc1", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains aromatic ring, ketone, halide. Its descriptor profile: MW=200.64, LogP=3.03, TPSA=17.07, HBD=0, HBA=1, QED=0.54.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=C(CCCCl)c1ccc(F)cc1", "compound_name": null, "pubchem_cid": null, "MW": 200.64, "LogP": 3.03, "TPSA": 17.07, "QED": 0.54, "functional_groups": ["aromatic ring", "ketone", "halide"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": 0.0, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01149", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTBP_METS3\n\n### Answer:", "output": "Belongs to the TBP family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01150", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMMKKIDVKILDPRVGKEFPLPTYATSGSAGLDLRACLDDAVELAPGDTTLVPTGLAIHIADPSLAAMMLPRSGLGHKHGIVLGNLVGLIDSDYQGQLMISVWNRGQDSFTIQPGERIAQMIFVPVVQAEFNLVEDFDATDRGEGGFGHSGRQ\n\n### Sequence 2:\nMVRAEIETGAPFRSVKEAVTLFGERILLGDNYISKSVERSSCKSIQDELVEAKENLKKAEEENKVLSQLIESLTQELETTKEKLNHSLRNFPEHPQVEDDLKFIEESTVNEPDNITEIKMNRFDRNEVYGDRLEKRRSVKFANPPLLTKVIVGKEEKNQVMVKKQTKKMKPLVPLAAWLFARNRSS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01151", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMVNMKASMFLTFAGLVLLFVVCYASESEEKEFPKEMLSSIFAVDNDSKQEERDCVGYMRECKEKLCCSGYVCSSRWKWCVLPAPWRR\n\n### Sequence 2:\nMVNMKASMFLTFAGLVLLFVVCYASESEEKEFPKEMLSSIFAVDNDSKQEERDCVGYMRECKEKLCCSGYVCSSRWKWCVLPAPWRR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01152", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nTGTYRGDSETQLERVNVYYNEASCGRYVPRAVLMDLEPGTMDSVRSGPYGQIFRPDNFVFGQSGAGNNWAKGHYTEGAELIDSVLDVVRKEAENCDCLQGFQVCHSLGGGTGSGMGTLLISKIREEYPDRMMXTFSVFPSPKVSDTVVEPYNATLSVHQLVENADECMVLDNEALYDICFRTLKLVTPTFGDLNHLISATMSGVTCCLRFPGQLNSDLRKLAVNLIPFPRLHFFMVGFAPLTSRGSQQYRALTVPELTQQMRDAKNMMCAADPRHGRYLTASAMFRGKMSTKEVDEQMINVQNKNSSYFVEWIPNNVKSSVCDIPPVGLKMACTFIGNSTSIQEMFRRVRDQFTAMFRXKAFLHWYTGEGMDEMEFTEAESNMNDLVSEYQQYQDATAEPEGXYEEDYDEA\n\n### Sequence 2:\nMTISVFDLFTIGIGPSSSHTVGPMRAANQFVVALRRRGHLDDLEAMRVDLFGSLAATGAGHGTMSAILLGLEGCQPETITTEHKERRLAEIAASGVTRIGGVIPVPLTERDIDLHPDIVLPTHPNGMTFTAAGPHGRVLATETYFSVGGGFIVTEQTSGNSGQHPCSVALPYVSAQELLDICDRLDVSISEAALRNETCCRTENEVRAALLHLRDVMVECEQRSIAREGLLPGGLRVRRRAKVWYDRLNAEDPTRKPEFAEDWVNLVALAVNEENASGGRVVTAPTNGAAGIVPAVLHYAIHYTSAGAGDPDDVTVRFLLTAGAIGSLFKERASISGAEVGCQGEVGSAAAMAAAGLAEILGGTPRQVENAAEIAMEHSLGLTCDPIAGLVQIPCIERNAISAGKAINAARMALRGDGIHRVTLDQVIDTMRATGADMHTKYKETSAGGLAINVAVNIVEC\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01153", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMGTSELLKQIYDINLSYLLLAQRIINQEKVSAMFRLGIDEAMADALARLTLPEMVKLAETNQLVCQFRFDDHQSISRLTRESRVEDLQQIHTGILLSSRLLRNVTKEQETPKKRAAS\n\n### Sequence 2:\nMGTSELLKHIYDINLSYLLLAQRLINDEKASAMFRLGIDETMADALAQLTLPQMVKLAETNQLVCHFRFNESQTIERLTKESRVDDLQQIHTGILLSSHLLQELSSKDASPTKKRA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01154", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nISPG_SALAR\n\n### Answer:", "output": "4-hydroxy-3-methylbut-2-en-1-yl diphosphate synthase (flavodoxin), 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01155", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLGEVSKILAKSSMAIAFTGAGISAESGIPTFRGKDGLWRKYRAEELATPEAFKRDPKLVWEFYKWRIKKILEAKPNPAHIALAELEKMGIIKAVITQNVDDLHREAGSKNVIELHGNIFRVKCTSCSYREYLKESDRIGWLLSQELPRCPKCGSLLRPDVVWFGEALPEKELTTAFSLAKKADVVLVVGTSGVVYPAAYIPYIVKESGGIVVEINIEPSAITPIADFFLRGKAGEVLPKLVEEIRRISK\n\n### Sequence 2:\nMIEEAPRIIAHSRFLIAFTGAGVSAESGIPTFRDRGGLWENYRIEEVATPEAFRKDPNLVWSFYKMRMKIMKGAKPNNAHLALAELEKMGILKAVITQNIDNLHREAGNQHIVELHGNIYRVKCTRCDYMENLLESGKLEDFLKEKNLPKCPECASLLRPDVVWFGEPLPQEALQKAFKLAERADVCLVVGTSAQVFPAAYVPYIVKDNGGSVIEINTKESGITPIADVFIRGKAGEVMQSLLVKVKR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01156", "category": "Mutation", "instruction": "Describe the effect of mutation P36017-L66Q on protein Vacuolar protein sorting-associated protein 21.", "input": "### Wild-type:\nMNTSVTSIKLVLLGEAAVGKSSIVLRFVSNDFAENKEPTIGAAFLTQRVTINEHTVKFEIWDTAGLERFASLAPMYYRNAQAALVVYDVTKPQSFIKARHWVKELHEQASKDIIIALVGNKIDMLQEGGERKVAREEGEKLAEEKGLLFFETSAKTGENVNDVFLGIGEKIPLKTAEEQNSASNERESNNQRVDLNAANDGTSANSACSC\n\n### Mutant:\nMNTSVTSIKLVLLGEAAVGKSSIVLRFVSNDFAENKEPTIGAAFLTQRVTINEHTVKFEIWDTAGQERFASLAPMYYRNAQAALVVYDVTKPQSFIKARHWVKELHEQASKDIIIALVGNKIDMLQEGGERKVAREEGEKLAEEKGLLFFETSAKTGENVNDVFLGIGEKIPLKTAEEQNSASNERESNNQRVDLNAANDGTSANSACSC\n\n### Answer:", "output": "Enables GTP hydrolysis.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01157", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEEVTFFEKAKRYIGNKHLYTEFLKILNLYSQDILDLDDLVEKVDFYLGSNKELFTWFKNFVGYQEAFGPSYKRLPKSDTFMPCSGRDDMCWEVLNDEWVGHPVWASEDSGFIAHRKNQYEETLFKIEEERHEYDFYIESNLRTIQCLETIVNKSMTIYKKVIRKVYDKERGFEIIDALHETAPVVLKRLKQKDEEWRRAQREWNKVWRELEQKVFFKSLDHLGLTFKQADKKLLTTKQLISEISSIKVDQTNKKSQLDFDFPDKNIFYDILCLADTFITHTTAYSNPDKERLKDLLKYFISLFFSISFEKIEESSIFNLFANTNIYIFFRHWTTIYERLLEIKQMNERVTKEINTRSTVTFAKDLDLLSSQLSEMGLDFVGEDAYKQVLRLSRRLINGDLEHQWFEESLRQAYNNKAFKLYTIDKVTQSLVKHAHTLMTDAKTAEIMALFVKDRNASTTSAKDQIIYRLQVRSHMSNTENMFRIEFDKRTLHVSIQYIALDDL\n\n### Answer:", "output": "<SEQ_3Di>DVVCLLVVLVVVQPDDVVSVVLVVLVVCVVVVNDDVVVNLVVVCVRCVVPDPSSVVVCVVVVVVPDDDPFKDFDDPVVQDDDDPPADPVNVVPDDRTIDGPCCVVVVVPDDDDDDADVLVVLVVVLVVVLVVLVVLLVLLVVLLVVLCVVPVVPPCSNLVNQCVLDVVVCSVVVVVVVVVCSVVVNVSSVVVNVVSVVVVVVSVVVSLVVNVPSVLVRPADCLVPVVVVCVVCLDLVNVVVQLVVQVVVVVVVQFSDKDFDPDDVLLLLLLVLLLLLLVQDPPDDPVVSVVLSVLSLVVVCLQQVPDNPVSVVVLKAKAKDASLVVSLVVLVVVSSVLLVVVVVCQVVLLCCLQPPDDDDPPCVPPRDDCPCVVVVLSPHPCRPSNVLSVVSSVVSNPVDDLVSNVVSCCVSVVNPNSVCSCVNVSSVVSSVSSVVCPPDPLSVVSVVLSSVRSVVSDDDPCCVQVSVVVNVVSPDQQAWMKMWMADNVRGMIGIHTGGRPSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01158", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "TCL1A, IGHM, IGHD, MS4A1", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as IGHM, MS4A1, IGHD support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "intestine", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TCL1A", "IGHM", "IGHD", "MS4A1"]}}
{"id": "eval_01159", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMTDGLNQLYNLVAQDYALTDTIEALSRMLHRGTIPLDTFVKQGRELARQQFLVRWHIQRITSPLS\n\n### Answer:", "output": "<SEQ_3Di>DDPLVVVLVVLVVLLVVLVVVLVVLVVCVVVVVDDPVVNVVVVVVSVVVNVVSVVVNCVSCVPVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01160", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKLHELKPAEGSRQVRNRVGRGTSSGNGKTAGRGQKGQKARSGGGVRLGFEGGQTPLFRRLPKRGFTNINRKDYAVVNLDTLNRFEDGTEVTPVVLKEAGIVKNEKAGIKVLADGELTKKLTVKAAKFSKSAQEAIEAAGGSIEVI\n\n### Sequence 2:\nMQLPRVTLRQMTWTTSAIVLLGITLLLWSAFRHQESTLAIRAINQGTTMPDGFSVWHHLDANGIRFKSITPKNDTLLITFDSSAQSAAAKVVLDRTFPHGYIIAQQDEDNQASQWLSLLRDTSHRFG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01161", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSARAPKELRLALPPCLLNRTFASPNASGSGNTGARGPGAGGSGTCITQVGQQLFQSFSSTLVLIVLVTLIFCLIVLSLSTFHIHKRRMKKRKMQRAQEEYERDHCSGSRGGGGLPRPGRQAPTHTKETRLERQPRDSPFCAPSNASSSSSSSPGLLCQGPCAPPPPPPASSPQGAHAASSCLDTAGEGLLQTVVLS\n\n### Sequence 2:\nMPTFDVVSEVDKHELQNAIDQVNREIGTRFDFRGTEAHIEKSEEELILVAESEFQLQQMRTILDTKLAKRGVDVDCLEAKEPEIIGKRARQSIQVRQGIDKDTARKIIKIIKESKLKVQAAIQGEQVRISGKKRDDLQQVIALLREADLDLPLQYINFRD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01162", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nQSEPELKLESVVIVSRHGVRAPTKATQLMQDVTPDAWPTWPVKLGWLTPRGGELIAYLGHYQRQRLVADGLLAKKGCPQSGQVAIIADTDERTRKTGEAFAAGLAPDCAITVHTQVDTSSPDPLFNPLKTGVCQLDNANVTDAILSRAGGSIADFTGHRQTAFRELERVLNFPQSNLCLKREKQDESCSLTQALPSELKVSADNVALTGAVSLASMLTEIFLLQQAQGMPEPGWGRITDSHQWNTLLSLHNAQFYLLQRTPEVARSRATPLLDLIKTALTPHPPQKQAYGVTLPTSVLFIAGHDTNLANLGGALELNWTLPGQPDNTPPGGELVFERWRRLSDNSQWIQVSLVFQTLQQMRDKTPLSLNTPPGEVKLTLAGCEERNAQGMCSLAGFTQIVNEARIPACSL\n\n### Answer:", "output": "<SEQ_2D>CCCCCCEEEEEEEEEECCCBCCSCCCHHHHHTCSSCCCCCSSSTTCBCHHHHHHHHHHHHHHHHHHHHTTSSCSSSCCCTTTEEEEECSSHHHHHHHHHHHHHHSTTSCCCEECCSCTTSCCTTTCTTTTTSSCCCHHHHHHHHHHHTTSSHHHHHHTTHHHHHHHHHHHTGGGSHHHHHCCCCCCCCCHHHHSCCCEEECSSCEEECHHHHHHHHHHHHHHHHHHTTCSSGGGGTCCSHHHHHHHHHHHHHHHHHHTTSHHHHHHHTHHHHHHHHHHHSCCCCEECGGGCEESCSEEEEEECHHHHHHHHHHHTCCCCCTTCSSSSCTTCEEEEEEEEETTTTEEEEEEEEEECCHHHHHTTCCCCSSSCCEEEECCCTTCSSBCTTSCEEHHHHHHHHHHHSCGGGCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01163", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nIGBP1_RAT\n\n### Answer:", "output": "Immunoglobulin-binding protein 1, Alpha4 phosphoprotein, CD79a-binding protein 1, Protein phosphatase 2/4/6 regulatory subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01164", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMFSNRLPPPKHSQGRVSTALSSDRVEPAILTDQIAKNVKLDDFIPKRQSNFELSVPLPTKAEIQECTARTKSYIQRLVNAKLANSNNRASSRYVTETHQAPANLLLNNSHHIEVVSKQMDPLLPRFVGKKARKVVAPTENDEVVPVLHMDGSNDRGEADPNEWKIPAAVSNWKNPNGYTVALERRVGKALDNENNTINDGFMKLSEALENADKKARQEIRSKMELKRLAMEQEMLAKESKLKELSQRARYHNGTPQTGAIVKPKKQTSTVARLKELAYSQGRDVSEKIILGAAKRSEQPDLQYDSRFFTRGANASAKRHEDQVYDNPLFVQQDIESIYKTNYEKLDEAVNVKSEGASGSHGPIQFTKAESDDKSDNYGA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHTTCCGGGTSCTTSSCTTCCCCCCTTTTHHHHHHHHHHHHHHHHHHHHHHSSCTTTGGGCCCCCCCCCCCCTTSTTCCEEEECCCCTTSCSSTTCCCCCCCCCCCCSCCCCCCCCCCCCCCCCSHHHHHCCCCCCCSSCCTTCCCCCGGGTCCCCCCCCCCCCCTHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHSSSCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01165", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKREELLVEEIKDLTLEELKGYADFYKILDKVYGFTAESVVRGVKILKDMIKEADLRFLSFTANLVSTGLRGLFADLIKQGYFNVIITTGGTIDHDIARSFGGKYYKGLFEYDDSMLRELEIHRLGNILVPMESYGKVIEDVVRKYMNEIVSIKKEWPVYELLWEFGKRISDENSILKAAYEKKVPIIVPGIIDGSFGTNLFIYSQFTQLKLNLFEDMKLIKDLIFSCKKSGALIIGGGISKHHTIWWNQFKDGLDYAIYITTAQEYDGSLSGAKPREAISWNKIKPTSENVVIYGDATIILPILSASLLG\n\n### Sequence 2:\nMKKHKMLSLLAVSGLMGIGILAGCSNDSSSSSKGTINIVSREEGSGTRGAFIELFGIESKNKKGEKVDHTSDAATVTNSTSVMLTTVSKDPSAIGYSSLGSLNSSVKVLKIDGKNATVKDIKSGSYKISRPFNIVTKEGKEKEATKDFIDYILSKDGQAVVEKNGYIPLDNAKAYQAKVSSGKVVIAGSSSVTPVMEKIKEAYHKVNAKVDVEIQQSDSSTGITSAIDGSADIGMASRELDKTESSKGVKATVIATDGIAVVVNKKNKVNDLSTKQVKDIFTGKTTSWSDLSK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01166", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMETE_BURO1\n\n### Answer:", "output": "5-methyltetrahydropteroyltriglutamate--homocysteine methyltransferase, Cobalamin-independent methionine synthase, Methionine synthase, vitamin-B12 independent isozyme.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01167", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nUVRC_YERPA\n\n### Answer:", "output": "The UvrABC repair system catalyzes the recognition and processing of DNA lesions.  UvrC both incises the 5' and 3' sides of the lesion.  The N-terminal half is responsible for the 3' incision and the C-terminal half is responsible for the 5' incision.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01168", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "S100A4, CD4, IL7R, CD45RO, CD3D", "output": "The marker profile suggests functional enrichment consistent with the typical biology of memory CD4+ T cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "breast", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_pathway", "top_genes": ["S100A4", "CD4", "IL7R", "CD45RO", "CD3D"]}}
{"id": "eval_01169", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: SCCCCCCCCS\nDescriptors: MW=178.37, LogP=3.19, TPSA=0.0, HBD=2, HBA=2", "output": "Descriptors: MW=178.37, LogP=3.19, TPSA=0.0, HBD=2, HBA=2, QED=0.43, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "SCCCCCCCCS", "compound_name": null, "pubchem_cid": null, "MW": 178.37, "LogP": 3.19, "TPSA": 0.0, "QED": 0.43, "functional_groups": [], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01170", "category": "Mutation", "instruction": "Describe the effect of mutation O29918-K531D on protein Dolichyl-phosphooligosaccharide-protein glycotransferase 1.", "input": "### Wild-type:\nMDRKVLMLAVILFALAVRFQNFGEIFDSGIYYTGYDSYYHMRLVEVMVKESFRPDYDYYINYPFGLKITWPPLFDYILAFPGMLFGFHSSEIFAVFLPVILGVLSVVLICLTALQIVNNQTFALISAFIYAAAPVAVWKTVLGQADHHALVIFLFLLSAYLLLKDGVWKILAGLPMLFMALAWLGSPIYGALLAFSALVHFDRKALRLVAASYLIPAISFVLYPPVGISFFGLAAFLFVGSVVKGYEDRFRNATIYYIALSLATVLIIYFIPLPHFEFVKGGINYIFGANIYLPTISEAR\n\n### Mutant:\nMDRKVLMLAVILFALAVRFQNFGEIFDSGIYYTGYDSYYHMRLVEVMVKESFRPDYDYYINYPFGLKITWPPLFDYILAFPGMLFGFHSSEIFAVFLPVILGVLSVVLICLTALQIVNNQTFALISAFIYAAAPVAVWKTVLGQADHHALVIFLFLLSAYLLLKDGVWKILAGLPMLFMALAWLGSPIYGALLAFSALVHFDRKALRLVAASYLIPAISFVLYPPVGISFFGLAAFLFVGSVVKGYEDRFRNATIYYIALSLATVLIIYFIPLPHFEFVKGGINYIFGANIYLPTISEAR\n\n### Answer:", "output": "Significantly reduces catalytic activity. This mutation affects the S/T recognition in the sequon.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01171", "category": "Mutation", "instruction": "Describe the effect of mutation Q3TBT3-K337R on protein Stimulator of interferon genes protein.", "input": "### Wild-type:\nMPYSNLHPAIPRPRGHRSKYVALIFLVASLMILWVAKDPPNHTLKYLALHLASHELGLLLKNLCCLAEELCHVQSRYQGSYWKAVRACLGCPIHCMAMILLSSYFYFLQNTADIYLSWMFGLLVLYKSLSMLLGLQSLTPAEVSAVCEEKKLNVAHGLAWSYYIGYLRLILPGLQARIRMFNQLHNNMLSGAGSRRLYILFPLDCGVPDNLSVVDPNIRFRDMLPQQNIDRAGIKNRVYSNSVYEILENGQPAGVCILEYATPLQTLFAMSQDAKAGFSREDRLEQAKLFCRTLEEILED\n\n### Mutant:\nMPYSNLHPAIPRPRGHRSKYVALIFLVASLMILWVAKDPPNHTLKYLALHLASHELGLLLKNLCCLAEELCHVQSRYQGSYWKAVRACLGCPIHCMAMILLSSYFYFLQNTADIYLSWMFGLLVLYKSLSMLLGLQSLTPAEVSAVCEEKKLNVAHGLAWSYYIGYLRLILPGLQARIRMFNQLHNNMLSGAGSRRLYILFPLDCGVPDNLSVVDPNIRFRDMLPQQNIDRAGIKNRVYSNSVYEILENGQPAGVCILEYATPLQTLFAMSQDAKAGFSREDRLEQAKLFCRTLEEILED\n\n### Answer:", "output": "Abolished sumoylation by TRIM38, leading to decreased stability. Sumoylation of cGAS at K337 prevents its polyubiquitination and degradation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01172", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLLLKNCLTIFFKNKNYVSWRFLEVNDRETWAEDYALDLDKYALEAYWLEAKEVKQDDASVLSCFVFKYQGWFFFFIYYYLFKTQAELDAEAKIDEELSDDEGNIDEDLEDFQNEAKVSTSLELF\n\n### Sequence 2:\nMLLLKNCLTIFFKNKNYVSWRFLEVNDRETWAEDYALDLDKYALEAYWLEAKEVKQDDASVLSCFVFKYQGWFFFFIYYYLFKTQAELDAEAKIDEELSDDEGNIDEDLEDFQNEAKVSTSLELF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01173", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nKIF1B_RAT\n\n### Answer:", "output": "Involved in the translocation of lysosomes from perinuclear regions to the cell periphery.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01174", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAITTAASRLGTEPFSDAPKVELRPKASREEVESVIRAVYRHVLGNDYILASERLVSAESLLRDGNLTVREFVRSVAKSELYKKKFFYNSFQTRLIELNYKHLLGRAPYDESEVVYHLDLYQNKGYDAEIDSYIDSWEYQSNFGDNVVPYYRGFETQVGQKTAGFNRIFRLYRGYANSDRAQVEGTKSRLARELASNKASTIVGPSGTNDSWGFRASADVAPKKNLGNAVGEGDRVYRLEVTGIRSPGYPSVRRSSTVFIVPYERLSDKIQQVHKQGGKIVSVTSA\n\n### Sequence 2:\nMAITTAASRLGVAPYNESRPVELRPDFSLDDAKMVIRAVYRQVLGNDYIMDSERLKGAESLLTNGSISVREFVRTVAKSELYKKKFLYNNFQTRVIELNYKHLLGRAPFSEDEVIFHLDLYENQGFDADIDSYIDSVEYQENFGENIVPYYR-FNNQVGDRTVGFTRMFRLYRGYANSDRSQLERSSSRLATELGQNTVSAIVGPSGSNAGWAYRPSRAGNTPAKALGGTVPFGQASKLFRVEITAISAPGYPKVRRSNKAVIVPFEQLNQTLQQINRLGGKVASITPA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01175", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPSAGSHHNDKLHFKKGDTVIVLSGKHKGQTGKVLLALPRDQKVVVEGVNVITKNVKPSMTNPQGGQEQRELALHASKVALVDPETGKATRVRKQIVDGKKVRVAVASGKT\n\n### Answer:", "output": "<SEQ_3Di>DAPCVVVVLDDPDDQFAWWFFPDDPSHGDIFGWHAARRGSQKTDGQPAKDKAWQQCPVVPPRRTDIDIDHDIGHPPRIGGADPVPGHRFDWDKDDDDDPDIAIQTVHDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01176", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMVGVTAFGNFDLASLAIYSFWIFLAGLIYYLQTENMREGYPLENEDGTPAANQGPFPLPKPKTFILPHGRGTLTVPGPESEDRPIALARTAVSEGFPHAPTGDPMKDGVGPASWVARRDLPELDGHGHNKIKPMKAAAGFHVSAGKNPIGLPVRGCDLEIAGKVVDIWVDIPEQMARFLEVELKDGSTRLLPMQMVKVQSNRVHVNALSSDLFAGIPTIKSPTEVTLLEEDKICGYVAGGLMYAAPKRKSVVAAMLAEYA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHTTSSSCSCBCTTSSBCSCCCSSCCCCCEEEECSTTCCEEEESCCCCCCSCCCEEESSSSTTCCEEESSCHHHHTCGGGCCCCCCSSCCBCTTSSBSEEEGGGSTTCEEEESCCCTTCEEEETTSSEEEEEEEEEEEGGGTEEEEEEEEETTSCEEEEEGGGCEEETTEEEESSCCGGGGGGSCCCSSTTCCCHHHHHHHHHHHHHHHHHTTTSCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01177", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nSLC32A1:low GAD1:high GAD2:medium\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as SLC32A1, GAD1, GAD2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "tumor", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SLC32A1", "GAD1", "GAD2"]}}
{"id": "eval_01178", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSMNPPPPETSNPNKPKRQTNQLQYLLRVVLKTLWKHQFAWPFQQPVDAVKLNLPDYYKIIKTPMDMGTIKKRLENNYYWNAQECIQDFNTMFTNCYIYNKPGDDIVLMAEALEKLFLQKINELPTEE\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCTTSCCCCCHHHHHHHHTHHHHHHTSTTCGGGSSCCCTTTTTCTTHHHHCSSCCCHHHHHHHHHTTCCSSHHHHHHHHHHHHHHHHHHSCTTSHHHHHHHHHHHHHHHHHHTCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01179", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRS20_MYCPA\n\n### Answer:", "output": "Binds directly to 16S ribosomal RNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01180", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nIF2_PORGI\n\n### Answer:", "output": "Belongs to the TRAFAC class translation factor GTPase superfamily. Classic translation factor GTPase family. IF-2 subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01181", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGRP1_HORVU\n\n### Answer:", "output": "Responsible for plasticity of the cell wall.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01182", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDXR_BURPS\n\n### Answer:", "output": "1-deoxy-D-xylulose 5-phosphate reductoisomerase, DXP reductoisomerase, 1-deoxyxylulose-5-phosphate reductoisomerase, 2-C-methyl-D-erythritol 4-phosphate synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01183", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "LAG3, HAVCR2, PDCD1, TOX", "output": "The expression pattern is consistent with exhausted CD8 T cell in a context associated with lung adenocarcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "liver", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["LAG3", "HAVCR2", "PDCD1", "TOX"]}}
{"id": "eval_01184", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMDITDQTIEYVSTLAKLSLQPEEKEKAKKDLGNILGYIDKMKELNTDGIEPMSHVLPMKNVFREDVVTNEENREELLKNAPKQMDGCFMVPKTVD\n\n### Sequence 2:\nMKVDKETIEYVANLARLRLSEKEKEKLALDLESIISYVDKLNELDTSKIIPTDHVIPIKNVFRDDEVRDSYPKDKMLMNAPEKEDGCFKVPKVVE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01185", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMILLWSCLLVAVVGILGTATPQPGNSSLHRLTRQLLQQYHKEVRPVYNWAEATTVYLDLCVHAVLDVDVQNQKLKTSMWYREVWNDEFLSWNSSLFDDIQEISLPLSAIWAPDIIINEFVDVERSPDLPYVYVNSSGTIRNHKPIQVVSACSLQTYAFPFDIQNCSLTFNSILHTVEDIDLGFLRNQEDIENDKRSFLNDSEWQLLSVTSTYHIRQSSAGDFAQIRFNVVIRRCPLAYVVSLLIPSIFLMLVDLGSFYLPPNCRARIVFKTNVLVGYTVFRVNMSDEVPRSAGCTSLIGVFFTVCMALLVLSLSKSILLIKFLYEERHSEQERPLMCLRGDSDANESRLYLRAPCAEDTESPVRQEHQVPSDTLKDFWFQLQSINNSLRTRDQVYQKEVEWLAILCHFDQLLFRIYLAVLGLYTVTLCSLWALWSRM\n\n### Sequence 2:\nMAPLWACILVAA-GILATDTHHPQDSALYHLSKQLLQKYHKEVRPVYNWTKATTVYLDLFVHAILDVDAENQILKTSVWYQEVWNDEFLSWNSSMFDEIREISLPLSAIWAPDIIINEFVDIERYPDLPYVYVNSSGTIENYKPIQVVSACSLETYAFPFDVQNCSLTFKSILHTVEDVDLAFLRSPEDIQHDKKAFLNDSEWELLSVSSTYSILQSSAGGFAQIQFNVVMRRHPLVYVVSLLIPSIFLMLVDLGSFYLPPNCRARIVFKTSVLVGYTVFRVNMSNQVPRSVGSTPLIGHFFTICMAFLVLSLAKSIVLVKFLHDEQRGGQEQPFLCLRGDTDADRPRVEPRAQRAVVTESSLYGEHLAQPGTLKEVWSQLQSISNYLQTQDQTDQQEAEWLVLLSRFDRLLFQSYLFMLGIYTITLCSLWALWG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01186", "category": "Mutation", "instruction": "Describe the effect of mutation K0I7Q2-T297M on protein Flavonoid 4'-O-methyltransferase 3.", "input": "### Wild-type:\nMAVDKEVQLHAQAWEHALSYINSTALSAAVELEIPDILEDHGGLMSLSELSAASGCPREPLYRLMRFLIFHGIFTKSDDCYAQSPLSRLFTRENLGPYMLMQATPVTRSPAGLSGEALKTGTSLYLKSIRGEDSWSDPAYGYHMKAFTNAMIAHARLTAAAIVSNYPAAFDGLRSVVDVGGRHGTAIGRLVEAFPWVRGIAFDLPEIVADAPPRKGVDFVGGDMFESVPKADAVMLMWILHDWSDDKCIEILKKCKEAIPASTGKVMIVDAIINEDGEGDEFSGARLSLDMIMLAVTAQG\n\n### Mutant:\nMAVDKEVQLHAQAWEHALSYINSTALSAAVELEIPDILEDHGGLMSLSELSAASGCPREPLYRLMRFLIFHGIFTKSDDCYAQSPLSRLFTRENLGPYMLMQATPVTRSPAGLSGEALKTGTSLYLKSIRGEDSWSDPAYGYHMKAFTNAMIAHARLTAAAIVSNYPAAFDGLRSVVDVGGRHGTAIGRLVEAFPWVRGIAFDLPEIVADAPPRKGVDFVGGDMFESVPKADAVMLMWILHDWSDDKCIEILKKCKEAIPASTGKVMIVDAIINEDGEGDEFSGARLSLDMIMLAVMAQG\n\n### Answer:", "output": "Almost inclusive cirsimaritin (CIRM) inhibition and gained ability to produce ladanein (LAD) from scutellarein-7-methyl ether (SCU7Me); when associated with T-292 and M-296.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01187", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "GATA1, KLF1, HBA1, HBB", "output": "The marker profile suggests functional enrichment consistent with the typical biology of erythroid progenitor.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "spleen", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["GATA1", "KLF1", "HBA1", "HBB"]}}
{"id": "eval_01188", "category": "Mutation", "instruction": "Describe the effect of mutation Q16635-L182P on protein Tafazzin.", "input": "### Wild-type:\nMPLHVKWPFPAVPPLTWTLASSVVMGLVGTYSCFWTKYMNHLTVHNREVLYELIEKRGPATPLITVSNHQSCMDDPHLWGILKLRHIWNLKLMRWTPAAADICFTKELHSHFFSLGKCVPVCRGDGVYQKGMDFILEKLNHGDWVHIFPEGKVNMSSEFLRFKWGIGRLIAECHLNPIILPLWHVGMNDVLPNSPPYFPRFGQKITVLIGKPFSALPVLERLRAENKSAVEMRKALTDFIQEEFQHLKTQAEQLHNHLQPGR\n\n### Mutant:\nMPLHVKWPFPAVPPLTWTLASSVVMGLVGTYSCFWTKYMNHLTVHNREVLYELIEKRGPATPLITVSNHQSCMDDPHLWGILKLRHIWNLKLMRWTPAAADICFTKELHSHFFSLGKCVPVCRGDGVYQKGMDFILEKLNHGDWVHIFPEGKVNMSSEFLRFKWGIGRLIAECHLNPIILPPWHVGMNDVLPNSPPYFPRFGQKITVLIGKPFSALPVLERLRAENKSAVEMRKALTDFIQEEFQHLKTQAEQLHNHLQPGR\n\n### Answer:", "output": "In BTHS; does not affect mitochondrial localization.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01189", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nC1QB CD163 C1QA CD68 MRC1\n</cell>", "output": "This cell is most consistent with macrophage, a tissue macrophage. Key markers such as MRC1, CD68, C1QB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "intestine", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["C1QB", "CD163", "C1QA", "CD68", "MRC1"]}}
{"id": "eval_01190", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGAMDPEFMNKYQAVIIGFGKAGKTLAVTLAKAGWRVALIEQSNAMYGGTCINIGCIPTKTLVHDAQQHTDFVRAIQRKNEVVNFLRNKNFHNLADMPNIDVIDGQAEFINNHSLRVHRPEGNLEIHGEKIFINTGAQTVVPPIPGITTTPGVYDSTGLLNLKELPGHLGILGGGYIGVEFASMFANFGSKVTILEAASLFLPREDRDIADNIATILRDQGVDIILNAHVERISHHENQVQVHSEHAQLAVDALLIASGRQPATASLHPENAGIAVNERGATVVDKRLHTTADNIWAMGDVTGGLQFTYISLDDYRIVRDELLGEGKRSTDDRKNVPYSVFMTPPLSRVGMTEEQARESGADIQVVTLPVAAIPRARVMNDTRGVLKAIVDNKTQRMLGASLLCVDSHEMINIVKMVMDAGLPYSILRDQIFTHPSMSESLNDLFSLVK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCBSEEEESCSHHHHHHHHHHHHTTCCEEEEESCGGGTTCHHHHTSHHHHHHHHHHHHTTCCHHHHHHHHHHHHHHHHHHHHHHHTTSTTEEEEESEEEECSSSEEEEECTTSCEEEEBSSEEECCCEEECCCSCTTTTTCTTEECHHHHTTCSSCCSEEEEESCHHHHHHHHHHHHHTTCEEEEECCSSSSSTTSCHHHHHHHHHHHHTTSEEEECSCCCCEEECCSSSEEEECSSSCEEESEEEECSCEEECCTTTCGGGTTCCBCSSCCBCCCSSCBCSSTTEEECSGGGTSCCCHHHHHHHHHHHHHHHTTSSCCCGGGCCSCCEEECSSSCEEEEECCHHHHHHHTCCEEEEEEEGGGCHHHHHTTCCCCEEEEEEETTTCCEEEEEEESTTHHHHHHHHHHHHHHTCCTHHHHSCCCCSSCSTTHHHHHHTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01191", "category": "Mutation", "instruction": "Describe the effect of mutation Q99LQ7-A96H on protein Plasmanylethanolamine desaturase 1.", "input": "### Wild-type:\nMAGAEDAPGRQPELDEDETAEGRRWGAQHAGARELAALYSPGKRFQEWCSVILCFSLIAHNLVHLLLLARWEHTPLVILGVVAGALVADFLSGLVAWGADTWGSVDLPIVGKAFIRPFREHHIDPTAITRHDFIETNGDNCLVTLLPLLNMAYKFRTQSPETLEQLYPWECFVFCLTIFGTFTNQIHKWSHTYLGLPYWVTVLQDWHVILPRKHHRIHHVAPHETYFCITTGWLNYPLEVIGFWRRLEDLIQGLTGEKPRADDMKWAQKIK\n\n### Mutant:\nMAGAEDAPGRQPELDEDETAEGRRWGAQHAGARELAALYSPGKRFQEWCSVILCFSLIAHNLVHLLLLARWEHTPLVILGVVAGALVADFLSGLVHWGADTWGSVDLPIVGKAFIRPFREHHIDPTAITRHDFIETNGDNCLVTLLPLLNMAYKFRTQSPETLEQLYPWECFVFCLTIFGTFTNQIHKWSHTYLGLPYWVTVLQDWHVILPRKHHRIHHVAPHETYFCITTGWLNYPLEVIGFWRRLEDLIQGLTGEKPRADDMKWAQKIK\n\n### Answer:", "output": "Enhancement of plasmanylethanolamine desaturase activity. Affects endoplasmic reticulum membrane localization.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01192", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "Cn1cc(S(C)(=O)=O)cc1-c1c2c(=O)n(C)c(=O)n(CC3CC3)c2nn1Cc1ccnc2ccc(Cl)cc12", "output": "This molecule is relatively large and polar, with moderately lipophilic character. It contains aromatic ring, halide, pyridine. Its descriptor profile: MW=553.04, LogP=2.97, TPSA=113.78, HBD=0, HBA=6, QED=0.32.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "Cn1cc(S(C)(=O)=O)cc1-c1c2c(=O)n(C)c(=O)n(CC3CC3)c2nn1Cc1ccnc2ccc(Cl)cc12", "compound_name": null, "pubchem_cid": null, "MW": 553.04, "LogP": 2.97, "TPSA": 113.78, "QED": 0.32, "functional_groups": ["aromatic ring", "halide", "pyridine"], "moleculenet_labels": {"label": 3.18}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01193", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTKSELIERIVTHQGLLSSKDVELAIKTMLEQMSQCLATGDRIEIRGFGSFSLHYRAPRVGRNPKTGQSVSLDGKFVPHFKPGKELRDRVNEDEEEGF\n\n### Sequence 2:\nMTKSELIERIVTHQGLLSSKDVELAIKTMLEQMSQCLATGDRIEIRGFGSFSLHYRAPRVGRNPKTGQSVSLEGKFVPHFKPGKELRDRVNEDEHE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01194", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nOBG_PARXL\n\n### Answer:", "output": "Belongs to the TRAFAC class OBG-HflX-like GTPase superfamily. OBG GTPase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01195", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nYFYB_SCHPO\n\n### Answer:", "output": "Cytoplasm. Nucleus.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01196", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMGKLTLLLLALLVWLQYSLWFGKNGIHDYSRVNDDVVAQQATNAKLKARNDQLFAEIDDLNGGQEAIEERARNELSMTKPGETFYRLVPDASKRAATAGQTHR\n\n### Sequence 2:\nMGKLTLLLLALLVWLQYSLWFGKNGIHDYSRVNDDVVAQQATNAKLKARNDQLFAEIDDLNGGQEAIEERARNELSMTKPGETFYRLVPDASKRAATAGQTHR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01197", "category": "Mutation", "instruction": "Describe the effect of mutation Q02ML8-D124A on protein CRISPR-associated nuclease/helicase Cas3 subtype I-F/YPEST.", "input": "### Wild-type:\nMNILLVSQCEKRALSETRRILDQFAERRGERTWQTPITQAGLDTLRRLLKKSARRNTAVACHWIRGRDHSELLWIVGDASRFNAQGAVPTNRTCRDILRKEDENDWHSAEDIRLLTVMAALFHDIGKASQAFQAKLRNRGKPMADAYRHEWVSLRLFEAFVGPGSSDEDWLRRLADKRETGDAWLSQLARDDRQSAPPGPFQKSRLPPLAQAVGWLIVSHHRLPNGDHRGSASLARLPAPIQSQWCGARDADAKEKAACWQFPHGLPFASAHWRARTALCAQSMLERPGLLARGPALLHD\n\n### Mutant:\nMNILLVSQCEKRALSETRRILDQFAERRGERTWQTPITQAGLDTLRRLLKKSARRNTAVACHWIRGRDHSELLWIVGDASRFNAQGAVPTNRTCRDILRKEDENDWHSAEDIRLLTVMAALFHAIGKASQAFQAKLRNRGKPMADAYRHEWVSLRLFEAFVGPGSSDEDWLRRLADKRETGDAWLSQLARDDRQSAPPGPFQKSRLPPLAQAVGWLIVSHHRLPNGDHRGSASLARLPAPIQSQWCGARDADAKEKAACWQFPHGLPFASAHWRARTALCAQSMLERPGLLARGPALLHD\n\n### Answer:", "output": "In a disruption mutant, does not restore biofilm formation, restores crRNA production.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01198", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL19_ESCF3\n\n### Answer:", "output": "50S ribosomal protein L19.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01199", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMQLRKEYLIAYDIEDNKTRTIIYKQLLAYGLKAVQKSVFWGYVSIAELNAIKRLFDSSLTISDKVFITRVNMHEQKLDYSFGYDDKTFKDWDEYGHI\n\n### Sequence 2:\nMIQTESRLEVADNTGAREVLCIKVLGGSKRRYAGIGDIIKVSVKEATPRGRVKKGEIYNAVVVRTAKGVRRQDGSLIKFDGNAAVLLNNKLEPIGTRIFGPVTRELRSERFMKIVSLAPEVL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01200", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMQRCEITGTGLISGNQISHSHRLTRRVWKPNLQVTTLVVNGSPIKVKVCARTLKTLKGASEVEVMRILKANIATLSERLLKHLNK\n\n### Sequence 2:\nMYLDAAEKQEIFGKYGKSNSDTGSTEAQIALFSYRISHLTEHMKLNRKDYSTERALTMLVAKRRRLLNYLKDKDITRYRSIVKELGLRK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01201", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "NC(=O)C(=CN1C(=O)C(=Cc2cccc(O)c2)SC1=S)C(N)=O", "output": "This molecule is medium-sized and polar, with hydrophilic character. It contains aromatic ring, amide, hydroxyl, phenol. Its descriptor profile: MW=349.39, LogP=0.45, TPSA=126.72, HBD=3, HBA=6, QED=0.31.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "NC(=O)C(=CN1C(=O)C(=Cc2cccc(O)c2)SC1=S)C(N)=O", "compound_name": null, "pubchem_cid": null, "MW": 349.39, "LogP": 0.45, "TPSA": 126.72, "QED": 0.31, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "phenol"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01202", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSHHHHHHGRSMSELIKENMHMKLYMEGTVNNHHFKCTSEGEGKPYEGTQTMRIKVVEGGPLPFAFDILATSFMYGSRTFINHTQGIPDFWKQSFPEGFTWERVTTYEDGGVLTATQDTSLQDGCLIYNVKLRGVNFPSNGPVMQKKTLGWEAATEMLYPADGGLEGRGDMALKLVGGGHLICNLKTTYRSKNPAKNLKMPGVYFVDHRLERIKEADKETYVEQHEVAVARYCDLPSKLGHKLN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCSCSSEEEEEEEEEEETTEEEEEEEEEEEETTTTEEEEEEEEEESCSCSSCGGGGGGGCCCTTSCEECTTCCCTTTTTTTTCEEEEEEEEETTSCEEEEEEEEEEETTEEEEEEEEEEECCCTTSTTTTTCEEEECCEEEEEEEETTEEEEEEEEEEEETTSCEEEEEEEEEEEESCSSSCCCCCCSEEEEEEEEEEEEETTTTEEEEEEEEEEECCCSCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01203", "category": "Mutation", "instruction": "Describe the effect of mutation Q91XW8-K254A on protein Inactive peptidyl-prolyl cis-trans isomerase FKBP6.", "input": "### Wild-type:\nMGGSTRDPGALEGAGILGQSPYERLSQRMLDISGDRGVLKDIIREGTGDTVTPDASVLVKYSGYLEHMDKPFDSNCFRKTPRLMKLGEDITLWGMELGLLSMRKGELARFLFKPAYAYGTLGCPPLIPPNATVLFEIELIDFLDSAESDKFCALSAEQQEQFPLQKVLKVAATEREFGNYLFRQNRFCDAKVRYKRALLLLHRRLATCEEQHLVEPAVLLVLLNLSFVYLKLDRPAMALRYGEQALLIDKRNAKALFRCGQACLLLTEYERARDFLVRAQKEQPCNHDINNELKKLSSHY\n\n### Mutant:\nMGGSTRDPGALEGAGILGQSPYERLSQRMLDISGDRGVLKDIIREGTGDTVTPDASVLVKYSGYLEHMDKPFDSNCFRKTPRLMKLGEDITLWGMELGLLSMRKGELARFLFKPAYAYGTLGCPPLIPPNATVLFEIELIDFLDSAESDKFCALSAEQQEQFPLQKVLKVAATEREFGNYLFRQNRFCDAKVRYKRALLLLHRRLATCEEQHLVEPAVLLVLLNLSFVYLKLDRPAMALRYGEQALLIDKRNAAALFRCGQACLLLTEYERARDFLVRAQKEQPCNHDINNELKKLSSHY\n\n### Answer:", "output": "Abolishes interaction with HSP90. The mutation in Fkbp6 leads to derepression of LINE1 retrotransposon and reduced DNA methylation due to deficient nuclear accumulation of Miwi2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01204", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGYPPARPVVSCQAADYENFSCTWSPSQISGLPTRYLTSYRKKTVPSTGPWPCPQDPLGAARCVVHGAEFWSQYRINVTEVNPLGASTRLLDVSLQSILRPDPPQGLRVESVPGYPRRLRASWTYPASWPSQPHFLLKFRLQYRPAQHPAWSTVEPAGLEEVITDAVAGLPHAVRVSARDFLDAGTWSTWSPEAWGTPSTG\n\n### Answer:", "output": "<SEQ_3Di>DDAFAEWAKAWAAQEQFKTKIFIGGPDCPQPDKDKFKWKFFDDCVPRDTDGFDDPVRPDRMGMDGRDDQATKMWMWIWMDGPHYIHIDIDIDHNQARYEHAFFADWDKEADPPAQAKIKIFTHGDPSYDDDVPWDKWKKKWKDFPPDDDIDIDTDDDGMDMGRGHDGPTKMWMWMWIDTPVNGYDIHDIDDIYIDGYDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01205", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "CC(=NNC(=O)CC(=O)Nc1ccccc1Cl)c1cc2cc(N=Nc3ccccc3Cl)ccc2oc1=O", "output": "The SMILES represents an unnamed molecule with MW=536.38, LogP=6.38, TPSA=125.49. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "CC(=NNC(=O)CC(=O)Nc1ccccc1Cl)c1cc2cc(N=Nc3ccccc3Cl)ccc2oc1=O", "compound_name": null, "pubchem_cid": null, "MW": 536.38, "LogP": 6.38, "TPSA": 125.49, "QED": 0.09, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01206", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "LST1, CDKN1C, MS4A7, FCGR3A", "output": "The expression pattern is consistent with CD16+ monocyte in a context associated with Alzheimer disease, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "bone marrow", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["LST1", "CDKN1C", "MS4A7", "FCGR3A"]}}
{"id": "eval_01207", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMQTSPLLTQLMEALRCLPGVGPKSAQRMAFTLLQRDRSGGMRLAQALTRAMSEIGHCADCRTFTEQDVCNICTNPRRQENGQICVVESPADIYAIEQTGQFSGRYFVLMGHLSPLDGIGPDDIGLDRLEQRLESETLTEVILATNPTVEGEATANYIGELCAQYGVSASRIAHGVPVGGELEMVDGTTLSHSLAGRHKLIF\n\n### Sequence 2:\nMQTSPLLTQLMEALRCLPGVGPKSAQRMAFTLLQRDRSGGMRLAQALTRAMSEIGHCADCRTFTEQDVCNICTNPRRQENGQICVVESPADIYAIEQTGQFSGRYFVLMGHLSPLDGIGPDDIGLDRLEQRLESETLTEVILATNPTVEGEATANYIGELCAQYGVSASRIAHGVPVGGELEMVDGTTLSHSLAGRHKLIF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01208", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCD123_BOVIN\n\n### Answer:", "output": "Cell division cycle protein 123 homolog.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01209", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSSTT_YERPS\n\n### Answer:", "output": "Serine/threonine transporter SstT, Na(+)/serine-threonine symporter.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01210", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMIVGLSNPKKEYHSTRHNVGSWYLYSLAESYLRNFKNEKKFFGFTTSLNIESNYIRLLIPNIFMNINGQSVFKMASFYNINLSEILIVHDDLELQPGISKLKYSYGHNGHNGLRDIVNTFNKNINFYRFRIGIGRPINRDQIASFVLSNPTKKEKILIQKSILHAIEKNVLSNILKF\n\n### Sequence 2:\nMAKTTAAITFVTGNAKKLQEVQQILGQGFPFELTNRKIDLPELQGEPEDISREKCRLAAAEVKGPVMVEDTSLCFNALHGLPGPYIKWFLDKTGHVGLNNLLAAYPDKSAYAQCIFAFTTGPGAEIQTFVGRTEGKIVPARGPTDFGWDPVFQPDGFEETYAEMDKTIKNSISHRGRSLSALCAYFDTHKAELEKQLAA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01211", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CCOP(=O)(OCC)C(c1ccccc1)N(C(=O)NC(F)(F)F)c1ccc(C)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "CCOP(=O)(OCC)C(c1ccccc1)N(C(=O)NC(F)(F)F)c1ccc(C)cc1", "compound_name": null, "pubchem_cid": null, "MW": 444.39, "LogP": 6.0, "TPSA": 67.87, "QED": 0.4, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "halide"]}}
{"id": "eval_01212", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "PRG2, IL5RA, EPX", "output": "This cell is most consistent with eosinophil, an eosinophil. Key markers such as EPX, IL5RA, PRG2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "eosinophil", "tissue": "liver", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PRG2", "IL5RA", "EPX"]}}
{"id": "eval_01213", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNLAGMIDHTALKAETSRAQIETLCKEALEYKFASVCVNPTNVALAAELLKSDDAVKVCTVIGFPLGANTPEVKAFETQDAIKNGATEIDMVLNIGALKDGDLSLVERDIRAVVEAANGTLVKVIFENCLLTKEEIKTAAELSVKAGADFVKTSTGFSTGGATVEDIRLMRETVGPDIGVKASGGVRDFEGAKAMIDAGATRIGASAGIAIVTGGTSDSDY\n\n### Sequence 2:\nMRKLTLAFAAASLLFTLNSAVVARASTPQPLWVGTNVAQLAEQAPIHWVSVAQIENSLLGRPPMAVGFDIDDTVLFSSPGFWRGQKTFSPGSEDYLKNPQFWEKMNNGWDEFSMPKEVARQLIAMHVKRGDSIWFVTGRSQTKTETVSKTLQDDFLIPAANMNPVIFAGDKPGQNTKTQWLQAKQIKVFYGDSDNDITAAREAGARGIRVLRAANSSYKPLPMAGALGEEVIVNSEY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01214", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nEAIVNAQPKCNPNLHYWTTQDEGAAIGLAWIPYFGPAAEGIYTEGLMHNQDGLICGLRQLANETTQALQLFLRATTELRTFSILNRKAIDFLLQRWGGTCHILGPDCCIEPHDWTKNITDKIDQIIHDFVD\n\n### Answer:", "output": "<SEQ_2D>CCCCCCSSCCCCCEEEEEECCCCCCSTTTTSTTSSCCTTTSEECCEECCCTTHHHHHHHHHHHHHHHHHHHHHHCCSSEECTHHHHHHHHHHHHTTCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01215", "category": "Mutation", "instruction": "Describe the effect of mutation Q4VP08-A68R on protein Protein LURE 1.2.", "input": "### Wild-type:\nMKLPIIFLTLLIFVSSCTSTLINGSSDEERTYSFSPTTSPFDPRSLNQELKIGRIGYCFDCARACMRAGKYIRTCSFERKLCRCSISDIK\n\n### Mutant:\nMKLPIIFLTLLIFVSSCTSTLINGSSDEERTYSFSPTTSPFDPRSLNQELKIGRIGYCFDCARACMRRGKYIRTCSFERKLCRCSISDIK\n\n### Answer:", "output": "Enhanced interaction with PRK6. Abnormal pollen tube attraction. The mutation in PRK6 does not affect the interaction with AtLURE1 peptides.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01216", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAKAPVRARKRVRKQVSDGVAHIHASFNNTIVTITDRQGNALGWATAGGSGFRGSRKSTPFAAQVAAERCADAVKEYGIKNLEVMVKGPGPGRESTIRALNAAGFRITNITDVTPIPHNGCRPPKKRRV\n\n### Sequence 2:\nMAKAPIRARKRVRKQVSDGVAHIHASFNNTIVTITDRQGNALGWATAGGSGFRGSRKSTPFAAQVAAERCAEAVKEYGIKNLEVMVKGPGPGRESTIRALNAAGFRITNITDVTPIPHNGCRPPKKRRV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01217", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIILVMLNTRFFNPYCFQPSSSLCRPSYSLFSGILACISSSKILPFENFFKSTLLGGFFTVVGHSLCRYLSCHYLPSSHRLTGKQRFSHSQSLFPPHPLCFLRRCCYYLLRLGKVTLEGSLWYRPNPLAFFPWKELKDGSFTTSVGSALFDIGDWFGFLSFMSYSCDMPTP\n\n### Sequence 2:\nMTTILRIDSSIKGEAAVSRRLTQRILDRLLEAHPDATVVSRDLAQGIRQIDGPWLGSVFTAPEQRTADQQEIARTAGAVMAEVKEADILVIALPVYNFGAPAQLKSWVDHIARRGESFVYTETGPVGLLTGKRAIVAFTSDGTPLGSELDHASGWLRQVLGFVGITDVDFVAADRMVFGADEAMARAEAAVAALAA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01218", "category": "Mutation", "instruction": "Describe the effect of mutation Q9HTR2-R264K on protein Phosphorylcholine phosphatase.", "input": "### Wild-type:\nMTFAKGILAALALAAAVGQASATELEHWPAPAARQLNALIEANANKGAYAVFDMDNTSYRYDLEESLLPYLEMKGVLTRDRLDPSLKLIPFKDQAGHKESLFSYYYRLCEIDDMVCYPWVAQVFSGFTLRELKGYVDELMAYGKPIPATYYDGDKLATLDVEPPRVFSGQRELYNKLMENGIEVYVISAAHEELVRMVAADPRYGYNAKPENVIGVTTLLKNRKTGELTTARKQIAEGKYDPKANLDLEVTPYLWTPATWMAGRQAAILTYIDRWKRPILVAGDTPDSDGYMLFNGTAEN\n\n### Mutant:\nMTFAKGILAALALAAAVGQASATELEHWPAPAARQLNALIEANANKGAYAVFDMDNTSYRYDLEESLLPYLEMKGVLTRDRLDPSLKLIPFKDQAGHKESLFSYYYRLCEIDDMVCYPWVAQVFSGFTLRELKGYVDELMAYGKPIPATYYDGDKLATLDVEPPRVFSGQRELYNKLMENGIEVYVISAAHEELVRMVAADPRYGYNAKPENVIGVTTLLKNRKTGELTTARKQIAEGKYDPKANLDLEVTPYLWTPATWMAGKQAAILTYIDRWKRPILVAGDTPDSDGYMLFNGTAEN\n\n### Answer:", "output": "Almost significant change in activity. Slight increase in inhibition by tetramethylammonium chloride.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01219", "category": "Mutation", "instruction": "Describe the effect of mutation Q9JIL4-A14K on protein Na.", "input": "### Wild-type:\nMASTFNPRECKLSAQEGQNYGFFLRIEKDTDGHLIRVIEEGSPAEKAGLLDGDRVLRINGVFVDKEEHAQVVELVRKSGNSVTLLVLDGDSYEKAVKNQVDLKELDQSQREAALNDKKPGPGMNGAVEPCAQPRLCYLVKEGNSFGFSLKTIQGKKGVYLTDIMPQGVAMKAGVLADDHLIEVNGENVENASHEEVVEKVTKSGSRIMFLLVDKETARCHSEQKTQFKRETASLKLLPHQPRVVVIKKGSNGYGFYLRAGPEQKGQIIKDIEPGSPAEAAGLKNNDLVVAVNGKSVEALD\n\n### Mutant:\nMASTFNPRECKLSKQEGQNYGFFLRIEKDTDGHLIRVIEEGSPAEKAGLLDGDRVLRINGVFVDKEEHAQVVELVRKSGNSVTLLVLDGDSYEKAVKNQVDLKELDQSQREAALNDKKPGPGMNGAVEPCAQPRLCYLVKEGNSFGFSLKTIQGKKGVYLTDIMPQGVAMKAGVLADDHLIEVNGENVENASHEEVVEKVTKSGSRIMFLLVDKETARCHSEQKTQFKRETASLKLLPHQPRVVVIKKGSNGYGFYLRAGPEQKGQIIKDIEPGSPAEAAGLKNNDLVVAVNGKSVEALD\n\n### Answer:", "output": "Improves interaction of the first PDZ domain with SCARB1. The mutant behaved unlike wild-type PDZK1 and exacerbated their hepatic SR-BI and plasma lipoprotein abnormalities.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01220", "category": "Mutation", "instruction": "Describe the effect of mutation Q12809-T74M on protein Potassium voltage-gated channel subfamily H member 2.", "input": "### Wild-type:\nMPVRRGHVAPQNTFLDTIIRKFEGQSRKFIIANARVENCAVIYCNDGFCELCGYSRAEVMQRPCTCDFLHGPRTQRRAAAQIAQALLGAEERKVEIAFYRKDGSCFLCLVDVVPVKNEDGAVIMFILNFEVVMEKDMVGSPAHDTNHRGPPTSWLAPGRAKTFRLKLPALLALTARESSVRSGGAGGAGAPGAVVVDVDLTPAAPSSESLALDEVTAMDNHVAGLGPAEERRALVGPGSPPRSAPGQLPSPRAHSLNPDASGSSCSLARTRSRESCASVRRASSADDIEAMRAGVLPPPP\n\n### Mutant:\nMPVRRGHVAPQNTFLDTIIRKFEGQSRKFIIANARVENCAVIYCNDGFCELCGYSRAEVMQRPCTCDFLHGPRMQRRAAAQIAQALLGAEERKVEIAFYRKDGSCFLCLVDVVPVKNEDGAVIMFILNFEVVMEKDMVGSPAHDTNHRGPPTSWLAPGRAKTFRLKLPALLALTARESSVRSGGAGGAGAPGAVVVDVDLTPAAPSSESLALDEVTAMDNHVAGLGPAEERRALVGPGSPPRSAPGQLPSPRAHSLNPDASGSSCSLARTRSRESCASVRRASSADDIEAMRAGVLPPPP\n\n### Answer:", "output": "In LQT2. unknown", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01221", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDNCH_PARMW\n\n### Answer:", "output": "1,4-dihydroxy-2-naphthoyl-CoA hydrolase, DHNA-CoA hydrolase, DHNA-CoA thioesterase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01222", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMREYKLVVLGSGGVGKSALTVQFVQGIFVEKYDPTIEDSYRKQVEVDAQQCMLEILDTAGTEQFTAMRDLYMKNGQGFALVYSITAQSTFNDLQDLREQILRVKDTDDVPMILVGNKCDLEDERVVGKEQGQNLARQWNNCAFLESSAKSKINVNEIFYDLVRQINR\n\n### Answer:", "output": "<SEQ_2D>CEEEEEEEEESTTSSHHHHHHHHHHCCCCCSCCCCSCEEEEEEEESSSSEEEEEEEECCSSCCCHHHHHHHHHHCSEEEEEEESSCTHHHHTTHHHHHHHHHHHSSSCCCEEEEEECTTCGGGCCSCHHHHHHHHHHTTSCEEEECCSSSCTTTTHHHHHHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01223", "category": "Mutation", "instruction": "Describe the effect of mutation Q8S403-A312T on protein Phosphate transporter PHO1.", "input": "### Wild-type:\nMVKFSKELEAQLIPEWKEAFVNYCLLKKQIKKIKTSRKPKPASHYPIGHHSDFGRSLFDPVRKLARTFSDKLFSNSEKPEILQVRRRRGSSETGDDVDEIYQTELVQLFSEEDEVKVFFARLDEELNKVNQFHKPKETEFLERGEILKKQLETLAELKQILSDRKKRNLSGSNSHRSFSSSVRNSDFSAGSPGELSEIQSETSRTDEIIEALERNGVSFINSATRSKTKGGKPKMSLRVDIPDAVAGAEGGIARSIATAMSVLWEELVNNPRSDFTNWKNIQSAEKKIRSAFVELYRGLG\n\n### Mutant:\nMVKFSKELEAQLIPEWKEAFVNYCLLKKQIKKIKTSRKPKPASHYPIGHHSDFGRSLFDPVRKLARTFSDKLFSNSEKPEILQVRRRRGSSETGDDVDEIYQTELVQLFSEEDEVKVFFARLDEELNKVNQFHKPKETEFLERGEILKKQLETLAELKQILSDRKKRNLSGSNSHRSFSSSVRNSDFSAGSPGELSEIQSETSRTDEIIEALERNGVSFINSATRSKTKGGKPKMSLRVDIPDAVAGAEGGIARSIATAMSVLWEELVNNPRSDFTNWKNIQSAEKKIRSAFVELYRGLG\n\n### Answer:", "output": "In pho1-6; disruptionDisruption of Pi accumulation. We identified two pho2 suppressors as carrying missense mutations in PHO1, which has been implicated in Pi loading to the xylem. In support of the genetic interaction between PHO1 and PHO2, we found that the protein level of PHO1 is increased in pho2, whereas such accumulation is ameliorated in both pho2 suppressors. Results from cycloheximide and endosomal C protease inhibitor E-64d treatments further suggest that PHO1 degradation is PHO2 dependent and involves multivesicular body-mediated vacuolar proteolysis. Using the transient expression system of tobacco (Nicotiana tabacum) leaves, we demonstrated that PHO1 and PHO2 are partially colocalized and physically interact in the endomembranes, where the ubiquitin conjugase activity of PHO2 is required for PHO1 degradation. In addition, reduced PHO1 expression caused by PHO1 mutations impede Pi uptake, indicating a functional association between xylem loading and acquisition of Pi.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01224", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSYS_ACIET\n\n### Answer:", "output": "Serine--tRNA ligase, Seryl-tRNA synthetase, SerRS, Seryl-tRNA(Ser/Sec) synthetase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01225", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C(CCl)c1ccc(Cl)cc1Cl", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains aromatic ring, ketone, halide. Its descriptor profile: MW=223.49, LogP=3.41, TPSA=17.07, HBD=0, HBA=1, QED=0.55.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=C(CCl)c1ccc(Cl)cc1Cl", "compound_name": null, "pubchem_cid": null, "MW": 223.49, "LogP": 3.41, "TPSA": 17.07, "QED": 0.55, "functional_groups": ["aromatic ring", "ketone", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": null, "SR-ARE": 1.0, "SR-ATAD5": 1.0, "SR-HSE": 1.0, "SR-MMP": 1.0, "SR-p53": 1.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01226", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGLVPRGSHPESQLDLRVQELIKLICNVQAMEEKTELQSPEHPLDQHYRNLHCALRPLDHESYEFKVISQYLQSTHAPTHSDYTMTLLDLFEVEKDGEKEAFREDLHNRMLLWHGSRMSNWVGILSHGLRIAPPEAPITGYMFGKGIYFADMSSKSANYCFASRLKNTGLLLLSEVALGQCNELLEANPKAEGLLQGKHSTKGLGKMAPSSAHFVTLNGSTVPLGPASDTGILNPDGYTLNYNEYIVYNPNQVRMRYLLKVQFNFLQLW\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCSCHHHHHHHHHHCCCCCCCCCCCCCCCSCHHHHHHHTTTEEEEEECTTSHHHHHHHHHHHHTSCTTCTTEEEEEEEEEEEEETTHHHHSCTTCSCEEEEEEECCGGGHHHHHHHCCCCCCSSCCSSCCTTSSSEEEESSHHHHHGGGCCCSSSCEEEEEEEEEECCSEEEESSCCTTTTGGGTTCSEEEECEEEEECGGGCEEETTEEECCSCEEECCCCCSSSCCEEECEEEESSGGGEEEEEEEEEEEEECCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01227", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nCLCA1:low MUC2:high TFF3:high\n</cell>", "output": "This cell is most consistent with goblet cell, an intestinal goblet cell. Key markers such as CLCA1, MUC2, TFF3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "goblet cell", "tissue": "lymph node", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CLCA1", "MUC2", "TFF3"]}}
{"id": "eval_01228", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMEGSLEREAPAGALAAVLKHSSTLPPESTQVRGYDFNRGVNYRALLEAFGTTGFQATNFGRAVQQVNAMIEKKLEPLSQDEDQHADLTQSRRPLTSCTIFLGYTSNLISSGIRETIRYLVQHNMVDVLVTTAGGVEEDLIKCLAPTYLGEFSLRGKELRENGINRIGNLLVPNENYCKFEDWLMPILDQMVMEQNTEGVKWTPSKMIARLGKEINNPESVYYWAQKNHIPVFSPALTDGSLGDMIFFHSYKNPGLVLDIVEDLRLINTQAIFAKCTGMIILGGGVVKHHIANANLMRNGADYAVYINTAQEFDGSDSGARPDEAVSWGKIRVDAQPVKVYADASLVFPLLVAETFAQKMDAFMHEKNED\n\n### Sequence 2:\nMEGSLEREAPAGALAAVLKHSSTLPPESTQVRGYDFNRGVNYRALLEAFGTTGFQATNFGRAVQQVNAMIEKKLEPLSQDEDQHADLTQSRRPLTSCTIFLGYTSNLISSGIRETIRYLVQHNMVDVLVTTAGGVEEDLIKCLAPTYLGEFSLRGKELRENGINRIGNLLVPNENYCKFEDWLMPILDQMVMEQNTEGVKWTPSKMIARLGKEINNPESVYYWAQKNHIPVFSPALTDGSLGDMIFFHSYKNPGLVLDIVEDLRLINTQAIFAKCTGMIILGGGVVKHHIANANLMRNGADYAVYINTAQEFDGSDSGARPDEAVSWGKIRVDAQPVKVYADASLVFPLLVAETFAQKMDAFMHEKNED\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01229", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTHIG_SYNP6\n\n### Answer:", "output": "Thiazole synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01230", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: VEGFR2\nLigand SMILES: O=C([O-])CCCCCCCC(=O)[O-]\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "ClinTox", "canonical_smiles": "O=C([O-])CCCCCCCC(=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 186.21, "LogP": -0.78, "TPSA": 80.26, "QED": 0.47, "functional_groups": [], "moleculenet_labels": {"FDA_APPROVED": 1, "CT_TOX": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_01231", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nASLPHPKIVKKHTKKFKRHHSDRYHRVAENWRKQKGIDSVVRRRFRGNISQPKIGYGSNKKTKFLSPSGHKTFLVANVKDLETLTMHTKTYAAEIAHNISAKNRVVILARAKALGIKVTNPKGRLALEA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCSCCSCCCCTTTTTCTTSCSSCCCCCCSSCHHHHTCTTSCCCCCGGGCCCTTTTTBCTTSCEEEEECSHHHHHTTGGGTTTEEEEECTTCCHHHHHHHHHHHHHHTCEESCCSTTCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01232", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nBH133_ARATH\n\n### Answer:", "output": "Belongs to the bHLH protein family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01233", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSRVAKAPVNIPAGVEVKLDGQLLTVKGKNGELSRKIHESVEVKQDNGQFTFTPREGFVEANAQSGTARALVNAMVIGVTEGFTKKLVLVGVGYRAQLKGNAIALSLGYSHPVEHTLPVGITAECPSQTEIVLKGADKQLIGQVAADIRAYRRPEPYKGKGVRYADEVVRIKEAKKK\n\n### Sequence 2:\nMIEINVYYKKWYNVIRKPKSFVKNIINTSLIDLNIYEYKPIISIVLANNKLLQQLNYEYRKKNKPTNVLSFPYNKLDKNCYLGEIFISLDVLMNESVDLNIPIEHHTSHMLIHGLLHILDYDHEEPLDQYIMESIEIKLLDKLGIKNPYVSRET\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01234", "category": "Mutation", "instruction": "Describe the effect of mutation P22033-I505T on protein Methylmalonyl-CoA mutase, mitochondrial.", "input": "### Wild-type:\nMLRAKNQLFLLSPHYLRQVKESSGSRLIQQRLLHQQQPLHPEWAALAKKQLKGKNPEDLIWHTPEGISIKPLYSKRDTMDLPEELPGVKPFTRGPYPTMYTFRPWTIRQYAGFSTVEESNKFYKDNIKAGQQGLSVAFDLATHRGYDSDNPRVRGDVGMAGVAIDTVEDTKILFDGIPLEKMSVSMTMNGAVIPVLANFIVTGEEQGVPKEKLTGTIQNDILKEFMVRNTYIFPPEPSMKIIADIFEYTAKHMPKFNSISISGYHMQEAGADAILELAYTLADGLEYSRTGLQAGLTIDE\n\n### Mutant:\nMLRAKNQLFLLSPHYLRQVKESSGSRLIQQRLLHQQQPLHPEWAALAKKQLKGKNPEDLIWHTPEGISIKPLYSKRDTMDLPEELPGVKPFTRGPYPTMYTFRPWTIRQYAGFSTVEESNKFYKDNIKAGQQGLSVAFDLATHRGYDSDNPRVRGDVGMAGVAIDTVEDTKILFDGIPLEKMSVSMTMNGAVIPVLANFIVTGEEQGVPKEKLTGTIQNDILKEFMVRNTYIFPPEPSMKIIADIFEYTAKHMPKFNSISISGYHMQEAGADAILELAYTLADGLEYSRTGLQAGLTIDE\n\n### Answer:", "output": "In MMAM; decreased protein expression; decreased methylmalonyl-CoA mutase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01235", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMPKAENKKSRAAEIVTREYTINLHKRLHGVGFKKRAPRAVKEIKKFASKIMGTTDVRVDPRLNKFVWNQGIRSVPYRVRVRLARKRNDDEDAKEKLYTLVTYVPVTSFKGLQTQNVDAE\n\n### Sequence 2:\nMNINATLFAQIIVFFGLVWFTMKFVWPPIAKALDERAAKVAEGLAAAERGKSDFEQAEKKVAELLAEGRNQVSEMVANAEKRAAKIVEEAKEQASSEAARIAAQAKADVEQELFRARESLREQVAVLAVKGAESILRSEVDASKHAKLLDTLKQEL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01236", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNDK_BRUMA\n\n### Answer:", "output": "Nucleoside diphosphate kinase, NDK, NDP kinase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01237", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAAKIRKGDKVIVLTGRDKGRTGEVFAVRPSEGRALVRGVNMVKRHQKQTPQQEGGIISKESPIHLSNLALLDKDGKPTRVGFKIGKDGTKVRIAKSSGAEIDG\n\n### Sequence 2:\nMLKPLGDRVVLKIEEKEQTVGGFVLAGSAQEKTKTAQVVATGQGVRTLNGDLVAPSVKTGDRVLVEAHAGLDVKDGDEKYIIVGEANILAIIEE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01238", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPHNX_PARPJ\n\n### Answer:", "output": "Belongs to the HAD-like hydrolase superfamily. PhnX family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01239", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALAAAQTNAAANLEKIKKLRNVSAWKVILYNDDIHNFTYVTDVIVKVVGQISKAKAHTITVEAHSTGQALILSTWKSKAEKYCQELQQNGLTVSIIHE\n\n### Answer:", "output": "<SEQ_3Di>DQAFEEEDAPLFLQVLLQVLQVVLCVVPVHHYDYDHDVVCVPVAVVQPVVQDDGFKYKDKLQQVQVCVVVVFFDFDDDDPVLLVQFDPVQQVSQFPLHTGFWDFWFKKWKWKKFFCVLPVDQAQAPVCQVVVQVVVLLVQAAAEAEALLFCLFLQLQLPFQPADQFDDDDSDTDLLHGRCQDPSNLVRLVSVLVCLVQQRAPNVGHHFVRLQCVLCVRYGIYMDMPVSVVVSVVSPGDMDTAFHHHYPPGGRAHEMITTTMTGTPSDPCPVVVVCSVRPRCLDLSNVVSRCVSPRRQATRGPVNNVVVVVRVNSVRRVVNNVSHHHNHNDPLVLQLRVLSSVLSSCCNVVVDPNNVSSVSSSVSSSDDPVVRVVVVVVLWKWKKFFSRLSDDLPLQLVQLCVLPVPDDSVNSSVQSVSRSRNGMGTSDHHDPVSQVSSVVSSVVSPTDMDMTGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01240", "category": "Mutation", "instruction": "Describe the effect of mutation P00898-P289L on protein Anthranilate synthase component 1.", "input": "### Wild-type:\nMQTPKPTLELLTCDAAYRENPTALFHQVCGDRPATLLLESADIDSKDDLKSLLLVDSALRITALGDTVTIQALSDNGASLLPLLDTALPAGVENDVLPAGRVLRFPPVSPLLDEDARLCSLSVFDAFRLLQGVVNIPTQEREAMFFGGLFAYDLVAGFEALPHLEAGNNCPDYCFYLAETLMVIDHQKKSTRIQASLFTASDREKQRLNARLAYLSQQLTQPAPPLPVTPVPDMRCECNQSDDAFGAVVRQLQKAIRAGEIFQVVPSRRFSLPCPSPLAAYYVLKKSNPSPYMFFMQDND\n\n### Mutant:\nMQTPKPTLELLTCDAAYRENPTALFHQVCGDRPATLLLESADIDSKDDLKSLLLVDSALRITALGDTVTIQALSDNGASLLPLLDTALPAGVENDVLPAGRVLRFPPVSPLLDEDARLCSLSVFDAFRLLQGVVNIPTQEREAMFFGGLFAYDLVAGFEALPHLEAGNNCPDYCFYLAETLMVIDHQKKSTRIQASLFTASDREKQRLNARLAYLSQQLTQPAPPLPVTPVPDMRCECNQSDDAFGAVVRQLQKAIRAGEIFQVVPSRRFSLPCPSPLAAYYVLKKSNLSPYMFFMQDND\n\n### Answer:", "output": "Decrease in feedback control by tryptophan. The mutant form of the complex with the substitution displayed a marked decrease in affinity for tryptophan with little or no change in substrate affinity or catalytic capacity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01241", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nFBID_MYCUA\n\n### Answer:", "output": "Belongs to the CofC family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01242", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMFKNTFQSGFLSILYSIGSKPLQIWDKKVRNGHIKRITDNDIQSLVLEVEGTNVSTTYITCPADPKKTLGIKLPFLVMIIKNLKKYFTFEVQVLDDKNVRRRFRASNYQSTTRVKPFICTMPMRLDDGWNQIQFNLSDFTRRAYGTNYIETLRVQIHANCRIRRVYFSDRLYSEDELPAEFKLYLPVQNKAKQ\n\n### Sequence 2:\nMFKNTFQSGFLSILYSIGSKPLQIWDKKVRNGHIKRITDNDIQSLVLEIEGTNVSTTYITCPADPKKTLGIKLPFLVMIIKNLKKYFTFEVQVLDDKNVRRRFRASNYQSTTRVKPFICTMPMRLDDGWNQIQFNLLDFTRRAYGTNYIETLRVQIHANCRIRRVYFSDRLYSEDELPAEFKLYLPVQNKAKQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01243", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nGZMB LILRA4 TCF4 CLEC4C IRF7\n</cell>", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as GZMB, LILRA4, IRF7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "lung", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["GZMB", "LILRA4", "TCF4", "CLEC4C", "IRF7"]}}
{"id": "eval_01244", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSQSFEGELKTLLRSGKVILGTRKTLKLLKTGKVKGVVVSSTLRQDLKDDIMTFSKFSDIPIYLYKGSGYELGTLCGKPFMVSVIGIVDEGESKILEFIKEVKQ\n\n### Sequence 2:\nELKKALETGKVILGSNETIRLAKTGGAKLIIVARNAPKEIKDDIYYYAKLSDIPVYEFEGTSVELGTLLGKPFVVASLAIVDPGESRILALVK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01245", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "ABCA3, SFTPC, SFTPB, SFTPA1", "output": "The marker profile suggests functional enrichment consistent with the typical biology of alveolar type II cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "spleen", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["ABCA3", "SFTPC", "SFTPB", "SFTPA1"]}}
{"id": "eval_01246", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "MS4A1, IGHA1, IGHG1, CD27", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as MS4A1, IGHG1, CD27 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "tumor", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["MS4A1", "IGHA1", "IGHG1", "CD27"]}}
{"id": "eval_01247", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nKTVVVTTILESPYVMMKKNHEMLEGNERYEGYCVDLAAEIAKHCGFKYKLTIVGDGKYGARDADTKIWNGMVGELVYGKADIAIAPLTITLVREEVIDFSKPFMSLGISIMIKKGTPIESAEDLSKQTEIAYGTLDSGSTKEFFRRSKIAVFDKMWTYMRSAEPSVFVRTTAEGVARVRKSKGKYAYLLESTMNEYIEQRKPCDTMKVGGNLDSKGYGIATPKGSSLGNAVNLAVLKLNEQGLLDKLKNKWWYDKGEC\n\n### Answer:", "output": "<SEQ_2D>CCEEEEECCBTTTBEECTTGGGSCGGGGEESHHHHHHHHHHHHHTCCEEEEECTTCCCCCBCTTTCCBCHHHHHHHTTSCSEECSSCBCCHHHHTTEEECSCSEEECEEEEEETTCSCCSHHHHHTCSSSEEEEBTTSHHHHHHHHCCSHHHHHHHHHHHHCSSCCCBSSHHHHHHHHHHTTTSEEEEEEHHHHHHHHTSTTCCEEEESCCSCCEEECCEEETTCTTHHHHHHHHHHHHHHTHHHHHHHHHTTTTCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01248", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMEVAAKLSGARISAQKARLVADQIRGKKVGEALNLLAFSSKKAAEILKKVLESAVANAEHNEGADVDDLKVSTVFVNEGRSLKRIMPRAKGRADRIVKRSCHITVKVADK\n\n### Sequence 2:\nMGAEYGCLPSTSQALYVILLIVLVRMSLVFVQFPKYLRCYRCLLETKELGCLLGSDICLAPPGSSCMTLLIKNSSSGSDIMVSDCRHKEQMSDCSYTRSSPVFGFWIFSRCCLREFCNNPQNRVFYIP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01249", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPSAI_SPIOL\n\n### Answer:", "output": "Belongs to the PsaI family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01250", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSKLKITLKKSKIGRIEKHIRTCEALGLHKIGQSVIKEDNDAIRGMIRHIAFMVDVEEVK\n\n### Sequence 2:\nMATKKKTFEEAIAELETIVEALENGSASLEDSLDMYQKGIELTKLCQDKLQSAEQRMAKVVTEAGEEIPFEADGE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01251", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGAMT1_ARATH\n\n### Answer:", "output": "Methylates the carboxyl group of several gibberellins (GAs).  Substrate preference is GA9 > GA20 > GA3 > GA4 > GA34 > GA51 > GA1 > GA19 > GA12.  No activity with diterpenes abietic acid and ent-kaurenoic acid.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01252", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: VEGFR2\nLigand SMILES: Cc1nn(C)c2c1N(c1ccccc1)C(=O)CC(=O)N2C\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "Cc1nn(C)c2c1N(c1ccccc1)C(=O)CC(=O)N2C", "compound_name": null, "pubchem_cid": null, "MW": 284.32, "LogP": 1.76, "TPSA": 58.44, "QED": 0.75, "functional_groups": ["aromatic ring", "amide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_01253", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UJY1-K141N on protein Heat shock protein beta-8.", "input": "### Wild-type:\nMADGQMPFSCHYPSRLRRDPFRDSPLSSRLLDDGFGMDPFPDDLTASWPDWALPRLSSAWPGTLRSGMVPRGPTATARFGVPAEGRTPPPFPGEPWKVCVNVHSFKPEELMVKTKDGYVEVSGKHEEKQQEGGIVSKNFTKKIQLPAEVDPVTVFASLSPEGLLIIEAPQVPPYSTFGESSFNNELPQDSQEVTCT\n\n### Mutant:\nMADGQMPFSCHYPSRLRRDPFRDSPLSSRLLDDGFGMDPFPDDLTASWPDWALPRLSSAWPGTLRSGMVPRGPTATARFGVPAEGRTPPPFPGEPWKVCVNVHSFKPEELMVKTKDGYVEVSGKHEEKQQEGGIVSKNFTNKIQLPAEVDPVTVFASLSPEGLLIIEAPQVPPYSTFGESSFNNELPQDSQEVTCT\n\n### Answer:", "output": "In HMN2A; strengthen interaction with HSPB1; no effect on cytoskeleton architecture; no effect on cytoplasmic location; increased interaction with BAG3. Both mutations target the same amino acid, which is essential to the structural and functional integrity of the small heat-shock protein alphaA-crystallin. This positively charged residue, when mutated in other small heat-shock proteins, results in various human disorders. Coimmunoprecipitation experiments showed greater binding of both HSPB8 mutants to the interacting partner HSPB1. Expression of mutant HSPB8 in cultured cells promoted formation of intracellular aggregates. The novel c. 423G-->T () missense mutation of small heat-shock protein 22-kDa protein 8 (encoded by HSPB8) is responsible for distal hereditary motor neuropathy type (dHMN) II. The mutation in HSPB8 exhibited increased binding to Bag3.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01254", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "CLCA1, TFF3, MUC2", "output": "The expression pattern is consistent with goblet cell in a context associated with rheumatoid arthritis, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "goblet cell", "tissue": "lymph node", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CLCA1", "TFF3", "MUC2"]}}
{"id": "eval_01255", "category": "Mutation", "instruction": "Describe the effect of mutation P0C6S4-A13R on protein Disintegrin eristostatin.", "input": "### Wild-type:\nQEEPCATGPCCRACKFKRAGKVCRVARGDWNDDYCTGKSCDCPKNPWNG\n\n### Mutant:\nQEEPCATGPCCRRCKFKRAGKVCRVARGDWNDDYCTGKSCDCPKNPWNG\n\n### Answer:", "output": "No change in inhibition of wound closure, but gain in platelet aggregation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01256", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAPMTAPTAMYTVPFGLSAAFKINGFPAVDGGVILATISDEEDVAPLEVEKAEDVNDGRDLELVSDTTLDVGFDRAVELVVVVTKVGRSLLTSDTDGIKFFLTISADVSF\n\n### Sequence 2:\nMKGATLTRADLCEAVHEEVGLTRQDCAGLVERTLDLVAEALEQGETVKLSGFGVFQVRAKRARMGRNPKTGEPAEIEPRRVIGFRASQVMKARIDRALGG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01257", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BRAF\nLigand SMILES: O=C(c1ccc([N+](=O)[O-])cc1[N+](=O)[O-])N(N=CC=Cc1ccccc1)c1nnc(-c2ccccc2)c(-c2ccccc2)n1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and B-Raf kinase (BRAF) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "O=C(c1ccc([N+](=O)[O-])cc1[N+](=O)[O-])N(N=CC=Cc1ccccc1)c1nnc(-c2ccccc2)c(-c2ccccc2)n1", "compound_name": null, "pubchem_cid": null, "MW": 571.55, "LogP": 6.37, "TPSA": 157.62, "QED": 0.11, "functional_groups": ["aromatic ring", "amide", "nitro"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "BRAF", "has_evidence": false}}
{"id": "eval_01258", "category": "Mutation", "instruction": "Describe the effect of mutation Q16518-L510M on protein Retinoid isomerohydrolase.", "input": "### Wild-type:\nMSIQVEHPAGGYKKLFETVEELSSPLTAHVTGRIPLWLTGSLLRCGPGLFEVGSEPFYHLFDGQALLHKFDFKEGHVTYHRRFIRTDAYVRAMTEKRIVITEFGTCAFPDPCKNIFSRFFSYFRGVEVTDNALVNVYPVGEDYYACTETNFITKINPETLETIKQVDLCNYVSVNGATAHPHIENDGTVYNIGNCFGKNFSIAYNIVKIPPLQADKEDPISKSEIVVQFPCSDRFKPSYVHSFGLTPNYIVFVETPVKINLFKFLSSWSLWGANYMDCFESNETMGVWLHIADKKRKKYL\n\n### Mutant:\nMSIQVEHPAGGYKKLFETVEELSSPLTAHVTGRIPLWLTGSLLRCGPGLFEVGSEPFYHLFDGQALLHKFDFKEGHVTYHRRFIRTDAYVRAMTEKRIVITEFGTCAFPDPCKNIFSRFFSYFRGVEVTDNALVNVYPVGEDYYACTETNFITKINPETLETIKQVDLCNYVSVNGATAHPHIENDGTVYNIGNCFGKNFSIAYNIVKIPPLQADKEDPISKSEIVVQFPCSDRFKPSYVHSFGLTPNYIVFVETPVKINLFKFLSSWSLWGANYMDCFESNETMGVWLHIADKKRKKYL\n\n### Answer:", "output": "Does not affect isomerohydrolase activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01259", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nCOL1A1:low LUM:low DCN:medium PDGFRA:medium\n</cell>", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as DCN, PDGFRA, COL1A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "ovary", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["COL1A1", "LUM", "DCN", "PDGFRA"]}}
{"id": "eval_01260", "category": "Mutation", "instruction": "Describe the effect of mutation P21860-M899V on protein Receptor tyrosine-protein kinase erbB-3.", "input": "### Wild-type:\nMRANDALQVLGLLFSLARGSEVGNSQAVCPGTLNGLSVTGDAENQYQTLYKLYERCEVVMGNLEIVLTGHNADLSFLQWIREVTGYVLVAMNEFSTLPLPNLRVVRGTQVYDGKFAIFVMLNYNTNSSHALRQLRLTQLTEILSGGVYIEKNDKLCHMDTIDWRDIVRDRDAEIVVKDNGRSCPPCHEVCKGRCWGPGSEDCQTLTKTICAPQCNGHCFGPNPNQCCHDECAGGCSGPQDTDCFACRHFNDSGACVPRCPQPLVYNKLTFQLEPNPHTKYQYGGVCVASCPHNFVVDQTS\n\n### Mutant:\nMRANDALQVLGLLFSLARGSEVGNSQAVCPGTLNGLSVTGDAENQYQTLYKLYERCEVVMGNLEIVLTGHNADLSFLQWIREVTGYVLVAMNEFSTLPLPNLRVVRGTQVYDGKFAIFVMLNYNTNSSHALRQLRLTQLTEILSGGVYIEKNDKLCHMDTIDWRDIVRDRDAEIVVKDNGRSCPPCHEVCKGRCWGPGSEDCQTLTKTICAPQCNGHCFGPNPNQCCHDECAGGCSGPQDTDCFACRHFNDSGACVPRCPQPLVYNKLTFQLEPNPHTKYQYGGVCVASCPHNFVVDQTS\n\n### Answer:", "output": "Almost complete gain of ERBB2 and ERBB3 phosphorylation in the presence or in the absence of NRG1 stimulation, suggesting preservation of downstream signaling; does not affect the subcellular localization at the cell membrane. The consequences of the variant were evaluated using quantitative real-time PCR (RT-qPCR) on patient-derived fibroblasts or immunoblot assays on Neuro-2a cells overexpressing WT or mutant proteins, revealing either increased expression or unaltered phosphorylation of the mutant receptors.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01261", "category": "Mutation", "instruction": "Describe the effect of mutation P36639-L150A on protein Oxidized purine nucleoside triphosphate hydrolase.", "input": "### Wild-type:\nMGASRLYTLVLVLQPQRVLLGMKKRGFGAGRWNGFGGKVQEGETIEDGARRELQEESGLTVDALHKVGQIVFEFVGEPELMDVHVFCTDSIQGTPVESDEMRPCWFQLDQIPFKDMWPDDSYWFPLLLQKKKFHGYFKFQGQDTILDYTLREVDTV\n\n### Mutant:\nMGASRLYTLVLVLQPQRVLLGMKKRGFGAGRWNGFGGKVQEGETIEDGARRELQEESGLTVDALHKVGQIVFEFVGEPELMDVHVFCTDSIQGTPVESDEMRPCWFQLDQIPFKDMWPDDSYWFPLLLQKKKFHGYFKFQGQDTILDYTAREVDTV\n\n### Answer:", "output": "Reduces 2-oxo-dATPase and 8-oxo-dGTPase activities and increases thermolability.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01262", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMDSKTPVTLAKVIKVLGRTGSRGGVTQVRVEFLEDTSRTIVRNVKGPVRENDILVLMESEREARRLR\n\n### Sequence 2:\nMDSKTPVTLAKVIKVLGRTGSRGGVTQVRVEFLDDTSRTIVRNVKGPVRENDILVLMESEREARRLR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01263", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Z0S5-L68M on protein Claudin-15.", "input": "### Wild-type:\nMSVAVETFGFFMSALGLLMLGLTLSNSYWRVSTVHGNVITTNTIFENLWYSCATDSLGVSNCWDFPSLLALSGYVQGCRALMITAILLGFLGLFLGMVGLRCTNVGNMDLSKKAKLLAIAGTLHILAGACGMVAISWYAVNITTDFFNPLYAGTKYELGPALYLGWSASLLSILGGICVFSTCCCSSKEEPATRAGLPYKPSTVVIPRATSDESDISFGKYGKNAYV\n\n### Mutant:\nMSVAVETFGFFMSALGLLMLGLTLSNSYWRVSTVHGNVITTNTIFENLWYSCATDSLGVSNCWDFPSMLALSGYVQGCRALMITAILLGFLGLFLGMVGLRCTNVGNMDLSKKAKLLAIAGTLHILAGACGMVAISWYAVNITTDFFNPLYAGTKYELGPALYLGWSASLLSILGGICVFSTCCCSSKEEPATRAGLPYKPSTVVIPRATSDESDISFGKYGKNAYV\n\n### Answer:", "output": "Decreases formation of tight-junction strand-like structures. This mutation may enhance the structure and function of the tight junctions, potentially improving cell adhesion and paracellular barriers.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01264", "category": "Mutation", "instruction": "Describe the effect of mutation Q8IWA4-A245E on protein Mitofusin-1.", "input": "### Wild-type:\nMAEPVSPLKHFVLAKKAITAIFDQLLEFVTEGSHFVEATYKNPELDRIATEDDLVEMQGYKDKLSIIGEVLSRRHMKVAFFGRTSSGKSSVINAMLWDKVLPSGIGHITNCFLSVEGTDGDKAYLMTEGSDEKKSVKTVNQLAHALHMDKDLKAGCLVRVFWPKAKCALLRDDLVLVDSPGTDVTTELDSWIDKFCLDADVFVLVANSESTLMNTEKHFFHKVNERLSKPNIFILNNRWDASASAPEYMEDVRRQHMERCLHFLVEELKVVNALEAQNRIFFVSAKEVLSARKQKAQGMP\n\n### Mutant:\nMAEPVSPLKHFVLAKKAITAIFDQLLEFVTEGSHFVEATYKNPELDRIATEDDLVEMQGYKDKLSIIGEVLSRRHMKVAFFGRTSSGKSSVINAMLWDKVLPSGIGHITNCFLSVEGTDGDKAYLMTEGSDEKKSVKTVNQLAHALHMDKDLKAGCLVRVFWPKAKCALLRDDLVLVDSPGTDVTTELDSWIDKFCLDADVFVLVANSESTLMNTEKHFFHKVNERLSKPNIFILNNRWDASASEPEYMEDVRRQHMERCLHFLVEELKVVNALEAQNRIFFVSAKEVLSARKQKAQGMP\n\n### Answer:", "output": "Increases GTPase activity. Enables dimerization. Enhances mitochondrial fusion.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01265", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "Cc1ccc(S(=O)(=O)NC(=O)N[C@@H](C)Cc2ccccc2)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, sulfonamide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "Cc1ccc(S(=O)(=O)NC(=O)N[C@@H](C)Cc2ccccc2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 332.43, "LogP": 2.61, "TPSA": 75.27, "QED": 0.88, "functional_groups": ["aromatic ring", "amide", "sulfonamide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": null, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "sulfonamide"]}}
{"id": "eval_01266", "category": "Mutation", "instruction": "Describe the effect of mutation F8WQK8-R670D on protein Asp/Glu-specific dipeptidyl-peptidase.", "input": "### Wild-type:\nMNKRFFPTLLLAFVCSTLAYADGGMWLMQQINGQVARMKSLGMQLEAADIYNPNGSSLKDAVVMFDGGCTGVLVSNQGLLLTNHHCGYDQIQKHSSVQHNYLKDGFWSYSLAEELVNPGLEVEIVDEITDVTAAVKKELERIKKPSGLEFLSPRYLSSLAPEIVGKKAASRPGYRYEIKAFYGGNRYYMFTKKVFRDVRLVAAPPSSIGKFGSDTDNWAWPRHTGDFSIFRLYADKNGNPAEYSKDNVPYRPKRWVKVNAQGVKEGDFALIMGYPGTTYKFFTADEVTEWSEIDNNIRIE\n\n### Mutant:\nMNKRFFPTLLLAFVCSTLAYADGGMWLMQQINGQVARMKSLGMQLEAADIYNPNGSSLKDAVVMFDGGCTGVLVSNQGLLLTNHHCGYDQIQKHSSVQHNYLKDGFWSYSLAEELVNPGLEVEIVDEITDVTAAVKKELERIKKPSGLEFLSPRYLSSLAPEIVGKKAASRPGYRYEIKAFYGGNRYYMFTKKVFRDVRLVAAPPSSIGKFGSDTDNWAWPRHTGDFSIFRLYADKNGNPAEYSKDNVPYRPKRWVKVNAQGVKEGDFALIMGYPGTTYKFFTADEVTEWSEIDNNIRIE\n\n### Answer:", "output": "Loss of catalytic activity. Substitution of R(670) to D in DPP11 interacts with an acidic residue of the substrate.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01267", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "HNF4A, TTR, ALB", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as TTR, ALB, HNF4A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "brain", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["HNF4A", "TTR", "ALB"]}}
{"id": "eval_01268", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "C=CC#N", "output": "This molecule is small and relatively nonpolar, with hydrophilic character. It contains nitrile. Its descriptor profile: MW=53.06, LogP=0.7, TPSA=23.79, HBD=0, HBA=1, QED=0.37.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "C=CC#N", "compound_name": null, "pubchem_cid": null, "MW": 53.06, "LogP": 0.7, "TPSA": 23.79, "QED": 0.37, "functional_groups": ["nitrile"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01269", "category": "Mutation", "instruction": "Describe the effect of mutation Q14677-R146E on protein Clathrin interactor 1.", "input": "### Wild-type:\nMLNMWKVRELVDKATNVVMNYSEIESKVREATNDDPWGPSGQLMGEIAKATFMYEQFPELMNMLWSRMLKDNKKNWRRVYKSLLLLAYLIRNGSERVVTSAREHIYDLRSLENYHFVDEHGKDQGINIRQKVKELVEFAQDDDRLREERKKAKKNKDKYVGVSSDSVGGFRYSERYDPEPKSKWDEEWDKNKSAFPFSDKLGELSDKIGSTIDDTISKFRRKDREDSPERCSDSDEEKKARRGRSPKGEFKDEEETVTTKHIHITQATETTTTRHKRTANPSKTIDLGAAAHYTGDKASP\n\n### Mutant:\nMLNMWKVRELVDKATNVVMNYSEIESKVREATNDDPWGPSGQLMGEIAKATFMYEQFPELMNMLWSRMLKDNKKNWRRVYKSLLLLAYLIRNGSERVVTSAREHIYDLRSLENYHFVDEHGKDQGINIRQKVKELVEFAQDDDRLEEERKKAKKNKDKYVGVSSDSVGGFRYSERYDPEPKSKWDEEWDKNKSAFPFSDKLGELSDKIGSTIDDTISKFRRKDREDSPERCSDSDEEKKARRGRSPKGEFKDEEETVTTKHIHITQATETTTTRHKRTANPSKTIDLGAAAHYTGDKASP\n\n### Answer:", "output": "Abolished binding to VTI1B. Rescued binding to VTI1B R-23 mutant. The mutation in Vti1b affects its interaction with epsinR, but the functional consequences are not mentioned in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01270", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSDYTNTVTVEITSDDSLVADVIVVPVASGAVPRLPEDSKFRAYEDILQKLGVTGSKDELTRIPLDGSKHVVLAFIGVGKAFSATELMFAAGSAVRQIGARCIQIDFSVTQKDKLSAIVEGAFLGAYRFDKYRSKKSECPEVIRVSHNINGITDSECRQIIARAKVIAGSVGLAKDLVNTTGDDLYPAQFASFVAKDLEGIDHISVESWDEKRLQEKSCGGILGVGRGSNFPPRLVKISYTPGQYKKHLSLVGKGITFDTGGLSLKPASAMLGMKYDMTGAANVFAVLRIVALLRLSVRVTGWLCLAENMLSGSAIRPGDILRTYSGKTVEVTNTDAEGRLVLADGLALAGDERPDVIIDIATLTGAAKVALGESCSGLMGNNPDLLCSLEVAAKSVGEKFLAVPIDDDALQKALKSDIADIVNVPTSNKVPGMQFGAVFLKEFETTNDSGEAIPWAHLDVAGPANASYDTGFNSCGPTGVAVRSLVEFCRCLSSAPERKLF\n\n### Sequence 2:\nMGLLTGDALFSVAVAVAIFLLLVDLMHRRQRWAARYPPGPVPVPGLGNLLQVDFENMAYSCDKLRHQFGDVFSLQFVWTPVVVVNGLLAVREALVNNSTDTSDRPTLPTNALLGFGPKAQGVIGAYYGPAWREQRRFSVSSLRNFGLGKKSLEQWVTEEAACLCAAFTNHAGQPFCPKALLNKAVCNVISSLIYARRFDYDDPMVLRLLEFLEETLRENSSLKIQVLNSIPLLLRIPCVAAKVLSAQRSFIALNDKLLAEHNTGWAPDQPPRDLTDAFLTEMHKAQGNSESSFNDENLRLLVSDLFGAGMVTTSVTLSWALLLMILHPDVQRHVQEEIDEVIGQVRCPEMADQAHMPFTNAVIHEVQRFADIVPMGVPHMTSRDTEVQGFLIPKGTMLFTNLSSVLKDETVWEKPLHFHPGHFLDAEGRFVKREAFMPFSAGPRICLGEPLARMELFLFFTSLLQRFSFSVPEGQPRPSDRGAPYLVVLPSPYQLCAVLR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01271", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: ClCCCl\nDescriptors: MW=98.96, LogP=1.46, TPSA=0.0, HBD=0, HBA=0", "output": "Descriptors: MW=98.96, LogP=1.46, TPSA=0.0, HBD=0, HBA=0, QED=0.44, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "ClCCCl", "compound_name": null, "pubchem_cid": null, "MW": 98.96, "LogP": 1.46, "TPSA": 0.0, "QED": 0.44, "functional_groups": ["halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01272", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMATINQLVRKPRVKKVVKSNVPALEACPQKRGVCTRVYTTTPKKPNSALRKVCRIRLTNGFEVTSYIGGEGHNLQEHSVVLIRGGRVKDLPGVRYHTVRGALDCAGVKDRKQGRSKYGVKRPKS\n\n### Sequence 2:\nMPEPAKSAPAPKKGSKKAVTKAQKKDGKKRKRSRKESYSVYVYKVLKQVHPDTGISSKAMGIMNSFVNDIFERIAGEASRLAHYNKRSTITSREIQTAVRLLLPGELAKHAVSEGTKAVTKYTSSK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01273", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NRF2-F29R on protein SH2B adapter protein 1.", "input": "### Wild-type:\nMNGAPSPEDGASPSSPPLPPPPPPSWREFCESHARAAALDFARRFRLYLASHPQYAGPGAEAAFSRRFAELFLQHFEAEVARASGSLSPPILAPLSPGAEISPHDLSLESCRVGGPLAVLGPSRSSEDLAGPLPSSVSSSSTTSSKPKLKKRFSLRSVGRSVRGSVRGILQWRGTVDPPSSAGPLETSSGPPVLGGNSNSNSSGGAGTVGRGLVSDGTSPGERWTHRFERLRLSRGGGALKDGAGMVQREELLSFMGAEEAAPDPAGVGRGGGVAGPPSGGGGQPQWQKCRLLLRSEGEG\n\n### Mutant:\nMNGAPSPEDGASPSSPPLPPPPPPSWRERCESHARAAALDFARRFRLYLASHPQYAGPGAEAAFSRRFAELFLQHFEAEVARASGSLSPPILAPLSPGAEISPHDLSLESCRVGGPLAVLGPSRSSEDLAGPLPSSVSSSSTTSSKPKLKKRFSLRSVGRSVRGSVRGILQWRGTVDPPSSAGPLETSSGPPVLGGNSNSNSSGGAGTVGRGLVSDGTSPGERWTHRFERLRLSRGGGALKDGAGMVQREELLSFMGAEEAAPDPAGVGRGGGVAGPPSGGGGQPQWQKCRLLLRSEGEG\n\n### Answer:", "output": "Abolishes self-association and interaction with INSR and IGF1R. The mutation in SH2-B has been shown to cause activation of JAK2 and other kinases, leading to potentiation of cytokine and growth factor receptor signaling.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01274", "category": "Mutation", "instruction": "Describe the effect of mutation Q5BPF3-A86S on protein Protein BREAKING OF ASYMMETRY IN THE STOMATAL LINEAGE.", "input": "### Wild-type:\nMASQWTIPKLVTWRVRDWASCFLACKIPLDGDEDGANNNGNTTNNNNLTFKRIKRKIKSTKKKRSERKLSLSPPGTRHHHLHLRSASVSPTTSGSQHRRLSWPQPPVSEESGFIVFCFDREDGGFDVVKEGKQEKKETESSSEKSPRTVNRKLIYGDQGVGGTEKNNSPETKGTEQDQNDNTSCQGTKDVSSDVTERTKEEEDIDASDKSSGSSHSDEGRGSFAFPILGVEWMGSPAKMPESDDLSPKKQKPVALGFQCCRF\n\n### Mutant:\nMASQWTIPKLVTWRVRDWASCFLACKIPLDGDEDGANNNGNTTNNNNLTFKRIKRKIKSTKKKRSERKLSLSPPGTRHHHLHLRSSSVSPTTSGSQHRRLSWPQPPVSEESGFIVFCFDREDGGFDVVKEGKQEKKETESSSEKSPRTVNRKLIYGDQGVGGTEKNNSPETKGTEQDQNDNTSCQGTKDVSSDVTERTKEEEDIDASDKSSGSSHSDEGRGSFAFPILGVEWMGSPAKMPESDDLSPKKQKPVALGFQCCRF\n\n### Answer:", "output": "Increased ASK7-mediated phosphorylation; when associated with A-85, A-87, A-89 and A-91. The mutation in the GSK3 protein leads to a dysfunctional change in the phosphorylation activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01275", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPEILDKVDRRLLEELKNNARENIATLSKKLGIPRTTVHYRIKRLVEEGIIEKFTIKPNYKKLNLGTTAFILIRYDPDSGLTQREVAEQIARIPGVYEVHLVAGEWDLLLKVRASNAEEIGRIVIDKLREIRGVGQTVTMVSFVTVKEEI\n\n### Sequence 2:\nMAEILDKVDRRLLEELKINARENIATLSKKLGIPRTTVHYRIKRLVEEGIIERFTIKPNYKKLNLGTTAFILIRYDPDSGLTQRQVAEQIAKIPGVYEVHIIAGEWDLLLKVRASNAEEVGRIVIDKLREIRGVGQTVTMVSFDTVKEEI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01276", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRPOC_STRPI\n\n### Answer:", "output": "DNA-directed RNA polymerase subunit beta', RNAP subunit beta', RNA polymerase subunit beta', Transcriptase subunit beta'.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01277", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nFCGR3A CDKN1C MS4A7 LST1\n</cell>", "output": "This cell is most consistent with CD16+ monocyte, a non-classical CD16+ monocyte. Key markers such as FCGR3A, MS4A7, CDKN1C support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "prostate", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["FCGR3A", "CDKN1C", "MS4A7", "LST1"]}}
{"id": "eval_01278", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDSBB_NITOC\n\n### Answer:", "output": "Required for disulfide bond formation in some periplasmic proteins.  Acts by oxidizing the DsbA protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01279", "category": "Mutation", "instruction": "Describe the effect of mutation P99999-C18A on protein Cytochrome c.", "input": "### Wild-type:\nMGDVEKGKKIFIMKCSQCHTVEKGGKHKTGPNLHGLFGRKTGQAPGYSYTAANKNKGIIWGEDTLMEYLENPKKYIPGTKMIFVGIKKKEERADLIAYLKKATNE\n\n### Mutant:\nMGDVEKGKKIFIMKCSQAHTVEKGGKHKTGPNLHGLFGRKTGQAPGYSYTAANKNKGIIWGEDTLMEYLENPKKYIPGTKMIFVGIKKKEERADLIAYLKKATNE\n\n### Answer:", "output": "Decreased covalent heme attachment.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01280", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGPDA_STAAN\n\n### Answer:", "output": "Glycerol-3-phosphate dehydrogenase [NAD(P)+], NAD(P)H-dependent glycerol-3-phosphate dehydrogenase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01281", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nATPA_CLOB8\n\n### Answer:", "output": "Produces ATP from ADP in the presence of a proton gradient across the membrane.  The alpha chain is a regulatory subunit.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01282", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSHMVIPAGKLERVDPTTVRQEGPWADPAQAVVQTGPNQYTVYVLAFAFGYQPNPIEVPQGAEIVFKITSPDVIHGFHVEGTNINVEVLPGEVSTVRYTFKRPGEYRIICTPHPFMFGTIVVKE\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCTTTTTTSSTTSCGGGCEEEEETTEEEEEEEEETTEEESBSEEEETTSEEEEEEEBSSSCEEEEETTSSCEEEECBTCCEEEEEECCSCEEEEEECSSCTTCEEEEEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01283", "category": "Mutation", "instruction": "Describe the effect of mutation P73826-K151A on protein Acetoacetyl-CoA reductase.", "input": "### Wild-type:\nMLSLGLEDKVIVVTGGNRGIGAAIVKLLQEMGAKVAFTDLATDGGNTEALGVVANVTDLESMTAAAAEITDKLGPVYGVVANAGITKDNFFPKLTPADWDAVLNVNLKGVAYSIKPFIEGMYERKAGSIVAISSISGERGNVGQTNYSATKAGVIGMMKSLAREGARYGVRANAVAPGFIDTEMTLAIREDIREKITKEIPFRRFGKPEEIAWAVAFLLSPVASSYVTGEVLRVNGAHHT\n\n### Mutant:\nMLSLGLEDKVIVVTGGNRGIGAAIVKLLQEMGAKVAFTDLATDGGNTEALGVVANVTDLESMTAAAAEITDKLGPVYGVVANAGITKDNFFPKLTPADWDAVLNVNLKGVAYSIKPFIEGMYERKAGSIVAISSISGERGNVGQTNYSATAAGVIGMMKSLAREGARYGVRANAVAPGFIDTEMTLAIREDIREKITKEIPFRRFGKPEEIAWAVAFLLSPVASSYVTGEVLRVNGAHHT\n\n### Answer:", "output": "5% enzymatic activity. The mutation in the FabG protein results in a change in its substrate selectivity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01284", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMPNTPDRKYKRILLKLSGEQLMGDEGFGIDPKVLDKMALEIGQLVGIGVQVGLVIGGGNLFRGAALSRAGLDRVTGDHMGMLATVMNALAMRDALERSNISSTVMSAIPMHGVTDQYDRRKALRLLDNGEVVIFSAGTGNPFFTTDSAACLRGIEVNAELVLKATKVDGVYSADPMKDPTAVRYSKLTYEKVLTEQLGVMDLTAICLCQDHNMPVRVFQMDKQGALLNIVVGGDEGTLIQGKETEEA\n\n### Sequence 2:\nMSETTHTAIDTAYPGNLFMVVAPSGAGKSTLVNALLAQDKAIRLSISHTTRSPRPGEQNGREYHFISVDEFRAARDRGEFLEWAEVHGNYYATSRVWIEEQMAQGTDVLLEIDWQGAQQVHKRFSNAVEIFILPPSLTALEERLKKRGQDEPNVIVRRLLAAGSEMSHASESDYVIINEVFDDALKQLQNVVHATRLRFSSQKARHAELFIELGIH\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01285", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSTSELATSYAALILADDGVDITADKLQSLIKAAKIEEVEPIWTTLFAKALEGKDVKDLLLNVGSGGGAAPLPEALLLRWRAADAAPAAEEKKEEEKEESDEDMGFGLFD\n\n### Sequence 2:\nMPPKSRQGAGRKVRRKEKKNVVHGHAHIKSTFNNTIVSITDPTGAVISWASAGQVGFKGSRKSTPFAAQMAAEAAARRAQEHGVRKVDVFVKGPGSGRETAIRSLQATGLEVGSIQDVTPVPHNGCRPPKRRRV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01286", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: C[C@@H]1O[C@@H](O[C@@H]2[C@@H](O)[C@H](OCCc3ccc(O)c(O)c3)O[C@H](CO[C@@H]3O[C@H](CO)[C@@H](O)[C@H](O)[C@H]3O)[C@H]2OC(=O)/C=C/c2ccc(O)c(O)c2)[C@H](O)[C@H](O)[C@H]1O\nDescriptors: MW=786.73, LogP=-3.19, TPSA=324.44, HBD=12, HBA=20", "output": "Descriptors: MW=786.73, LogP=-3.19, TPSA=324.44, HBD=12, HBA=20, QED=0.05, Lipinski violations=3. Overall the molecule is somewhat outside the typical oral drug-like ranges, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "C[C@@H]1O[C@@H](O[C@@H]2[C@@H](O)[C@H](OCCc3ccc(O)c(O)c3)O[C@H](CO[C@@H]3O[C@H](CO)[C@@H](O)[C@H](O)[C@H]3O)[C@H]2OC(=O)/C=C/c2ccc(O)c(O)c2)[C@H](O)[C@H](O)[C@H]1O", "compound_name": null, "pubchem_cid": null, "MW": 786.73, "LogP": -3.19, "TPSA": 324.44, "QED": 0.05, "functional_groups": ["aromatic ring", "carboxylic acid", "ester", "hydroxyl", "ether", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01287", "category": "Mutation", "instruction": "Describe the effect of mutation P24058-K325N on protein Argininosuccinate lyase.", "input": "### Wild-type:\nMASEARGDKLWGGRFSGSTDPIMEKLNSSIAYDQRLSEVDIQGSMAYAKALEKAGILTKTELEKILSGLEKISEEWSKGVFVVKQSDEDIHTANERRLKELIGDIAGKLHTGRSRNDQVVTDLKLFMKNSLSIISTHLLQLIKTLVERAAIEIDVILPGYTHLQKAQPIRWSQFLLSHAVALTRDSERLGEVKKRINVLPLGSGALAGNPLDIDREMLRSELEFASISLNSMDAISERDFVVEFLSFATLLMIHLSKMAEDLIIYSTSEFGFLTLSDAFSTGSSLMPQKKNPDSLELIRS\n\n### Mutant:\nMASEARGDKLWGGRFSGSTDPIMEKLNSSIAYDQRLSEVDIQGSMAYAKALEKAGILTKTELEKILSGLEKISEEWSKGVFVVKQSDEDIHTANERRLKELIGDIAGKLHTGRSRNDQVVTDLKLFMKNSLSIISTHLLQLIKTLVERAAIEIDVILPGYTHLQKAQPIRWSQFLLSHAVALTRDSERLGEVKKRINVLPLGSGALAGNPLDIDREMLRSELEFASISLNSMDAISERDFVVEFLSFATLLMIHLSKMAEDLIIYSTSEFGFLTLSDAFSTGSSLMPQKKNPDSLELIRS\n\n### Answer:", "output": "99% decrease in catalytic efficiency.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01288", "category": "Mutation", "instruction": "Describe the effect of mutation G4SW86-F241Y on protein Delta.", "input": "### Wild-type:\nMSEQESRDNAAVDAVRQKYGFGFSWLVLMIALPPLVYYLWICVTYYQGELVFTSDAAAWRRFWSHVAPPTWHAAGLYAAWFLGQAALQVWAPGPTVQGMKLPDGSRLDYRMNGIFSFLFTLAVVFGLVTMGWLDATVLYDQLGPLLTVVNIFTFVFAGFLYFWGLNGKQWERPTGRPFYDYFMGTALNPRIGSLDLKLFCEARPGMIFWLLMNLSMAAKQYELHGTVTVPMLLVVGFQSFFLIDYFIHEEAVLTTWDIKHEKFGWMLCWGDLVWLPFTYTLQAQYLVHHTHDLPVWGIIA\n\n### Mutant:\nMSEQESRDNAAVDAVRQKYGFGFSWLVLMIALPPLVYYLWICVTYYQGELVFTSDAAAWRRFWSHVAPPTWHAAGLYAAWFLGQAALQVWAPGPTVQGMKLPDGSRLDYRMNGIFSFLFTLAVVFGLVTMGWLDATVLYDQLGPLLTVVNIFTFVFAGFLYFWGLNGKQWERPTGRPFYDYFMGTALNPRIGSLDLKLFCEARPGMIFWLLMNLSMAAKQYELHGTVTVPMLLVVGFQSFYLIDYFIHEEAVLTTWDIKHEKFGWMLCWGDLVWLPFTYTLQAQYLVHHTHDLPVWGIIA\n\n### Answer:", "output": "Gain of catalytic activity; when associated with A-363.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01289", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BV86-W20Y on protein N-terminal Xaa-Pro-Lys N-methyltransferase 1.", "input": "### Wild-type:\nMTSEVIEDEKQFYSKAKTYWKQIPPTVDGMLGGYGHISSIDINSSRKFLQRFLREGPNKTGTSCALDCGAGIGRITKRLLLPLFREVDMVDITEDFLVQAKTYLGEEGKRVRNYFCCGLQDFTPEPDSYDVIWIQWVIGHLTDQHLAEFLRRCKGSLRPNGIIVIKDNMAQEGVILDDVDSSVCRDLDVVRRIICSAGLSLLAEERQENLPDEIYHVYSFALR\n\n### Mutant:\nMTSEVIEDEKQFYSKAKTYYKQIPPTVDGMLGGYGHISSIDINSSRKFLQRFLREGPNKTGTSCALDCGAGIGRITKRLLLPLFREVDMVDITEDFLVQAKTYLGEEGKRVRNYFCCGLQDFTPEPDSYDVIWIQWVIGHLTDQHLAEFLRRCKGSLRPNGIIVIKDNMAQEGVILDDVDSSVCRDLDVVRRIICSAGLSLLAEERQENLPDEIYHVYSFALR\n\n### Answer:", "output": "Nearly abolishes methyltransferase activity with CENPA. This mutation may affect the substrate recognition and catalytic activity of NRMT1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01290", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "TTR, ALB, HNF4A, APOB", "output": "This profile is most consistent with hepatocyte, which is commonly found in breast under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "breast", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TTR", "ALB", "HNF4A", "APOB"]}}
{"id": "eval_01291", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nEX7S_CHLMU\n\n### Answer:", "output": "Bidirectionally degrades single-stranded DNA into large acid-insoluble oligonucleotides, which are then degraded further into small acid-soluble oligonucleotides.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01292", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "COL1A1, LUM, DCN, COL3A1, PDGFRA, COL1A2", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as DCN, COL1A1, COL1A2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "spleen", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["COL1A1", "LUM", "DCN", "COL3A1", "PDGFRA", "COL1A2"]}}
{"id": "eval_01293", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPGSIPLIGERFPEMEVTTDHGVIKLPDHYVSQGKWFVLFSHPADFTPVCTTEFVSFARRYEDFQRLGVDLIGLSVDSVFSHIKWKEWIERHIGVRIPFPIIADPQGTVARRLGLLHAESATHTVRGVFIVDARGVIRTMLYYPMELGRLVDEILRIVKALKLGDSLKRAVPADWPNNEIIGEGLIVPPPTTEDQARARMESGQYRSLDWWFCWDTPASRDDVEEARRYLRRAAEKPAKLLYEEARTHLH\n\n### Answer:", "output": "<SEQ_2D>CEEECCTTSBCCCEEEEETTEEEEETHHHHTTTCEEEEECCSCTTCHHHHHHHHHHHHTHHHHHHTTEEEEEEESSCHHHHHHHHHHHHHHTCCCCCSCEEECTTCHHHHHTTCCCTTCSSSCCEEEEEECTTSBEEEEEEECTTBCCCHHHHHHHHHHHHHHHHHTCBBCTTTTSBTTTBTCEECCCCCSHHHHHHHHHHCCSEEEETTEEEECCSCHHHHHHHHHHHHHHHSCCSSCHHHHCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01294", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTHIG_SHEPC\n\n### Answer:", "output": "Belongs to the ThiG family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01295", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "COCC1CN=C(c2ccccc2Cl)c2cc(Br)ccc2N1C", "output": "This molecule is medium-sized and relatively nonpolar, with lipophilic character. It contains aromatic ring, ether, halide. Its descriptor profile: MW=393.71, LogP=4.4, TPSA=24.83, HBD=0, HBA=3, QED=0.77.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "COCC1CN=C(c2ccccc2Cl)c2cc(Br)ccc2N1C", "compound_name": null, "pubchem_cid": null, "MW": 393.71, "LogP": 4.4, "TPSA": 24.83, "QED": 0.77, "functional_groups": ["aromatic ring", "ether", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01296", "category": "Mutation", "instruction": "Describe the effect of mutation P03971-M546A on protein Muellerian-inhibiting factor.", "input": "### Wild-type:\nMRDLPLTSLALVLSALGALLGTEALRAEEPAVGTSGLIFREDLDWPPGSPQEPLCLVALGGDSNGSSSPLRVVGALSAYEQAFLGAVQRARWGPRDLATFGVCNTGDRQAALPSLRRLGAWLRDPGGQRLVVLHLEEVTWEPTPSLRFQEPPPGGAGPPELALLVLYPGPGPEVTVTRAGLPGAQSLCPSRDTRYLVLAVDRPAGAWRGSGLALTLQPRGEDSRLSTARLQALLFGDDHRCFTRMTPALLLLPRSEPAPLPAHGQLDTVPFPPPRPSAELEESPPSADPFLETLTRLVRA\n\n### Mutant:\nMRDLPLTSLALVLSALGALLGTEALRAEEPAVGTSGLIFREDLDWPPGSPQEPLCLVALGGDSNGSSSPLRVVGALSAYEQAFLGAVQRARWGPRDLATFGVCNTGDRQAALPSLRRLGAWLRDPGGQRLVVLHLEEVTWEPTPSLRFQEPPPGGAGPPELALLVLYPGPGPEVTVTRAGLPGAQSLCPSRDTRYLVLAVDRPAGAWRGSGLALTLQPRGEDSRLSTARLQALLFGDDHRCFTRMTPALLLLPRSEPAPLPAHGQLDTVPFPPPRPSAELEESPPSADPFLETLTRLVRA\n\n### Answer:", "output": "Enables AMH signaling.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01297", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nARCH_PYRIL\n\n### Answer:", "output": "Activates the tRNA-splicing ligase complex by facilitating the enzymatic turnover of catalytic subunit RtcB.  Acts by promoting the guanylylation of RtcB, a key intermediate step in tRNA ligation.  Can also alter the NTP specificity of RtcB such that ATP, dGTP or ITP is used efficiently (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01298", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMLERIKDSFTESIQTKIDAAEALPESIAKAAEMMVHCLLGGNKILACGNGGSAGDAQHFSAELLNRYEIERPPLPAIALSTDTSTITAIANDYSYDEIFSKQIFALGQPGDILLAISTSGNSGNVIKAMEAALSRDMTIVALTGKDGGAMAGLLSVGDVEIRVPSNVTARIQEVHLLVIHCLCDNIDRTLFPQDEQQ\n\n### Sequence 2:\nMADQLTDEQIAEFKEAFSLFDKDGDGCITTKELGTVMRSLGQNPTEAELQDMINEVDADGNGTIDFPELLNLMARKMKDTDSEEELKEAFRVFDKDQNGFISAAELRHVMTNLGEKLTDEEVDEMIREADVDGDGQINYEEFVKVMMAK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01299", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nMS4A1 CD79B IGHM IGHD\n</cell>", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as IGHD, CD79B, MS4A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "intestine", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MS4A1", "CD79B", "IGHM", "IGHD"]}}
{"id": "eval_01300", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMERIKELRDLMSQSRTRGILTKTTVDHMAIIKKYTSGRQEKNPALRMKWMMAMKYPITADKRIMEMIPERNEQGQTLWSKTNDAGSDRVMVSPLAVTWWNRNGPTTSTVHYPKVHKTYFEKVERLKHGTFGPVHFRNQVKIRRRVDINPGHADLSAKEAQDVIMEVVFPNEVGARILTSESQLTITKEKKEELQDCKIAPLMVAYMLERELVRKTRFLPVAGGTSSVYIEVLHLTQGTCWEQMYTPGGEVRNDDVDQSLIIAARNIVRRATVSADPLASLLEMCHSTQIGGIRMVDILRQNPTEEQAVDICKAAMGLRISSSFSFGGFTFKRTSGSSIKREEEVLTGNLQTLKIRVHEGYEEFTMVGRRATAILRKATRRLIQLIVSGRDEQSIAEAIIVAMVFSQEDCMIKAVRGDLNFVNRANQRLNPMHQLLRHFQKDAKVLFQNWGIEPIDNVMGMIGILPDMTPNAEMSLRGIRVSKMGVDEYSSTERVVVSIDRFLRVRDQRGNVLLSPEEVSETQGTEKLTITYSSSMMWEINGPESVLVNTYQWIIRNWEMIKIQWSQDPTMLYNKMEFEPFQSLVPKAARAQYSGFVRTLFQQMRDVLGTFDTVQIIKLLPFAAAPPEQSRMQFSSLTVNVRGSGMRVLVRGNSPVFNYNKATKRLTVLGKDAGSLTEDPDEGTAGVESAVLRGFLILGKEDKRYGPALSINELSNLAKGEKANVLIGQGDVVLVMKRKRDSSILTDSQTATKRIRMAIN\n\n### Sequence 2:\nMQSIFNNKYQFKNIETKYNTLWDTTKLYKWKNSGTNQFVIDTPPPTISGQLHIGHVFSYCHTDFIARYQRMLGKDVFYPIGFDDNGLPTERLVEKTKKIRATDISRKEFKTICTQVSHEFRIQFKQLFQSIGISYDWDLEYHTISKDIQKISQTSFINLYNKGKLYRKLQPIFWDCIDKTAIARAEVEENELSSFMNTIAFSTEAGKAINIATTRPELMPACVAVFFNPSDIRYQDLLGQNAIVPIFGNKVKILSDDQVKIDKGTGLVMCCTFGDEMDVYWWNKHNLDTKIIISKSGTIDHLNTLQGQDVCKQLHGLSITEARALIIEILERNNLLIQKQEITHNVKCAERSGAPIEILLSHQWFIKVVDIKHELLKQVQKINWHPQSMRKQIEIWIEGLNWDWCISRQRYFGVPFPVWYSKDGKIILPDINKLPIDPTNDLPEGYQDTEIEVETDVMDTWATSSLSTQFHNISATPADLRAQSHEIIRSWAFYTILQAYYHNNDIPWKNIMISGWCLAEDKTKMSKSKGNVLTPNKLLDEYGADVVRYWTANSKLGADTTFSNEILKLGKRFTTKLWNASKFVSMFIDQYSEPDLQYITETMDKWILSKLYKVIVKATESFNSFEYCIALDCIESFFWKDFCDNYLELSKKRAYGELISKQEHLSAVNTLSFVLRELLKMLAPFMPYVTEEIYRTLYSSNNSIHSHNTWPAADVNLYNESDELLGETFIEILNQVRKVKASAQLSVKYKINKLIINKHFPVSLENDLKAVCNADCIVYDNRQNDNKEQLLVSVEFENVQIT\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01301", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nATPD_CHLP8\n\n### Answer:", "output": "Belongs to the ATPase delta chain family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01302", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UKT5-S8R on protein F-box only protein 4.", "input": "### Wild-type:\nMAGSEPRSGTNSPPPPFSDWGRLEAAILSGWKTFWQSVSKERVARTTSREEVDEAASTLTRLPIDVQLYILSFLSPHDLCQLGSTNHYWNETVRDPILWRYFLLRDLPSWSSVDWKSLPDLEILKKPISEVTDGAFFDYMAVYRMCCPYTRRASKSSRPMYGAVTSFLHSLIIQNEPRFAMFGPGLEELNTSLVLSLMSSEELCPTAGLPQRQIDGIGSGVNFQLNNQHKFNILILYSTTRKERDRAREEHTSAVNKMFSRHNEGDDQQGSRYSVIPQIQKVCEVVDGFIYVANAEAHKR\n\n### Mutant:\nMAGSEPRRGTNSPPPPFSDWGRLEAAILSGWKTFWQSVSKERVARTTSREEVDEAASTLTRLPIDVQLYILSFLSPHDLCQLGSTNHYWNETVRDPILWRYFLLRDLPSWSSVDWKSLPDLEILKKPISEVTDGAFFDYMAVYRMCCPYTRRASKSSRPMYGAVTSFLHSLIIQNEPRFAMFGPGLEELNTSLVLSLMSSEELCPTAGLPQRQIDGIGSGVNFQLNNQHKFNILILYSTTRKERDRAREEHTSAVNKMFSRHNEGDDQQGSRYSVIPQIQKVCEVVDGFIYVANAEAHKR\n\n### Answer:", "output": "In esophagus cancer samples.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01303", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTRPA_GLUDA\n\n### Answer:", "output": "Tryptophan synthase alpha chain.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01304", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SHU7-G205E on protein Protein TIC 21, chloroplastic.", "input": "### Wild-type:\nMQSLLLPPASSSGVSAVALRPGFQHSFNHQSLSTRSLPLFNPLLLAPKKKTTISSYQSPPSLSVYGFQIGGSKPSFTPSTVAFSYPTSPSSVPGDNEVDKAKLAQVAKRLEKTSRYFKRLGSIGFWGQLVSTVVAAVILSFSIVVTGKPTSPATFYATASGIAAAFVSVFWSFGYIRLSERLRRTSIDPAKAPPRADVVKGLRSGIMVNILGMGSALLGMQATVGFLVAKALTTSANPFYQGVSQGYSPVLALDVFLVQASANTLLSHFLGLVCSLELLRSVTVPNSESVVVPKVA\n\n### Mutant:\nMQSLLLPPASSSGVSAVALRPGFQHSFNHQSLSTRSLPLFNPLLLAPKKKTTISSYQSPPSLSVYGFQIGGSKPSFTPSTVAFSYPTSPSSVPGDNEVDKAKLAQVAKRLEKTSRYFKRLGSIGFWGQLVSTVVAAVILSFSIVVTGKPTSPATFYATASGIAAAFVSVFWSFGYIRLSERLRRTSIDPAKAPPRADVVKGLRSEIMVNILGMGSALLGMQATVGFLVAKALTTSANPFYQGVSQGYSPVLALDVFLVQASANTLLSHFLGLVCSLELLRSVTVPNSESVVVPKVA\n\n### Answer:", "output": "In cia5-1; albino plants.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01305", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMNQAENRIDQILKRGVAEVIVEEDFKKLLLSGKKLRLKEGFDPSSPDIHLGHMVALRKLRQLQDLGHQVVLIVGDWTAQIGDPSGASVTRPMLSAEQVKANAKTYLEQFFKIVDKDKTEVRWQSEWYANFKLEDVVRLSSKFTVAQMLARDDFAKRYAAGKPISVTELLYPMLQAYDSVMVKSDVEFGGTDQKFNLLVGRELQEMLGQKPQQVLMVPILVGTDGVHKMSKSLGNYIGVAEEPTEIFGKCMSIPDELILQYFELVTDIPATEIADFKNQMANGLVNPMVLKKRLARELITQLYSAAAAEEADAQFTRVVQRGEIPEDMPECRLENGESICVIDFIIKADLAKSKSEARRLLEQGGVEINSAKISDPGTTVKCGDIIKAGKRRYSKAV\n\n### Sequence 2:\nMNQSENQIDQILKRGVAEVIVEEDFKKLLLSGRKLRLKEGFDPSSPDIHLGHMVALRKLRQLQDLGHQVVLIVGDWTAQIGDPSGASVTRPMLSAEQVKANAKTYLEQFFKIVDKDKTEVRWQSEWYGNFKLEDVVRLSSKFTVAQMLARDDFAKRYAAGKPISVTELLYPMLQAYDSVMVKSDVEFGGTDQKFNLLVGRELQEMVGQKPQQVLMVPILVGTDGVHKMSKSLGNYIGVAEDPSEIFGKCMSIPDELILQYFELVTDIPDQEIADFKAQMENGQVNPMILKKRLASELLTQLYNATAAQEADARFTRVVQRGEIPEDMPECRLENGQNTGVIDFIILSGLAKSKSEARRLLEQGAVEINSEKISDQNTPVKCGDIIKAGKRRYSKAI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01306", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CCCCCC#COC1CC(C)CCC1C(C)C\nDescriptors: MW=250.43, LogP=5.01, TPSA=9.23, HBD=0, HBA=1", "output": "Descriptors: MW=250.43, LogP=5.01, TPSA=9.23, HBD=0, HBA=1, QED=0.49, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "CCCCCC#COC1CC(C)CCC1C(C)C", "compound_name": null, "pubchem_cid": null, "MW": 250.43, "LogP": 5.01, "TPSA": 9.23, "QED": 0.49, "functional_groups": ["ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01307", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CCC(C)C1(NC(C)=O)CCN(C(CCc2ccccc2)C(=O)NC(Cc2cc(F)cc(F)c2)C(O)C2CC(OCc3cccnc3)C[NH2+]2)C1=O", "output": "The SMILES represents an unnamed molecule with MW=706.85, LogP=2.82, TPSA=137.47. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "BACE", "canonical_smiles": "CCC(C)C1(NC(C)=O)CCN(C(CCc2ccccc2)C(=O)NC(Cc2cc(F)cc(F)c2)C(O)C2CC(OCc3cccnc3)C[NH2+]2)C1=O", "compound_name": null, "pubchem_cid": null, "MW": 706.85, "LogP": 2.82, "TPSA": 137.47, "QED": 0.18, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether", "halide", "pyridine"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01308", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRSAQVYRWQIPMDAGVVLRDRRLKTRDGLYVCLREGECEGWGEISPLPGFSQETWEEAQCVLLAWVNNWLAGDCELPQMPSVAFGVSCALAELADTLPQAANFRAAPLCNGDPDDLILKLADMPGEKVAKVKVGLYEAVRDGMVVNLLLEAIPDLHLRLDANRAWTPLKGQQFAKYVNPDYRGRIAFLEEPCKTREDSRAFARETGIAIAWDESLREPDFAFVAEEGVRAVVIKPTLTGSLEKVHEQVQAAHALGLTAVISSSIESSLGLTQLARIAAWLTPDTIPGLDTLDLMQAQQVRRWPGSTLPFVDVDALERLL\n\n### Sequence 2:\nMTKESPIPVVVNGAAGKMGKEVIKAVSQAEDMMLVGAVDLNPKYRGEDAGEVAGCGPVEVPILDDLQSVLVLATQEKVQGVMVDFTHPDGVYDNIRSAIAYGVRPVVGTTGLSGEQIKDLGEFAEKASTGCLIVPNFSIGMILLQQAAVQASQYFDHVEIIELHHNQKADAPSGTAIKTAEMLSELGKSFNPAKVEETETIRGSRGGITEENIHIHSVRLPGFIAHQEVIFGSTGQIYTLRHDTTDRSSFMPGVLLAIRKVTQLKSLIYGLEKIL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01309", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "HAVCR2, PDCD1, LAG3, TOX", "output": "The expression pattern is consistent with exhausted CD8 T cell in a context associated with inflammatory bowel disease, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "kidney", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["HAVCR2", "PDCD1", "LAG3", "TOX"]}}
{"id": "eval_01310", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "CD1C, HLA-DQA1, CLEC10A", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as CLEC10A, HLA-DQA1, CD1C support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "kidney", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD1C", "HLA-DQA1", "CLEC10A"]}}
{"id": "eval_01311", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMKDTITSLINPADEKGSYLDAAALEQLNRYFQSGNMRVKAAKTISSSASSIISKTVAKSLLYGDITLPGGNMYPTRRYAACLRDLTYFLRYATYAMLAADPSILDERVLQGLKETYITLGVPIDRVIQALNAMKEVLTESLDTEASQEMAVYLDHIIAGL\n\n### Answer:", "output": "<SEQ_3Di>DAFLVRVQVVVCVVVVHDRDDVSVVVVVVVVVVPPQLVVLLVLCVVCVLVLLVQLCVPQPPDDCLDPPVHQCPDVVSVVVVSVLVSVLSVQLSVCSSRVHLPSVVVPPLDCNVVVCVVVVPDLVSVLRSLVSSLVSSCVRGDHVSSVSSVVNSVVVNVSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01312", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "HPGDS, TPSAB1, CPA3, TPSB2, KIT", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as TPSAB1, HPGDS, CPA3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "intestine", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["HPGDS", "TPSAB1", "CPA3", "TPSB2", "KIT"]}}
{"id": "eval_01313", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: CCCCCCCCCCCCCC(=O)OCC(O)CO\nDescriptors: MW=302.45, LogP=3.58, TPSA=66.76, HBD=2, HBA=4", "output": "Descriptors: MW=302.45, LogP=3.58, TPSA=66.76, HBD=2, HBA=4, QED=0.36, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "CCCCCCCCCCCCCC(=O)OCC(O)CO", "compound_name": null, "pubchem_cid": null, "MW": 302.45, "LogP": 3.58, "TPSA": 66.76, "QED": 0.36, "functional_groups": ["ester", "hydroxyl", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01314", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSALVYFSSSSENTHRFMQRLGLPATRIPLNERERIQVDEPYILVVPSYGGGMAGAVPRQVIRFLNDEHNRARIRGVIASGNRNFGDAWGCAGDVIAQKCGVPWLYRFELMGTQRDIDNVRKGVNEFWQQLSRSA\n\n### Sequence 2:\nMTLKALAQSLGITLKYLFSKPVTVPYPDAPVALKPRFHGRHVLTRHPNGLEKCIGCSLCAAACPAYAIYVEPAENDPENPVSAGERYAKVYEINMLRCIFCGLCEEACPTGAIVLGYDFEMADYEYSDLVYGKEDMLVDVVGTKPQRREAKRTGKPVKVGYVVPYVRPELEGFKAPTEGGKR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01315", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDGTP_SALSV\n\n### Answer:", "output": "Deoxyguanosinetriphosphate triphosphohydrolase, dGTP triphosphohydrolase, dGTPase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01316", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSPF27_ARATH\n\n### Answer:", "output": "Belongs to the SPF27 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01317", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCYSI_SALDC\n\n### Answer:", "output": "Sulfite reductase [NADPH] hemoprotein beta-component, SiR-HP, SiRHP.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01318", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "HPGDS, TPSB2, CPA3, TPSAB1, KIT", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as CPA3, TPSAB1, HPGDS support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "prostate", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["HPGDS", "TPSB2", "CPA3", "TPSAB1", "KIT"]}}
{"id": "eval_01319", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMVAIKNLVLVALTAVTALAMPSPLEERAATWTCMNEQKNPKTNKYENKRLLYNQNNAESNAHHAPLSDGKTGSSYPHWFTNGYDGDGKILKGRTPIKWGNSDCDRPPKHSKNGDGKNDHYLLEFPTFPDGHQYNFDSKKPKEDPGPARVIYTYPNKVFCGIVAHTRENQGDLKLCSH\n\n### Sequence 2:\nMVAIKNLFLLAATAVSVLAAPSPLDARA-TWTCINQQLNPKTNKWEDKRLLYSQAKAESNSHHAPLSDGKTGSSYPHWFTNGYDGNGKLIKGRTPIKFGKADCDRPPKHSQNGMGKDDHYLLEFPTFPDGHDYKFDSKKPKEDPGPARVIYTYPNKVFCGIVAHQRGNQGDLRLCSH\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01320", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRS21_PEDPA\n\n### Answer:", "output": "30S ribosomal protein S21.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01321", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nCD79A:medium IGHM:high CD79B:high TCL1A:low\n</cell>", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as CD79B, CD79A, IGHM support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "intestine", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD79A", "IGHM", "CD79B", "TCL1A"]}}
{"id": "eval_01322", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMICA2_MOUSE\n\n### Answer:", "output": "[F-actin]-monooxygenase MICAL2, ERK2-binding testicular protein 1, MICAL C-terminal-like protein, Mical-cL, Molecule interacting with CasL protein 2, MICAL-2, mMical2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01323", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "LEF1, CD3D, CD8B, SELL, CD8A", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as SELL, CD3D, LEF1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "skin", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["LEF1", "CD3D", "CD8B", "SELL", "CD8A"]}}
{"id": "eval_01324", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nNKAI1_MOUSE\n\n### Answer:", "output": "Cell membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01325", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAVKKTKSKQKVKRKNIVSGVAHIHSTHQNTIVAFSDEAGNVFAWSSAGAIGYKGTKKKTPYAAGLATTAAVEKAKEHGLKEVRVELKGTGSGKDAARKQIEALGIIIKEVKDVTPIPHNGTRPPKKVLKRDLK\n\n### Sequence 2:\nMSTTGKGGKAKGKTASSKQVSRSARAGLQFPVGRISRFLKHGRYSERVGTGAPVYLAAVLEYLAAEVLELAGNAAKDNKKTRIVPRHILLAIRNDEELNKLMANTTIADGGVLPNINPMLLPSKSKKTESRGQASQDL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01326", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CCOP(=O)(OCC)C(C#N)=Cc1cccc(F)c1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, nitrile, halide. Its descriptor profile: MW=283.24, LogP=3.96, TPSA=59.32, HBD=0, HBA=4, QED=0.59.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "CCOP(=O)(OCC)C(C#N)=Cc1cccc(F)c1", "compound_name": null, "pubchem_cid": null, "MW": 283.24, "LogP": 3.96, "TPSA": 59.32, "QED": 0.59, "functional_groups": ["aromatic ring", "nitrile", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01327", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSVGIVYGDQYRQLCCSSPKFGDRYALVMDLINAYKLIPELSRVPPLQWDSPSRMYEAVTAFHSTEYVDALKKLQMLHCEEKELTADDELLMDSFSLNYDCPGFPSVFDYSLAAVQGSLAAASALICRHCEVVINWGGGWHHAKRSEASGFCYLNDIVLAIHRLVSSTPPETSPNRQTRVLYVDLDLHHGDGVEEAFWYSPRVVTFSVHHASPGFFPGTGTWNMVDNDKLPIFLNGAGRGRFSAFNLPLEEGINDLDWSNAIGPILDSLNIVIQPSYVVVQCGADCLATDPHRIFRLTNFYPNLNLDSDCDSECSLSGYLYAIKKILSWKVPTLILGGGGYNFPDTARLWTRVTALTIEEVKGKKMTISPEIPEHSYFSRYGPDFELDIDYFPHESHNKTLDSIQKHHRRILEQLRNYADLNKLIYDYDQVYQLYNLTGMGSLVPR\n\n### Answer:", "output": "<SEQ_2D>CCEEEECCHHHHHHHTCCTTTTTHHHHHHHHHHHTTCGGGSEEECCCCCSSHHHHHHHHTTTSCHHHHHHHHHHHHHTTSSSCCCHHHHHHHHTTTCSTTSCCCTTHHHHHHHHHHHHHHHHHHHHTTSCSEEEETTCCCTTCBTTBCBTTBSSCHHHHHHHHHHTCCCCCCCCCCCCCEEEEECSSSCCHHHHHHTTTCTTEEEEEEEECCTTBSSCCSSCCCCCCCCCCCCCCCCGGGTTCEEEEEECTTCBHHHHHHHHHHHHHHHHHHHCCSEEEEECCSTTBTTSTTCCCBBCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHTCCCEEEECCCCSSHHHHHHHHHHHHHHHHHHHHCCCCCCCSBCCSSTTGGGGTTTCBSSCSCCCCCCCCCCTTCCHHHHHHHHHHHHHHHHHTTCCCCHHHHHHHHHTTTCGGGSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01328", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMDSNTVSSFQDILMRMSKMQLGSSSEDLNGMITQFESLKLYRDSLGEAVMRMGDLHSLQIRNGKWREQLSQKFEEIRWLIEEVRHRLKTTENSFEQITFMQALQLLLEVEQEIRTFSFQLI\n\n### Sequence 2:\nMDSNTVSSFQDILMRMSKMQLGSSSEDLNGMITQFESLKLYRDSLGEAVMRMGDLHSLQNRNGKWREQLSQKFEEIRWLIEEVRHRLKITENSFEQITFMQALQLLLEVEQEIRTFSFQLI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01329", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMKRTWQPNRRKRAKTHGFRARMRTPGGRKVLKRRRQKGRWRLTPAVRKR\n\n### Answer:", "output": "<SEQ_2D>CCCSCCCCHHHHHHHHSHHHHTTSHHHHHHHHHHHHHTCSCSSCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01330", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSLVTCPICHAPYVEEDLLIQCRHCERWMHAGCESLFTEDDVEQAADEGFDCVSCQPYVVK\n\n### Answer:", "output": "<SEQ_3Di>DEQAAPLVGRPDDPPFDWAAAPVPRDIDGCVSVPQDDPVSRVVCVVCHHHDPVCVVPPDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01331", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMIYKVLYQKDKIQNPRRETTQTLYLEAPSAVEARALVEKNTPYNIEFIQELSGNFLEYEEKSANFKLTTF\n\n### Sequence 2:\nMIYKVLYQKDKIQNPRRETTQTLYLEAPSAVEARALVEKNTPYNIEFIQELSGNFLEYEEKSANFKLTTF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01332", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "Sc1ccccc1", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring. Its descriptor profile: MW=110.18, LogP=1.98, TPSA=0.0, HBD=1, HBA=1, QED=0.48.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "Sc1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 110.18, "LogP": 1.98, "TPSA": 0.0, "QED": 0.48, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 1.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01333", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLSEGYLSGLEYWNDIHWSCASYNEQVAGEKEEETNSVATLSYSSVDETQVRSLYVSCKSSGKFISSVHSRESQHSRSQRVTVLQTNPNPVFESPNLAAVEICRDASRETYLVPSSCKSICKNYNDLQIAGGQVMAINSVTTDFPSESSFEYGPLLKSSEIPLPMEDSISTQPSDFPQKPIQRYSSYWRITSIKEKSSLQMQNPISNAVLNEYLEQKVVELYKQYIMDTVFHDSSPTQILASELIMTSVDQISLQVSREKNLETSKARDIVFSRLLQLMSTEITEISTPSLHISQYSNVNP\n\n### Sequence 2:\nMLSEGYLSGLEYWNDIHWSCASYNEQVAGEKEEETNSVATLSYSSVDETQVRSLYVSCKSSGKFISSVHSRESQHSRSQRVTVLQTNPNPVFESPNLAAVEICRDASRETYLVPSSCKSICKNYNDLQIAGGQVMAINSVTTDFPSESSFEYGPLLKSSEIPLPMEDSISTQPSDFPQKPIQRYSSYWRITSIKEKSSLQMQNPISNAVLNEYLEQKVVELYKQYIMDTVFHDSSPTQILASELIMTSVDQISLQVSREKNLETSKARDIVFSRLLQLMSTEITEISTPSLHISQYSNVNP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01334", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nFTSH_PELPD\n\n### Answer:", "output": "In the C-terminal section; belongs to the peptidase M41 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01335", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nMOBP PLP1 MOG MBP\n</cell>", "output": "This cell is most consistent with oligodendrocyte, an oligodendrocyte. Key markers such as PLP1, MOG, MBP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "oligodendrocyte", "tissue": "pancreas", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MOBP", "PLP1", "MOG", "MBP"]}}
{"id": "eval_01336", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSREGFQIPTNLDAAAAGTSQARTATLKYICAECSSKLSLSRTDAVRCKDCGHRILLKARTKRLVQFEAR\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCSCCBTTTCCCCCCCTTTCSCCSSSCCCEEECCCCSSCCEECCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01337", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLRDNIQGITKPAIRRLARRGGVKRISGLIYEETRGVLKVFLENVIRDAVTYTEHAKRKTVTAMDVVYALKRQGRTLYGF\n\n### Answer:", "output": "<SEQ_3Di>DPPCLCVCDPVNVVVVVVVVVDDDDDPCVSVVVSVVVVVVCVQLVVQQVVVQVVVVHDDRDVVSSQVSCVVVVRHDPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01338", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "CCCCCCCCCCCCCCCc1nnc(SCC(=O)O)[nH]1", "output": "The molecule contains the following functional groups: hydroxyl.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "CCCCCCCCCCCCCCCc1nnc(SCC(=O)O)[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 369.58, "LogP": 5.62, "TPSA": 78.87, "QED": 0.28, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["hydroxyl"]}}
{"id": "eval_01339", "category": "Mutation", "instruction": "Describe the effect of mutation Q640N1-A1003T on protein Adipocyte enhancer-binding protein 1.", "input": "### Wild-type:\nMAPVRTASLLCGLLALLTLCPEGNPQTVLTDDEIEEFLEGFLSELETQSPPREDDVEVQPLPEPTQRPRKSKAGGKQRADVEVPPEKNKDKEKKGKKDKGPKATKPLEGSTRPTKKPKEKPPKATKKPKEKPPKATKKPKEKPPKATKKPKEKPPKATKRPSAGKKFSTVAPLETLDRLLPSPSNPSAQELPQKRDTPFPNAWQGQGEETQVEAKQPRPEPEEETEMPTLDYNDQIEKEDYEDFEYIRRQKQPRPTPSRRRLWPERPEEKTEEPEERKEVEPPLKPLLPPDYGDSYVIPN\n\n### Mutant:\nMAPVRTASLLCGLLALLTLCPEGNPQTVLTDDEIEEFLEGFLSELETQSPPREDDVEVQPLPEPTQRPRKSKAGGKQRADVEVPPEKNKDKEKKGKKDKGPKATKPLEGSTRPTKKPKEKPPKATKKPKEKPPKATKKPKEKPPKATKKPKEKPPKATKRPSAGKKFSTVAPLETLDRLLPSPSNPSAQELPQKRDTPFPNAWQGQGEETQVEAKQPRPEPEEETEMPTLDYNDQIEKEDYEDFEYIRRQKQPRPTPSRRRLWPERPEEKTEEPEERKEVEPPLKPLLPPDYGDSYVIPN\n\n### Answer:", "output": "Enhances DNA-binding and phosphorylation by MAPK1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01340", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNKQQNNLQDIFLNSARKNKMPVTIYLSTGFQINGTVKGFDSFTVILDSEGKQMLIYKHAITTVTPEKPILFVDNES\n\n### Sequence 2:\nVNKQQNNLQDIFLNNARKNKIQIIIHLVNGFQLKGTVKGFDNFTVILDCDNKQMLIYKHAISTITPTKPILFADNE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01341", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSHFFAHLSRLKLINRWPLMRNVRTENVSEHSLQVAFVAHALAIIKNRKFNGNLNAERIALLAMYHDASEVITGDLPTPIKYHNPKIAHEYKKIEKVAQQKLIEMLPKELQHDFRCLLDEHYYSEEEKALVKQADALCAYLKCLEELSAGNNEFIQAKARLEKTLAIRQSPEMDYFMAVFVPSFSLSLDEISLDSLD\n\n### Sequence 2:\nMSHFFAHLSRLKLINRWPLMRNVRTENVSEHSLQVAFVAHALAIIKNRKFNGNLNAERIALLAMYHDASEVITGDLPTPIKYHNPKIAHEYKKIEKVAQQKLIEMLPKELQHDFRCLLDEHYYSEEEKALVKQADALCAYLKCLEELSAGNNEFIQAKARLENTLAIRQSPEMDYFMAVFVPSFSLSLDEISLDSLD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01342", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMKRTYQPNNRKRAKTHGFRARMKTKSGRNILARRRAKGRHQLTVSDE\n\n### Answer:", "output": "<SEQ_2D>CCCSCCCCHHHHHHHHSHHHHHHSHHHHHHHHHHHTTTCSCCCTTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01343", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMPSNHYGPIPGIPVGTMWRFRVQVSESGVHRPHVAGIHGRSNDGAYSLVLAGGYEDDVDHGNFFTYTGSGGRDLSGNKRTAEQSCDQKLTNTNRALALNCFAPINDQEGAEAKDWRSGKPVRVVRNVKGGKNSKYAPAEGNRYDGIYKVVKYWPEKGKSGFLVWRYLLRRDDDEPGPWTKEGKDRIKKLGLTMQYPEGYLEALANAHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCCCSCCTTCCTTCEESSHHHHHHHTSSCCSSCSEEEETTTEEEEEEESSCSTTCEECSSEEEEECSCSBCCTTTCSCCCBCSCCCSCHHHHHHHHHSSSCCCTTTCEECTTGGGSCEEEEEEEGGGGGGCSSSCSSSEEEEEEEEEEEEEEEECTTSSEEEEEEEEECCSSCCTTSHHHHHHHHHHTCCCBCCTTHHHHHTTCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01344", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMEEPQSDPSVEPPLSQETFSDLWKLLPENNVLSPLPSQA\n\n### Answer:", "output": "<SEQ_3Di>DVPVPDPVPVPPCPVCVPVCCVPPVPPDDDDPPDDDPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01345", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMVTFVSEMRIPVVIPYKPIQPKTRLSCILTDEEREDFAFMMLRDVVNAVKNAGCSPLILATAPVELDDVPVRILEEGLNETINGFCEGNDEPLAIVMADLALADRSAILTLLTSGGDLAIAPGRGGGTNAIYVRSAKMFQAQYYGMSYEKHVRYGTDAGLMVKIIDSFRLYCDIDEQDDLIEVFIHNTGYSREWLISHGFEIAMKKSRIGVKRPGYE\n\n### Sequence 2:\nMTKQNAFTREDLLRCSRGELFGPGNAQLPAPNMLMVDRITHISEEGGKFGKGELVAELDINPDLWFFACHFEGDPVMPGCLGLDAMWQLVGFFLGWQGLPGRGRALGSGEVKFFGQVLPSAKKVTYNIHIKRVLKGKLNMAIADGSVSVDGREIYTAEGLRVGVFTSTDNF\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01346", "category": "Mutation", "instruction": "Describe the effect of mutation P03562-C43A on protein Transcriptional activator protein.", "input": "### Wild-type:\nMRNSSSSTPPSIKAQHRAAKRRAIRRRRIDLNCGCSIYIHIDCRNNGFTHRGTYHCASSREWRLYLGDNKSPLFQDNQRRGSPLHQHQDIPLTNQVQPQPEESIGSPQGISQLPSMDDIDDSFWENLFK\n\n### Mutant:\nMRNSSSSTPPSIKAQHRAAKRRAIRRRRIDLNCGCSIYIHIDARNNGFTHRGTYHCASSREWRLYLGDNKSPLFQDNQRRGSPLHQHQDIPLTNQVQPQPEESIGSPQGISQLPSMDDIDDSFWENLFK\n\n### Answer:", "output": "72% loss of transactivation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01347", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nACT2_PLAF7\n\n### Answer:", "output": "Actin-2, Actin II.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01348", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMALTPTNLNNKMSLQMKMDCQEQQLTKKNNGFFQKLNVTEGAMQDLLKEIIKVDHILDRSDDEDDISSENPQTDFLHKGMLELEAEHDQDLSKQDKQETDVDEDPQASTSLQFSKKNLLELCLKGMFLKLNYWNTKIGLQVKELGADYIDGTEKIDNIIKKINVTENTVKSLLKDMLTLKGQIEKLEDRGLDLDQGTSTEVNTCNEVYELKKKVIERLEDLCKNVELLSAKLRMYQMEAEDTDSHSSEEIDTEEMEALLPQAPASFLVQKSPPRNTAWKRALRIFIMFDVLTVTGLLCYILFFGATFLFERVLLRMLGCRTTWDLREMREPFLNLEVEALLPS\n\n### Sequence 2:\nMSRVGIVLLNLGGPERIQDVGPFLYNLFADPEIIRLPIPALQKPLAWLISTLRSNKSQEAYRSIGGGSPLRRITDQQARELQSLLRQRNVDATTYVAMRYWHPFTESAVADMKADGIEQVVVLPLYPHFSISTSGSSFRELQRLRQGDERFAQLPLRAIRSWHDHPGYLKAMAQLMEREIDACVDPSTAHVFFSAHGVPKSYVEEAGDPYQKEIESCAELIMKTLGRDNPWTLAYQSRVGPVEWLQPYTEEALEELGEKGVKELVVVPISFVSEHIETLEEIDIEYREIATEAGVSNFRRVPALDIDPTFIASLADLVETSLAGPEVDLDEAAALPARTKLYPQEKWSWGWNNSSEVWNGRLAMLGFSAFLVELISGHGPLHALGLL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01349", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nLEUD_STAHJ\n\n### Answer:", "output": "3-isopropylmalate dehydratase small subunit, Alpha-IPM isomerase, IPMI, Isopropylmalate isomerase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01350", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSCIN_STAA1\n\n### Answer:", "output": "Belongs to the SCIN family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01351", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD4, IL2RA, CD3D, CTLA4, IKZF2", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as CD3D, CD4, IKZF2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "bone marrow", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD4", "IL2RA", "CD3D", "CTLA4", "IKZF2"]}}
{"id": "eval_01352", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "TNNI3, MYH6, MYH7, TNNT2", "output": "This profile is most consistent with cardiomyocyte, which is commonly found in prostate under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "prostate", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TNNI3", "MYH6", "MYH7", "TNNT2"]}}
{"id": "eval_01353", "category": "Mutation", "instruction": "Describe the effect of mutation Q8I4F7-D247A on protein Histone-lysine N-methyltransferase set-18.", "input": "### Wild-type:\nMAVKKNKNVPKQEDPLVQKDEISSIHARIKIFETPFATQVLNPKVSEFCANCLRGPAPGEKLLRCGGCNFSMYCSKECQATAWLVHKPECKRLKASFPNLPLTEVLFLSKVIDRIQFLEKNGDKLGIEAERKFSSLVDHKVDIRDDEEKMAHFEKIFEKMGAFRGEEMIEKGEFFDVFCKATINSHSIHTNAGNEVGMALDLGVSKYNHSCRPTCSMVFDGYRVCLRPLVPGVDAENTEEAFISYIDVGRSKYIRRRDLNSRWYFNCECTRCMDPEDDALTAIRCANPACDAPILTSETE\n\n### Mutant:\nMAVKKNKNVPKQEDPLVQKDEISSIHARIKIFETPFATQVLNPKVSEFCANCLRGPAPGEKLLRCGGCNFSMYCSKECQATAWLVHKPECKRLKASFPNLPLTEVLFLSKVIDRIQFLEKNGDKLGIEAERKFSSLVDHKVDIRDDEEKMAHFEKIFEKMGAFRGEEMIEKGEFFDVFCKATINSHSIHTNAGNEVGMALDLGVSKYNHSCRPTCSMVFDGYRVCLRPLVPGVDAENTEEAFISYIAVGRSKYIRRRDLNSRWYFNCECTRCMDPEDDALTAIRCANPACDAPILTSETE\n\n### Answer:", "output": "Decreased dimethylation of 'Lys-36' of histone H3.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01354", "category": "Mutation", "instruction": "Describe the effect of mutation Q99708-S327A on protein DNA endonuclease RBBP8.", "input": "### Wild-type:\nMNISGSSCGSPNSADTSSDFKDLWTKLKECHDREVQGLQVKVTKLKQERILDAQRLEEFFTKNQQLREQQKVLHETIKVLEDRLRAGLCDRCAVTEEHMRKKQQEFENIRQQNLKLITELMNERNTLQEENKKLSEQLQQKIENDQQHQAAELECEEDVIPDSPITAFSFSGVNRLRRKENPHVRYIEQTHTKLEHSVCANEMRKVSKSSTHPQHNPNENEILVADTYDQSQSPMAKAHGTSSYTPDKSSFNLATVVAETLGLGVQEESETQGPMSPLGDELYHCLEGNHKKQPFEESTR\n\n### Mutant:\nMNISGSSCGSPNSADTSSDFKDLWTKLKECHDREVQGLQVKVTKLKQERILDAQRLEEFFTKNQQLREQQKVLHETIKVLEDRLRAGLCDRCAVTEEHMRKKQQEFENIRQQNLKLITELMNERNTLQEENKKLSEQLQQKIENDQQHQAAELECEEDVIPDSPITAFSFSGVNRLRRKENPHVRYIEQTHTKLEHSVCANEMRKVSKSSTHPQHNPNENEILVADTYDQSQSPMAKAHGTSSYTPDKSSFNLATVVAETLGLGVQEESETQGPMSPLGDELYHCLEGNHKKQPFEESTR\n\n### Answer:", "output": "Abolishes BRCA1 interaction and ubiquitination. No activation of CHEK1 after DNA damage. The CtIP/BRCA1 complex is required for DNA damage-induced Chk1 phosphorylation and the G(2)/M transition checkpoint.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01355", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "TCF4, CLEC4C, LILRA4", "output": "This profile is most consistent with plasmacytoid dendritic cell, which is commonly found in brain under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "brain", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TCF4", "CLEC4C", "LILRA4"]}}
{"id": "eval_01356", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSCDGGKPAPAKLGDEQLAELREIFRSFDQNKDGSLTELELGSLLRSLGLKPSQDQLDTLIQKADRNNNGLVEFSEFVALVEPDLVKCPYTDDQLKAIFRMFDRDGNGYITAAELAHSMAKLGHALTAEELTGMIKEADRDGDGCIDFQEFVQAITSAAFDNAWG\n\n### Sequence 2:\nKLDEEQINELREIFRSFDRNKDGSLTQLELGSLLRALGVKPSPDQFETLIDKADTKSNGLVEFPEFVALVSPELLSPAKRTTPYTEEQLLRLFRIFDTDGNGFITAAELAHSMAKLGHALTVAELTGMIKEADSDGDGRINFQEFAKAINSAAFDDIWG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01357", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nTAGLN ACTA2 MYH11 CNN1\n</cell>", "output": "This cell is most consistent with smooth muscle cell, a smooth muscle cell. Key markers such as TAGLN, CNN1, MYH11 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "smooth muscle cell", "tissue": "intestine", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TAGLN", "ACTA2", "MYH11", "CNN1"]}}
{"id": "eval_01358", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nFCER1A HLA-DQA1 CD1C CLEC10A\n</cell>", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as CLEC10A, CD1C, FCER1A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "spleen", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["FCER1A", "HLA-DQA1", "CD1C", "CLEC10A"]}}
{"id": "eval_01359", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAKGTDSLKNSIEKYLKDKKAKVGVAVLGIEDNFKLNVNEKHHYPMQSTYKFHLALAVLDKLDKENISVDKKLFVKKSDLQPNTWSPLKDKYPNGNLELSFSEIIKSTVSHSDNNGCDILFRFVGGTNKVHNFISKLGVKNISIKATEEEMHKAWNVQYTNWTTPDATVQLLKKFYKNEILSKNSYDFLLNTMIETTTGPKRLKGLLPDGTVVAHKTGSSDTNNKGITAATNDIGIITLPNGKHFAIAVYVSDSSEKSDVNEKIIAEICKSVWDYLVKDGK\n\n### Answer:", "output": "<SEQ_2D>CCCHHHHHHHHHHHHTTCSSEEEEEEEESSSCCEEEESTTSCEECGGGHHHHHHHHHHHHHHHTTCCTTCEEEECGGGSCSSSCCHHHHHCTTCCEEEEHHHHHHHHHHSCCHHHHHHHHHHHTCHHHHHHHHHHTTCCSEECCCCHHHHHTCGGGGGGSEECHHHHHHHHHHHHTTSSSCHHHHHHHHHHHHSCCSCTTSTTTTSCTTCCEEEEEEECCBCTTSCEEEEEEEEEEECTTSCEEEEEEEEEEECSCHHHHHHHHHHHHHHHHHHHHCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01360", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "C=C1CCC=C(C(=O)O)CCC2C1CC2(C)C", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains hydroxyl. Its descriptor profile: MW=234.34, LogP=3.79, TPSA=37.3, HBD=1, HBA=1, QED=0.7.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "C=C1CCC=C(C(=O)O)CCC2C1CC2(C)C", "compound_name": null, "pubchem_cid": null, "MW": 234.34, "LogP": 3.79, "TPSA": 37.3, "QED": 0.7, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01361", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMRSLISVAVLSALPTAFSQANTSYTDYNVEANPDLFPLCLQHLNASFPDCASGPLSLTPVCDRSLSPKDRATALVSLFTFDELVNNTGNTGLGVSRLGLPNYQVWGEALHGVGRANFVESGNFSWATSFPMPITMMAALNKTLIHQIGTIVSTQLRAFSNAGLGGVDVYSPNINTFRHPVWGRGQETPGEDAFLTSVYGYEYITALQGGVDPETLKIIATAKHYAGYDIESWNNHSRLGNDMQITQQELSEYYTPPFIVASRDAKVRSVMCSYNAVNGVPSCANKFFLQTLLRDTFEFSEDGYVSGDCGAVYNVWNPHGYASNEAAASADSILAGTDIDCGTSYQWHSEDAFEDSLVSRSDIERGVIRLYSNLVQAGYFDGEDAPYRDITWDDVLSTDAWNIAYEAAVEGIVLLKNDETLPLSKDIKSVAVIGPWANVTEELQGNYFGPAPYLISPLTGFRDSGLDVHYALGTNLTSHSTSGFEEALTAAKQADAIIFAGGIDNTIEAEAMDRENITWPGNQLDLISKLSELGKPLVVLQMGGGQVDSSSLKDNDNVNALIWGGYPGQSGGHALADIITGKRAPAGRLVTTQYPAEYAEVFPAIDMNLRPNETSGNPGQTYMWYTGTPVYEFGHGLFYTTFEESTETTDAGSFNIQTVLTTPHSGYEHAQQKTLLNFTATVKNTGERESDYTALVYVNTTAGPAPYPKKWVVGFDRLGGLEPGDSQTLTVPVTVESVARTDEQGNRVLYPGSYELALNNERSVVVKFELKGEEAVILSWPEDTTSDFVSSIDGGLDRKQDVIA\n\n### Sequence 2:\nIAAVLVALLPGALAQANTSYVDYNVEANPNLTPQSVATIDLSFPDCENGPLSKTLVCDTSARPHDRAAALVSMFTFEELVNNTGNTSPGVPRLGLPPYQVWSEALHGLDRANFTDEGEYSWATSFPMPILTMSALNRTLINQIATIIATQGRAFNNVGRYGLDVYAPNINAFRSAMWGRGQETPGEDAYCLASAYAYEYITGIQGGVDPEHLKLVATAKHYAGYDLENWDGHSRLGNDMNITQQELSEYYTPQFLVAARDAKVHSVMCSYNAVNGVPSCANSFFLQTLLRDTFGFVEDGYVSSDCDSAYNVWNPHEFAANITGAAADSIRAGTDIDCGTTYQYYFGEAFDEQEVTRAEIERGVIRLYSNLVRLGYFDGNGSVYRDLTWNDVVTTDAWNISYEAAVEGIVLLKNDGTLPLAKSVRSVALIGPWMNVTTQLQGNYFGPAPYLISPLNAFQNSDFDVNYAFGTNISSHSTDGFSEALSAAKKSDVIIFAGGIDNTLEAEAMDRMNITWPGNQLQLIDQLSQLGKPLIVLQMGGGQVDSSSLKSNKNVNSLIWGGYPGQSGGQALLDIITGKRAPAGRLVVTQYPAEYATQFPATDMSLRPH--GNNPGQTYMWYTGTPVYEFGHGLFYTTFHASLPGTGKDKTSFNIQDLLTQPHPGFANVEQMPLLNFTVTITNTGKVASDYTAMLFANTTAGPAPYPNKWLVGFDRLASLEPHRSQTMTIPVTIDSVARTDEAGNRVLYPGKYELALNNERSVVLQFVLTGREAVIFKWP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01362", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPYRB_BACAA\n\n### Answer:", "output": "Belongs to the aspartate/ornithine carbamoyltransferase superfamily. ATCase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01363", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "CD8A, HAVCR2, PDCD1", "output": "The expression pattern is consistent with a healthy exhausted CD8 T cell without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "pancreas", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["CD8A", "HAVCR2", "PDCD1"]}}
{"id": "eval_01364", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "CLDN5, KDR, VWF, PECAM1, CDH5", "output": "The marker profile is enriched for vascular development and angiogenesis, consistent with endothelial cell function.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "endothelial cell", "tissue": "colon", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["CLDN5", "KDR", "VWF", "PECAM1", "CDH5"]}}
{"id": "eval_01365", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nMS4A1:low IGHG1:low CD79A:low IGHA1:low CD27:medium\n</cell>", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as IGHA1, CD27, IGHG1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "tumor", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MS4A1", "IGHG1", "CD79A", "IGHA1", "CD27"]}}
{"id": "eval_01366", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIGLVGKKVGMTRIFTEDGVSIPVTVIEVEANRVTQVKDLANDGYRAVQVTTGAKKANRVTKPEAGHFAKAGVEAGRGLWEFRLAEGEEYTVGQSISVELFADVKKVDVTGTSKGKGFAGTVKRWNFRTQDATHGNSLSHRVPGSIGQNQTPGKVFKGKKMAGQMGNERVTVQSLDVVRVDAERNLLLVKGGVPGATGCDLIVKPAVKA\n\n### Sequence 2:\nMIGLVGKKVGMTRIFTEEGVSIPVTVIEVEANRVTQVKDLANDGYRAVQVTTGAKKANRVTKPEAGHFAKAGVEAGRGLWEFRLADGEEYTVGQSISVELFAEVKKVDVTGTSKGKGFAGTVKRWNFRTQDATHGNSLSHRVPGSIGQNQTPGKVFKGKKMAGQLGNERVTVQSLDVVRVDAERNLLLVKGAVPGATGSDLIVKPAVKA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01367", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGLPVYITPGSGQFMTTDDMQSPCALPWYHPTKEISIPGEVKNLIEMCQVDTLIPVNNVGNNVGNVSMYTVQLGNQTGMAQKVFSIKVDITSTPLATTLIGEIASYYTHWTGSLRFSFMFCGTANTTLKLLLAYTPPGIDEPTTRKDAMLGTHVVWDVGLQSTISLVVPWVSASHFRLTADNKYSMAGYITCWYQTNLVVPPSTPQTADMLCFVSACKDFCLRMARDTDLHIQSGPIEQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCTTTTCCCTTCCCCCCBSSTTCCCCCCCCCSCCBCCHHHHHSSCEECCCCCBGGGTTSSGGGEEEECCCSSSSCEEEEEECCSSSTTSTTSTTHHHHTTEEEEESCEEEEEEECSCSSCEEEEEEEEECTTSCCCSSHHHHHTSEEEEEECSSSCEEEEEECCCCSSSSEESSCCSTTCCCEEEEEEEEEEECCSSSCSCEEEEEEEEECTTCEEEEECCCSSCCCCSCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01368", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nDIMLTQSPASLTVSLGQRATMSCRASQSVSTSRYSYMHWYQEKAGQPPQLLIKYASNLESGVPARFSGSGSGTDFTLNIHPVEAEDTATYYCQHSWEIPYTFGGGTKLEIKRADAAPTVSIFPPSSEQLTSGGASVVCFLNNFYPKDINVKWKIDGSERQNGVLNSWTDQDSKDSTYSMSSTLTLTKDEYERHNSYTCEATHKTSTSPIVKSFNRNEC\n\n### Answer:", "output": "<SEQ_2D>CCCEEEECSEEEEETTSCEEEEEEESSCCEETTEECEEEEEECTTCCCEEEEETTTEECTTCCTTEEEEEETTEEEEEESSCCGGGCEEEEEEECSSSSCCBCCCEEEEEECCCBCCEEEEECCCHHHHHTTEEEEEEEEEEEBSSCCEEEEEETTEECCTTEEEEECCCCSSSCCEEEEEEEEEEHHHHHTCCEEEEEEECSSCSCCEEEEEESTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01369", "category": "Mutation", "instruction": "Describe the effect of mutation P10696-R316L on protein Alkaline phosphatase, germ cell type.", "input": "### Wild-type:\nMQGPWVLLLLGLRLQLSLGIIPVEEENPDFWNRQAAEALGAAKKLQPAQTAAKNLIIFLGDGMGVSTVTAARILKGQKKDKLGPETFLAMDRFPYVALSKTYSVDKHVPDSGATATAYLCGVKGNFQTIGLSAAARFNQCNTTRGNEVISVMNRAKKAGKSVGVVTTTRVQHASPAGAYAHTVNRNWYSDADVPASARQEGCQDIATQLISNMDIDVILGGGRKYMFPMGTPDPEYPDDYSQGGTRLDGKNLVQEWLAKHQGARYVWNRTELLQASLDPSVTHLMGLFEPGDMKYEIHRD\n\n### Mutant:\nMQGPWVLLLLGLRLQLSLGIIPVEEENPDFWNRQAAEALGAAKKLQPAQTAAKNLIIFLGDGMGVSTVTAARILKGQKKDKLGPETFLAMDRFPYVALSKTYSVDKHVPDSGATATAYLCGVKGNFQTIGLSAAARFNQCNTTRGNEVISVMNRAKKAGKSVGVVTTTRVQHASPAGAYAHTVNRNWYSDADVPASARQEGCQDIATQLISNMDIDVILGGGRKYMFPMGTPDPEYPDDYSQGGTRLDGKNLVQEWLAKHQGARYVWNRTELLQASLDPSVTHLMGLFEPGDMKYEIHRD\n\n### Answer:", "output": "The BeWo AP (Nagao isozyme) remains unaltered with PI-glycan that is added soon after translation, not cotranslationally. The LS174T cell alkaline phosphatase cDNAs are actually much more dissimilar to the 'germ cell' alkaline phosphatase gene than they are to PLAP. Only 7 amino acid similarities exist between the LS174T cell enzyme and the alkaline phosphatase encoded by the germ cell alkaline phosphatase genomic DNA clone isolated by Millan and Manes (Pc. Natl. Acad. Sci. USA, 85: 3024-3028, 1988).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01370", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "SELL, CCR7, LEF1, CD8A, CD8B, CD3D, CD3E", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as SELL, CD8A, CD8B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "breast", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SELL", "CCR7", "LEF1", "CD8A", "CD8B", "CD3D", "CD3E"]}}
{"id": "eval_01371", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nCD45RO S100A4 IL7R CD3D CD4\n</cell>", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as S100A4, CD3D, IL7R support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "breast", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD45RO", "S100A4", "IL7R", "CD3D", "CD4"]}}
{"id": "eval_01372", "category": "Mutation", "instruction": "Describe the effect of mutation Q99741-T323R on protein Cell division control protein 6 homolog.", "input": "### Wild-type:\nMPQTRSQAQATISFPKRKLSRALNKAKNSSDAKLEPTNVQTVTCSPRVKALPLSPRKRLGDDNLCNTPHLPPCSPPKQGKKENGPPHSHTLKGRRLVFDNQLTIKSPSKRELAKVHQNKILSSVRKSQEITTNSEQRCPLKKESACVRLFKQEGTCYQQAKLVLNTAVPDRLPAREREMDVIRNFLREHICGKKAGSLYLSGAPGTGKTACLSRILQDLKKELKGFKTIMLNCMSLRTAQAVFPAIAQEICQEEVSRPAGKDMMRKLEKHMTAEKGPMIVLVLDEMDQLDSKGQDVLYTL\n\n### Mutant:\nMPQTRSQAQATISFPKRKLSRALNKAKNSSDAKLEPTNVQTVTCSPRVKALPLSPRKRLGDDNLCNTPHLPPCSPPKQGKKENGPPHSHTLKGRRLVFDNQLTIKSPSKRELAKVHQNKILSSVRKSQEITTNSEQRCPLKKESACVRLFKQEGTCYQQAKLVLNTAVPDRLPAREREMDVIRNFLREHICGKKAGSLYLSGAPGTGKTACLSRILQDLKKELKGFKTIMLNCMSLRTAQAVFPAIAQEICQEEVSRPAGKDMMRKLEKHMTAEKGPMIVLVLDEMDQLDSKGQDVLYTL\n\n### Answer:", "output": "In MGORS5. The mutation in the ORC4 gene is associated with Meier-Gorlin syndrome. This mutation affects the pre-replication complex formation and replication licensing, leading to developmental abnormalities including absent or hypoplastic patellae and small ears.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01373", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPAND_BORPA\n\n### Answer:", "output": "Catalyzes the pyruvoyl-dependent decarboxylation of aspartate to produce beta-alanine.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01374", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMVAIHKPAHDVILKPVVSEKSYAASDRGQYTFVVAPDANKVQIKQAIEEIFKVKVTNVNTLNRAGKKQRTRTGFGQRASQKRAIVTVAEGQTIDIFGN\n\n### Sequence 2:\nMQVLVRDNNVDQALKALKKKMQREGIFREMKLRGHYEKPSEKKAREKAEAVRRARKLARKKLQREGLLPSKPKPAFGADRRPSAAAR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01375", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nCD8A:low CD3D:high NKG7:medium GZMA:high PRF1:high CD8B:high\n</cell>", "output": "This cell is most consistent with cytotoxic T cell, an effector cytotoxic CD8+ T cell. Key markers such as CD8A, PRF1, CD8B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "lymph node", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD8A", "CD3D", "NKG7", "GZMA", "PRF1", "CD8B"]}}
{"id": "eval_01376", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSNKVQRFIEAERELSQLKHWLKTTHKISIEEFVVLFKVYEAEKISGKELRDTLHFEMLWDTSKIDVIIRKIYKKELISKLRSETDERQVFYFYSTSQKKLLDKITKEIEVLSVTN\n\n### Sequence 2:\nMTSEQLLAEIREANLTYLMLAQTLIRQDKAEAVFRLGLNEEAADILASLSAAQVLKLASRNTLLCSFRVDDELVWSLLTSHNTPRKAASEATNTLHANILMASRVSEVL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01377", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL15_FRATT\n\n### Answer:", "output": "Binds to the 23S rRNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01378", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMLAKRIIPCLDVRDGQVVKGVQFRNHEIIGDIVPLAKRYAEEGADELVFYDITASSDGRVVDKSWVSRVAEVIDIPFCVAGGIKSLEDAAKILSFGADKISINSPALADPTLITRLADRFGVQCIVVGIDTWYDAETGKYHVNQYTGDESRTRVTQWETLDWVQEVQKRGAGEIVLNMMNQDGVRNGYDLEQLKKVREVCHVPLIASGGAGTMEHFLEAFRDADVDGALAASVFHKQIINIGELKAYLATQGVEIRIC\n\n### Sequence 2:\nMDFTDQALVLRVGRFREADLWVRFLCRQRGIISAFAFGGCRSRRRFCGCLDIFNVVLMRVQGTRGGLYQSLQEATLLKGPDRLRRDWRRYGVAVNCLRFIEALGAGPDGADSAFTLTLEMLELLETVDVVPPLLPLLFRARFAFEQGYAPRFESCASCGEPFDGTAKDGGARFHVRDGVLYCGRCSAPTGATVAISRETLDALRFVQDNSPLRWSELCFSPTGRRECSRAVDGFIQYHIGLTWENGTFRRL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01379", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFDM_PSEPU\n\n### Answer:", "output": "Formaldehyde dismutase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01380", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD3E TCF7 CCR7 LTB SELL IL7R\n</cell>", "output": "This cell is most consistent with naive CD4+ T cell, a T-cell subtype with naive/central-memory-like profile. Key markers such as IL7R, CD3E, CCR7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "naive CD4+ T cell", "tissue": "lymph node", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD3E", "TCF7", "CCR7", "LTB", "SELL", "IL7R"]}}
{"id": "eval_01381", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCALR3_BOVIN\n\n### Answer:", "output": "Belongs to the calreticulin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01382", "category": "Mutation", "instruction": "Describe the effect of mutation Q93099-H225R on protein Homogentisate 1,2-dioxygenase.", "input": "### Wild-type:\nMAELKYISGFGNECSSEDPRCPGSLPEGQNNPQVCPYNLYAEQLSGSAFTCPRSTNKRSWLYRILPSVSHKPFESIDEGQVTHNWDEVDPDPNQLRWKPFEIPKASQKKVDFVSGLHTLCGAGDIKSNNGLAIHIFLCNTSMENRCFYNSDGDFLIVPQKGNLLIYTEFGKMLVQPNEICVIQRGMRFSIDVFEETRGYILEVYGVHFELPDLGPIGANGLANPHDFLIPIAWYEDRQVPGGYTVINKYQGKLFAAKQDVSPFNVVAWHGNYTPYKYNLKNFMVINSVAFDHADPSIFTV\n\n### Mutant:\nMAELKYISGFGNECSSEDPRCPGSLPEGQNNPQVCPYNLYAEQLSGSAFTCPRSTNKRSWLYRILPSVSHKPFESIDEGQVTHNWDEVDPDPNQLRWKPFEIPKASQKKVDFVSGLHTLCGAGDIKSNNGLAIHIFLCNTSMENRCFYNSDGDFLIVPQKGNLLIYTEFGKMLVQPNEICVIQRGMRFSIDVFEETRGYILEVYGVHFELPDLGPIGANGLANPRDFLIPIAWYEDRQVPGGYTVINKYQGKLFAAKQDVSPFNVVAWHGNYTPYKYNLKNFMVINSVAFDHADPSIFTV\n\n### Answer:", "output": "Increases in AKU. The mutation in the HGD gene is predicted to affect the activity of the homogentisate 1, 2-dioxygenase protein complex through stabilization of the individual protomer structure or through facilitation of protomer-protomer interactions.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01383", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nISCA1_MACFA\n\n### Answer:", "output": "Iron-sulfur cluster assembly 1 homolog, mitochondrial, HESB-like domain-containing protein 2, Iron-sulfur assembly protein IscA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01384", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMQGQSQQQAYDRGITIFSPDGRLYQVEYAREAVKRGTASIGIRTEDGVVLVVDKRIRSPLMERTSVEKIHKADDHIGIASAGHVADARQLIDFARRQAQVNQLRYDEPIGVETLTKAVTDHIQQYTQVGGARPFGVALIIGGIEDGEPRLYETDPSGTPYEWQALAVGADRGDIQEYLEEHYSESLDTEAGIGLALEALASVNDDSLTPEGIGLATVTVDDDDGFSQLSDEEIEAHLEERDLLEAEDDADEDDEE\n\n### Sequence 2:\nMIETKPIDGKAFAAGLRARIAEEVAVLVRDHDLKPGLAVVLVGEDPASQVYVRNKAAQTAEAGMASFEYKLPADTAEADLLALVEKLNADPAVNGILVQLPLPAHLDSMKVLAAIDPAKDVDGFHVVNAGRLAVGLDALVPCTPLGCVMLLKHHLGNLSGLNAVVVGRSNIVGKPAAQLLLREDCTVTIAHSRTRDLPGMCRQADILVAAVGRPEMVRGDWIKPGATVIDVGINRVPKADGKTRLVGDVAYEEALGVAGLITPVPGGVGPMTIACLLQNTLTAARRQKGL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01385", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nHVLQLESASDKAHYILSKDGNRNNWYIGRGSDNNNDCTFHSYVHGTTLTLKQDYAVVNKHFHVGQAVVATDGNIQGTKWGGKWLDAYLRDSFVAKSKAWTQVWSGSAGGGVSVTVSQDLRFRNIWIKCANNSWNFFRTGPDGIYFIASDGGWLRFQIHSNGLGFKNIADSRSVPNAIMVENE\n\n### Answer:", "output": "<SEQ_3Di>DDDDDDDDPQAWDKDWDADPNRGAWIWARHGGNAHKTKTAGPVVRWIWIDHDPDIDIDAWDQDPNWIQGPVRWIADVVLVRDIVVVSVVVPDDDPQPDKDWFFFAWAAAFRKDFTNDQQAQWWKWWQWPDPDTHIDGHHWFAWDKDLDPQGIFIWTQDDRRGMIHGNHTHNTITGIMMTDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01386", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTATDNARQVTIIGAGLAGTLVARLLARNGWQVNLFERRPDPRIETGARGRSINLALAERGAHALRLAGLEREVLAEAVMMRGRMVHVPGTPPNLQPYGRDDSEVIWSINRDRLNRILLDGAEAAGASIHFNLGLDSVDFARQRLTLSNVSGERLEKRFHLLIGADGCNSAVRQAMASVVDLGEHLETQPHGYKELQITPEASAQFNLEPNALHIWPHGDYMCIALPNLDRSFTVTLFLHHQSPAAQPASPSFAQLVDGHAARRFFQRQFPDLSPMLDSLEQDFEHHPTGKLATLRLTTWHVGGQAVLLGDAAHPMVPFHGQGMNCALEDAVALAEHLQSAADNASALAAFTAQRQPDALAIQAMALENYVEMSSKVASPTYLLERELGQIMAQRQPTRFIPRYSMVTFSRLPYAQAMARGQIQEQLLKFAVANHSDLTSINLDAVEHEVTRCLPPLSHLS\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCEEEEECCSHHHHHHHHHHHHTTCEEEEECSSCCTTSCCSSCCCCCEEEECHHHHHHHHHTTCHHHHHTTCEEECEEEECCTTSCCEEEESSSSTTSCEEEEEHHHHHHHHHHHHHHTTCEEECSCEEEEEETTTTEEEEECTTCCEEEEECSSEEECCCTTCHHHHHHTTTSCCCCEEEECSEEEEEEEECHHHHHHTTCCTTSEEEEEETTEEEEEEECTTSCEEEEEEEESSCCSSCTTSCCGGGCCSHHHHHHHHHHHCTTTGGGCTTHHHHHHHSCCBCCEEEECSCCEETTTEEECGGGTCCCCSTTSCHHHHHHHHHHHHHHHHHHSSSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTGGGCHHHHHHHHHHHHHHHHSTTTCCCHHHHHHHCCSCHHHHHHHHHHHHHHHHHHHTTCSSGGGSCHHHHHHHHHHHCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01387", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CCOC", "output": "This molecule is small and relatively nonpolar, with hydrophilic character. It contains ether. Its descriptor profile: MW=60.1, LogP=0.65, TPSA=9.23, HBD=0, HBA=1, QED=0.43.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "FreeSolv", "canonical_smiles": "CCOC", "compound_name": null, "pubchem_cid": null, "MW": 60.1, "LogP": 0.65, "TPSA": 9.23, "QED": 0.43, "functional_groups": ["ether"], "moleculenet_labels": {"label": -2.1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01388", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: C=CCSCCCC#Cc1ccccc1C(=O)OC\nDescriptors: MW=274.39, LogP=3.52, TPSA=26.3, HBD=0, HBA=3", "output": "Descriptors: MW=274.39, LogP=3.52, TPSA=26.3, HBD=0, HBA=3, QED=0.34, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "C=CCSCCCC#Cc1ccccc1C(=O)OC", "compound_name": null, "pubchem_cid": null, "MW": 274.39, "LogP": 3.52, "TPSA": 26.3, "QED": 0.34, "functional_groups": ["aromatic ring", "ester", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01389", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNAPDKFASLLALTREPFPASTKSYLAGSQPGLRVPVRDIQLTNGEVVSVYDTSGPYTDPAVQIDVRKGLASVRGEWIAARGDTEGYEGRVRKALDDGQKAEDGDRLAQLRAEAAALQRQPLRARSGANVTQMHYAKKGIVTPEMEYVALRENGRREWMQQYMQDAAREQRLAGNPLGASIPKIITPEFVRDEVARGRAIIPANINHPEVEPMAIGRNFKVKINANIGNSAVTSSIEEEVEKLVWAIRWGADNVMDLSTGKNIHTTRDWIVRNSPVPIGTVPIYQALEKVGGIAEDLTWEIFRDTLIEQAEQGVDYFTIHAGVRLAYIQLTAARRTGIVSRGGSIMAKWCMAHHKESFLYTHFEDICDIMKAYDVAFSLGDGLRPGCASDANDEAQFAELHTLGELTQIAWKHDVQTMIEGPGHVPMHMIQANMTEQLKTCHEAPFYTLGPLTIDIAPGYDHIASAIGAAMIGWMGTAMLCYVTPKEHLGLPDRDDVKQGIIAYKIAAHAADVAKGHPGARARDDALSQARFDFRWQDQFNLGLDPDTAKEYHDETLPKDSAKVAHFCSMCGPKFCSMKITQEVREFAQQGLQSKAEEFNRTGGELYVPIHRAD\n\n### Sequence 2:\nMNAPDKFAQLLALTREPFPASTKNYLIGSRPDVRVPVRDIALTNGEQVSVYDTSGPYTDPAAEIDVRRGLPSVRGAWIEGRGDTERYEGRQRVALDDGSKSEDAARLAQLRAEAAALQRAPLRAKSGASHTGNVTQMHYAKKGIVTPEMEYVALRENGRREWMAQYQQDAAREQRLAGNPMGASIPKIITPEFVRDEVARGRAIIPANINHPEVEPMAIGRTFKVKINANIGNSAVTSSIEEEVEKLVWAIRWGADNVMDLSTGKNIHTTRDWIVRNSPVPIGTVPIYQALEKVGGIAEDLTWEIFRDTLVEQAEQGVDYFTIHAGVRLAFIHLTAQRRTGIVSRGGSIMAKWCMAHHRESFLYEHFEDICDIMKQYDVSFSLGDGLRPGCASDANDEAQFAELRTLGELTRTAWKHDVQTMIEGPGHVPMHMIQANMTEQLRTCHEAPFYTLGPLTIDIAPGYDHIASAIGAAMIGWMGTAMLCYVTPKEHLGLPDRDDVKQGIIAYKIAAHAADVAKGHPGARARDDALSQARFDFRWQDQFNLGLDPETAQQYHDETLPKDSAKVAHFCSMCGPKFCSMKITQEVREFAAQGMQEKAREFRGTGGELYVPIQPA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01390", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMELILSAGIGTLTASGVYLLLRPRTYQVIIGLSLLSFAVNLFIFGMGRLRVNAPPILDPGGVGDLARYTDPVPQALVLTAIVIGFAMTALFLVVLLASRGFTGTDHVDGREQRGD\n\n### Sequence 2:\nMEIIMIFVSGILTSISVYLVLSKSLIRIIMGTTLLTHAANLFLITMGGLKHGTVPIFEKGT----SSYVDPIPQALILTAIVIAFATTAFFLVLAFRTYKELGTDNVE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01391", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL22_BURO0\n\n### Answer:", "output": "The globular domain of the protein is located near the polypeptide exit tunnel on the outside of the subunit, while an extended beta-hairpin is found that lines the wall of the exit tunnel in the center of the 70S ribosome.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01392", "category": "Mutation", "instruction": "Describe the effect of mutation Q9HB75-F587H on protein p53-induced death domain-containing protein 1.", "input": "### Wild-type:\nMAATVEGPELEAAAAAGDASEDSDAGSRALPFLGGNRLSLDLYPGGCQQLLHLCVQQPLQLLQVEFLRLSTHEDPQLLEATLAQLPQSLSCLRSLVLKGGQRRDTLGACLRGALTNLPAGLSGLAHLAHLDLSFNSLETLPACVLQMRGLGALLLSHNCLSELPEALGALPALTFLTVTHNRLQTLPPALGALSTLQRLDLSQNLLDTLPPEIGGLGSLLELNLASNRLQSLPASLAGLRSLRLLVLHSNLLASVPADLARLPLLTRLDLRDNQLRDLPPELLDAPFVRLQGNPLGEASP\n\n### Mutant:\nMAATVEGPELEAAAAAGDASEDSDAGSRALPFLGGNRLSLDLYPGGCQQLLHLCVQQPLQLLQVEFLRLSTHEDPQLLEATLAQLPQSLSCLRSLVLKGGQRRDTLGACLRGALTNLPAGLSGLAHLAHLDLSFNSLETLPACVLQMRGLGALLLSHNCLSELPEALGALPALTFLTVTHNRLQTLPPALGALSTLQRLDLSQNLLDTLPPEIGGLGSLLELNLASNRLQSLPASLAGLRSLRLLVLHSNLLASVPADLARLPLLTRLDLRDNQLRDLPPELLDAPFVRLQGNPLGEASP\n\n### Answer:", "output": "Loss of the proteolytic cleavage producing PIDD-CC.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01393", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nTAPSQVLKIRRPDDWHLHLRDGDMLKTVVPYTSEIYGRAIVMPNLAPPVTTVEAAVAYRQRILDAVPAGHDFTPLMTCYLTDSLDPNELERGFNEGVFTAAKLYPANAVTNSSHGVTSVDAIMPVLERMEKIGMPLLVHGEVTHADIDIFDREARFIESVMEPLRQRLTALKVVFEHITTKDAADYVRDGNERLAATITPQHLMFNRNHMLVGGVRPHLYCLPILKRNIHQQALRELVASGFNRVFLGTDSAPHARHRKESSCGCAGCFNAPTALGSYATVFEEMNALQHFEAFCSVNGPQFYGLPVNDTFIELVREEQQVAESIALTDDTLVPFLAGETVRWSVKQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCEEEEECCCEEEECCCSHHHHHHHHHHHHTTCSEEEECCCCSSCCCSHHHHHHHHHHHHHTSCTTCCCEEEEEEECCTTCCHHHHHHHHHTTSEEEEEECCCCTTSCTTTCCSCTGGGHHHHHHHHHHTCCEEECCCCCCTTSCGGGHHHHHHHHTHHHHHHHSTTCCEEECSCCSHHHHHHHHTSCTTEEEEECHHHHHCCHHHHHTTSBCGGGCCSSCCCCHHHHHHHHHHHHTTCTTEEECCCBCCCCHHHHSSTTBCCCCCCTTTHHHHHHHHHHHTTCGGGHHHHHHTHHHHHHTCCCCCSEEEEECCCEECCSCEECSSSEECCTTTTCEESCEECC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01394", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVRESIPKEGENIKIQSYKHDGKIHRVWSETTILKGTDHVVIGGNDHTLVTESDGRTWITREPAIVYFHSEYWFNVICMFREDGIYYYCNLSSPFVCDEEALKYIDYDLDIKVYPNGKYHLLDEDEYEQHMNQMNYPHDIDIILRRNVDILQQWIEQKKGPFAPDFIKVWKERYKKIRQY\n\n### Sequence 2:\nMVRESIPKEGENIKIQSYKHDGKIHRVWSETTILKGTDHVVIGGNDHTLVTESDGRTWITREPAIVYFHSEYWFNVICMFREDGIYYYCNLSSPFVCDEEALKYIDYDLDIKVYPNGKYHLLDEDEYEQHMNQMNYPHDIDIILRRNVDILQQWIEQKKGPFAPDFIKVWKDRYKKIRQY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01395", "category": "Mutation", "instruction": "Describe the effect of mutation O35815-N134A on protein Ataxin-3.", "input": "### Wild-type:\nMESIFHEKQEGSLCAQHCLNNLLQGEYFSPVELSSIAHQLDEEERLRMAEGGVTSEDYRTFLQQPSGNMDDSGFFSIQVISNALKVWGLELILFNSPEYQRLRIDPINERSFICNYKEHWFTVRKLGKQWFNLNSLLTGPELISDTYLALFLAQLQQEGYSIFVVKGDLPDCEADQLLQMIKVQQMHRPKLIGEELAHLKEQSALKADLERVLEAADGPGMFDDDEDDLQRALAMSRQEIDMEDEEADLRRAIQLSMQGSSRGMCEDSPQTSSTDLSSEELRKRREAYFEKQQHQQQEAD\n\n### Mutant:\nMESIFHEKQEGSLCAQHCLNNLLQGEYFSPVELSSIAHQLDEEERLRMAEGGVTSEDYRTFLQQPSGNMDDSGFFSIQVISNALKVWGLELILFNSPEYQRLRIDPINERSFICNYKEHWFTVRKLGKQWFNLASLLTGPELISDTYLALFLAQLQQEGYSIFVVKGDLPDCEADQLLQMIKVQQMHRPKLIGEELAHLKEQSALKADLERVLEAADGPGMFDDDEDDLQRALAMSRQEIDMEDEEADLRRAIQLSMQGSSRGMCEDSPQTSSTDLSSEELRKRREAYFEKQQHQQQEAD\n\n### Answer:", "output": "Loss of enzyme activity. The enzyme activity of human AT3 (hAT3) is not affected by the mutation .", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01396", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMDLSSLSPAKGSVKNKKRVGRGQGSGNGTTAGKGNKGQQSRSGYKRPVSEGGQMPLYRRLPKFGFTKPNRKSVIGVNLSQIAKWIEDGRVSSEVGVEDFKQLCNASKSDYFKVLGNGELGTAVKITAHFVSKSAEEKIKQAGGEVILAERTLLEAERVRDLSTDEALLKPKAKLRKFKKQSKSS\n\n### Sequence 2:\nMDLSSLSPAKGSVKNKKRVGRGQGSGNGTTAGKGNKGQQSRSGYKRPVSEGGQMPLYRRLPKFGFTKPNRKSVIGVNLSQIAKWIEDGRVSSEVGVEDFKQLCNASKSDYFKVLGNGELGTAVKITAHFVSKSAEEKIKQAGGEVILAERTLLEAERVRDLSTDEALLKPKAKLRKFKKQSKSS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01397", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: ALK\nLigand SMILES: C[NH+](C)CC(c1ccc(O)cc1)C1(O)CCCCC1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "ClinTox", "canonical_smiles": "C[NH+](C)CC(c1ccc(O)cc1)C1(O)CCCCC1", "compound_name": null, "pubchem_cid": null, "MW": 264.39, "LogP": 1.32, "TPSA": 44.9, "QED": 0.77, "functional_groups": ["aromatic ring", "hydroxyl", "phenol"], "moleculenet_labels": {"FDA_APPROVED": 1, "CT_TOX": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01398", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMEGTGVVAVYGNGAITEAKKSPFSVKVGLAQMLRGGVIMDVVNAEQARIAEEAGACAVMALERVPADIRAQGGVARMSDPQMIKEIKQAVTIPVMAKARIGHFVEAQILEAIGIDYIDESEVLTLADEDHHINKHNFRIPFVCGCRNLGEALRRIREGAAMIRTKGEAGTGNIIEAVRHVRSVNGDIRVLRNMDDDEVFTFAKKLAAPYDLVMQTKQLGRLPVVQFAAGGVATPADAALMMQLGCDGVFVGSGIFKSGDPARRARAIVQAVTHYSDPEMLVEVSCGLGEAMVGINLNDEKVERFANRSEEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCHHHHHHHHHGGGTTEEEEEESSHHHHHHHHHTTCSEEEECSSCHHHHHHHCSCCCCCCHHHHHHHHHHCSSCEEEEEETTCHHHHHHHHHHTCSEEEEETTSCCSCSSCCCCGGGCSSCEEEEESSHHHHHHHHHHTCSEEEECCCTTSSCCHHHHHHHHHHHHHHHHHHHSCGGGHHHHHHHHTCCHHHHHHHHHHTSCSSEEEBCSSCCSHHHHHHHHHHTCSEEEESSTTTTSSSHHHHHHHHHHHHHTTTCHHHHHHHHTTCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01399", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nY4364_PSEA8\n\n### Answer:", "output": "Belongs to the UPF0761 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01400", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLTIALPKGRIGEDSLALFEKIFDTKFEFGKRKLILEAGGFRFMKVRNQDVPTYVYHQAADLGIVGLDVLEEKRLDIMRLLDLGFGRCDICIGIKAEESLDFSKPSYKVATKMENITRDFFSKKAIPVEIIKLYGSIELAPLVGLADMIVDIVETGETMRQNGLKPALKIMESSAFLIANKNSFYQKKAQILHLREQMSKVLYGS\n\n### Sequence 2:\nMITVALPKGRIAEDTLEIFRKIFGSSFMFEDRKLILEEGNFRFLMVRNQDIPTYVTEGAADIGVVGLDVLEEHKPNVVRLLDLQIGKCKVCIGIKNEEELDFSRSELKIATKMPNITRNYFAKLAVGVKIIKLYGSIELAPLVGLSDAIVDVVETGSTMKQNGLKVAGDIMQSSAYLIANKNSFIIKKDEILELYQKI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01401", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMSASSERELYEAWVELLSWMREYAQAKGVRFEKEADFPDFIYRMERPYDLPTTIMTASLSDGLGEPFLLADVSPRHAKLKRIGLRLPRAHIHLHAHYEPGKGLVTGKIPLTKERFFALADRAREALAFA\n\n### Answer:", "output": "<SEQ_2D>CCCCHHHHHHHHHHHHHHHHHHHHHHTTCCEEEEEECHHHHSCCSSCCCCSCCEEEEEEECSSSCEEEEEEECCTTCBSCCEEEEEGGGTEEEEEEEETTTEEEETTEECCHHHHHHHHHHHHHHHCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01402", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAIKEKQVQDYGENPIEVRDSDHYQNEYIEGFVEKWDELINWHARSSSEGEFFIKTLKEHGAKRVLDAATGTGFHSIRLIEAGFDVASVDGSVEMLVKAFENATRKDQILRTVHSDWRQVTRHIQERFDAVICLGNSFTHLFSEEDRRKTLAEFYSVLKHDGILILDQRNYDLILDEGFKSKHTYYYCGDNVKAEPEYVDDGLARFRYEFPDQSVYHLNMFPLRKDYVRRLLHEVGFQDITTYGDFQETYHQDDPDFYIHVAKKD\n\n### Sequence 2:\nMKKFYIAANWKMNMNRAEAKQLATEMKAGLKDGKNKYMIAPSFTLLQDVASVLKGSNILLGAQNMGLEEKGAHTGEVSVLQLLDVGVQAVILGHSERRHIYKETDDLINKKVKLALKHGLEVILCVGELLEEREAGHAEAVCERQIKKGLAEVSAKDLDKVTIAYEPVWAIGTGKNASPEDADAIHSSIRKTLAALYGEKAAKNMIIQYGGSMKPENAAGLLKKHNIDGGLIGGAGLKTETFLPIALFSE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01403", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "IRX2, GCG, TTR", "output": "This profile is most consistent with pancreatic alpha cell, which is commonly found in liver under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "liver", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["IRX2", "GCG", "TTR"]}}
{"id": "eval_01404", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CN1C(=O)[C@@](c2ccc(OC(F)F)cc2)(c2cccc(OCC3CC3)c2)N=C1N", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, ether, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BACE", "canonical_smiles": "CN1C(=O)[C@@](c2ccc(OC(F)F)cc2)(c2cccc(OCC3CC3)c2)N=C1N", "compound_name": null, "pubchem_cid": null, "MW": 401.41, "LogP": 3.11, "TPSA": 77.15, "QED": 0.77, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "ether", "halide"]}}
{"id": "eval_01405", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSQSLIVALDFPGKQEVEQFLHHFEGEELFVKVGMELFYKEGPAIITYLKEKGHKIFLDLKLHDIPNTVKSAMRSLASLDVDMVNVHAAGGSSMMKAAIEGLEEGKQEGKERPICIAVTQLTSTSEAMMKKEIGIEKTLEEAVAHYAKLTKESGLDGVVCSTLEVPKLREVCGNEFVTVTPGIRLASDDVNDQVRVATPKRARELGSSYIVVGRSITKAENPLEAYKTVKQQWEGVTV\n\n### Sequence 2:\nMFGNAFGVKKRRSDEAEKPFWISYADLMTAMMVLFLVVMVASLSSVTQRIQRAEQGEKTRGQDISRLCERLELHARNVNKTIVVDCHDNRISFGEAGRFDHNQFFLNAEGQKALQDVVPLVLEASNSEEGKKWFKQIVIEGFTDTDGSYLYNLHLSLQRSEWVMCSLLDSRSPLQKNISAEQQLQIRKLFLAGGVSFNNAKESKEASRRVELRMQFFGLKDKRDKADEVDFPPVVNKEVCQLVMPL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01406", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: Cc1ccc(NC(=O)c2ccccc2F)cc1-n1cnc2ccc(N3CCN(C)CC3)cc2c1=O\nDescriptors: MW=471.54, LogP=3.84, TPSA=70.47, HBD=1, HBA=5", "output": "Descriptors: MW=471.54, LogP=3.84, TPSA=70.47, HBD=1, HBA=5, QED=0.49, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "Cc1ccc(NC(=O)c2ccccc2F)cc1-n1cnc2ccc(N3CCN(C)CC3)cc2c1=O", "compound_name": null, "pubchem_cid": null, "MW": 471.54, "LogP": 3.84, "TPSA": 70.47, "QED": 0.49, "functional_groups": ["aromatic ring", "amide", "amine", "halide"], "moleculenet_labels": {"label": 2.75}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01407", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nG3P2_BACCE\n\n### Answer:", "output": "Belongs to the glyceraldehyde-3-phosphate dehydrogenase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01408", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMMFEHVLFLSVYLFSIGIYGLITSRNMVRALICLELILNSINLNLVTFSDLFDSRQLKGDIFAIFVIALAAAEAAIGLSILSSIHRNRKSTRINQSNLLN\n\n### Answer:", "output": "<SEQ_3Di>DVLPVQLVCLCVQLVVLVVQLVPPFFLVSNLVSVLSNLVSVLSVQVSVQPVDVPVPRPRVVVSVVSVVVSVVVSVVSVVVVVVVCVVPVDGGPVPDPPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01409", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nS100A8 S100A9 CSF3R CXCR2 FCGR3B\n</cell>", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as S100A8, CSF3R, CXCR2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "kidney", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["S100A8", "S100A9", "CSF3R", "CXCR2", "FCGR3B"]}}
{"id": "eval_01410", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSPD13_HUMAN\n\n### Answer:", "output": "Belongs to the Speedy/Ringo family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01411", "category": "Mutation", "instruction": "Describe the effect of mutation P00952-K230A on protein Tyrosine--tRNA ligase.", "input": "### Wild-type:\nMDLLAELQWRGLVNQTTDEDGLRKLLNEERVTLYCGFDPTADSLHIGHLATILTMRRFQQAGHRPIALVGGATGLIGDPSGKKSERTLNAKETVEAWSARIKEQLGRFLDFEADGNPAKIKNNYDWIGPLDVITFLRDVGKHFSVNYMMAKESVQSRIETGISFTEFSYMMLQAYDFLRLYETEGCRLQIGGSDQWGNITAGLELIRKTKGEARAFGLTIPLVTKADGTKFGKTESGTIWLDKEKTSPYEFYQFWINTDDRDVIRYLKYFTFLSKEEIEALEQELREAPEKRAAQKTLAE\n\n### Mutant:\nMDLLAELQWRGLVNQTTDEDGLRKLLNEERVTLYCGFDPTADSLHIGHLATILTMRRFQQAGHRPIALVGGATGLIGDPSGKKSERTLNAKETVEAWSARIKEQLGRFLDFEADGNPAKIKNNYDWIGPLDVITFLRDVGKHFSVNYMMAKESVQSRIETGISFTEFSYMMLQAYDFLRLYETEGCRLQIGGSDQWGNITAGLELIRKTKGEARAFGLTIPLVTKADGTAFGKTESGTIWLDKEKTSPYEFYQFWINTDDRDVIRYLKYFTFLSKEEIEALEQELREAPEKRAAQKTLAE\n\n### Answer:", "output": "Decreases the binding affinity between tRNA(Tyr) and TyrRS-Tyr-AMP complex. The K(d)(tRNA) value for the variant is increased relative to the wild-type enzyme.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01412", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPKIFRKKVCTTFITTEKPTDAYGDLDFTYSGRKHSNFLRLSDRTDPATVYSLVTRSWGFRAPNLVVSVLGGSGGPVLQTWLQDLLRRGLVRAAQSTGAWIVTGGLHTGIGRHVGVAVRDHQTASTGSSKVVAMGVAPWGVVRNRDMLINPKGSFPARYRWRGDPEDGVEFPLDYNYSAFFLVDDGTYGRLGGENRFRLRFESYVAQQKTGVGGTGIDIPVLLLLIDGDEKMLKRIEDATQAQLPCLLVAGSGGAADCLVETLEEARDRIRRYFPKGDPEVLQAQVERIMTRKELLTVYSSEDGSEEFETIVLRALVKACDELRLAVAWNRVDIAQSELFRGDIQWRSFHLEASLMDALLNDRPEFVRLLISHGLSLGHFLTPVRLAQLYSAVSPNSLIRNLLDQASHAPNVGQVLRTLLQAPWSDLLIWALLLNRAQMAIYFWEKGSNSVASALGACLLLRVMARLESEAEEAARRKDLAATFESMSVDLFGECYHNSEERAARLLLRRCPLWGEATCLQLAMQADARAFFAQDGVQSLLTQKWWGEMDSTTPIWALLLAFFCPPLIYTNLIVFRKSKRWSDFWGAPVTAFLGNVVSYLLFLLLFAHVLLVDFQPTKPSVSELLLYFWAFTLLCEELRQGLGLRHRLHLYLSDTWNQCDLLALTCFLLGVGCRLTPGLFDLGRTVLCLDFMIFTLRLLHIFTVNKQLGPKIVIVSKMMKDVFFFLFFLCVWLVAYGVATEGILRPQDRSLPSILRRVFYRPYLQIFGQIPQEEMDVALMIPGNCSMERGSWAHPEGPVAGSCVSQYANWLVVLLLIVFLLVANILLLNLLIAMFSYTFSKVHGNSDLYWKAQRYSLIREFHSRPALAPPLIIISHVRLLIKWLRRLSKEAERKLLTWESVHKENFLLAQARDKRDSDSERLKRTSQKVDTALKQLGQIREYDRRLRGLEREVQHCSRVLTWMAEALS\n\n### Answer:", "output": "<SEQ_3Di>DQVCQQAPFWACLVPIGGAAFWAWADEPDPDDDTATEGEGAPPDACVVVVCCVCPVVNDDQFQEEEEEAEFPPWPDDDVQVLLQLQFQPFVLCLLLPHEYEYAFFRGGCNVSNLVSQVVVVVVDPSVDGRQYEHEDEQLAEDDCVLGRGRPDNDHRYDDDPVDDDDQQHTHGGDRHRYYYYYYPNDSGDPDSTVSSVLSNQVVQQVAADDLVGQAHRRAYEYEYAHHAPVCLVVVLVNVVSLHAYEYECHDLVRVLLVVVVVDVCQLVSQCPRHVDDRPVVSVSSVSVVVPCLRYYYHYPDPGSPPVVVVSVVNVVSSDNVLRSCLSSLPPVVNVVVVVDPPPPDDAVSLLSSCLSCQLDPRLPVVVVSVVVPHDDVVRDDVVSVVVSVVSADCPFDVNVLVVVQVVVSRVVSSVVVQPVVCLVVQLVSCLVRPDQSSNVVSLLVDFPRLLVLLVSLQVLQSVLVVGDDPVSSVSSNVSSVVSLVVSQVLLVVVCVVPLVLSLCSQPPADPDDHGDGSLVSCQNNVSLVNLPDDSNVVSLLLLQFQFWQSPQDVVLLVVCVVPVVVLPDPRIHGHVVVSVCRSLQHLVNLLVLAVVLLVVLLVLLVCCLPPDPDLPDDDPSVVLNVVLLVLVVVVLVCCQPCDVVSVVVQVVDDVSVLVVVLSVLVVVLVVQSPPNPRNVVSSVSSVVSSVSSVVNVLVSPLLDDVVNLVSVCVVVLVVLVVVLVVVCVVVLQVLQVVLCVVKPADPDDPVLSVCSRRPQLVVVLVPDHDDQLQAQVNFAWDDPDDDDAGWTAGDDPRSHGIIGPPPSVVNVVSVVVSNCCRNPVRVVVSVVVVVVVCVVCSVVSVSVSSSCSVVVSVVSNPDDSDHSSPCVVVVVVVVVVVVVPDDDVSVVVSRVVSSVSSVVVVVVVVVVVCPDPVNVVVVVVVVVVVVVVVCVVVVVVVVVVVCVVVVVVVVVVVVVVVVVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01413", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMDKNELVQKAKLAEQAERYDDMAACMKSVTEQGAELSNEERNLLSVAYKNVVGARRSSWRVVSSIEQKTEGAEKKQQMAREYREKIETELRDICNDVLSLLEKFLIPNASQAESKVFYLKMKGDYYRYLAEVAAGDDKKGIVDQSQQAYQEAFEISKKEMQPTHPIRLGLALNFSVFYYEILNSPEKACSLAKTAFDEAIAELDTLSEESYKDSTLIMQLLRDNLTLWTS\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHHHHHHHHHTTCHHHHHHHHHHHHHTCSCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTTTTCCSHHHHHHHHHHHHHHHHHHHHHCCSSHHHHHHHHHHHHHHHHHHHHHHHSCTTCHHHHHHHHHHHHHHHHTSCCHHHHHHHHHHHHHHHHHCCCCCCCCSCHHHHHHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01414", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "O=C1C2(Cl)C3(Cl)C4(Cl)C(Cl)(Cl)C5(Cl)C3(Cl)C1(Cl)C5(Cl)C24Cl", "output": "The molecule contains the following functional groups: ketone, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "ESOL", "canonical_smiles": "O=C1C2(Cl)C3(Cl)C4(Cl)C(Cl)(Cl)C5(Cl)C3(Cl)C1(Cl)C5(Cl)C24Cl", "compound_name": null, "pubchem_cid": null, "MW": 490.64, "LogP": 4.62, "TPSA": 17.07, "QED": 0.45, "functional_groups": ["ketone", "halide"], "moleculenet_labels": {"label": -5.259}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["ketone", "halide"]}}
{"id": "eval_01415", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "Cc1ccccc1Nc1ncnc2ccccc12", "output": "The molecule contains the following functional groups: aromatic ring.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "Cc1ccccc1Nc1ncnc2ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 235.29, "LogP": 3.68, "TPSA": 37.81, "QED": 0.74, "functional_groups": ["aromatic ring"], "moleculenet_labels": {"label": 2.85}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring"]}}
{"id": "eval_01416", "category": "Mutation", "instruction": "Describe the effect of mutation O43520-G308V on protein Phospholipid-transporting ATPase IC.", "input": "### Wild-type:\nMSTERDSETTFDEDSQPNDEVVPYSDDETEDELDDQGSAVEPEQNRVNREAEENREPFRKECTWQVKANDRKYHEQPHFMNTKFLCIKESKYANNAIKTYKYNAFTFIPMNLFEQFKRAANLYFLALLILQAVPQISTLAWYTTLVPLLVVLGVTAIKDLVDDVARHKMDKEINNRTCEVIKDGRFKVAKWKEIQVGDVIRLKKNDFVPADILLLSSSEPNSLCYVETAELDGETNLKFKMSLEITDQYLQREDTLATFDGFIECEEPNNRLDKFTGTLFWRNTSFPLDADKILLRGCVI\n\n### Mutant:\nMSTERDSETTFDEDSQPNDEVVPYSDDETEDELDDQGSAVEPEQNRVNREAEENREPFRKECTWQVKANDRKYHEQPHFMNTKFLCIKESKYANNAIKTYKYNAFTFIPMNLFEQFKRAANLYFLALLILQAVPQISTLAWYTTLVPLLVVLGVTAIKDLVDDVARHKMDKEINNRTCEVIKDGRFKVAKWKEIQVGDVIRLKKNDFVPADILLLSSSEPNSLCYVETAELDGETNLKFKMSLEITDQYLQREDTLATFDGFIECEEPNNRLDKFTGTLFWRNTSFPLDADKILLRGCVI\n\n### Answer:", "output": "In PFIC1; greatly reduced expression due to proteosomal degradation; abolishes interaction with TMEM30A. The mutation causes progressive familial intrahepatic cholestasis type 1 (PFIC1). The mutant protein does not interact with CDC50A and is not detectable in the canalicular membrane.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01417", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFRL5_ARATH\n\n### Answer:", "output": "FRIGIDA-like protein 5.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01418", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRSMA_RUEST\n\n### Answer:", "output": "Specifically dimethylates two adjacent adenosines (A1518 and A1519) in the loop of a conserved hairpin near the 3'-end of 16S rRNA in the 30S particle.  May play a critical role in biogenesis of 30S subunits.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01419", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMKTYVPKQVEPRWVLIDAEGKTLGRLATKIATLLRGKHRPDWTPNVAMGDFVVVVNADKIRVTGKKLEQKIYTRYSGYPGGLKKIPLEKMLATHPERVLEHAVKGMLPKGPLGRRLFKRLKVYAGPDHPHQAQRPEKLEV\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCEEECSSSBTTTTHHHHHHTTTTCSSSCCCTTTCCCCCCEEESCSCCBCSSSHHHHCEEEEECSSTTCEEEEEHHHHHHSCTHHHHHHHHHTTSCCSHHHHHHHHHCEECSSSCCCCCSCCCBCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01420", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMATK_ARAHI\n\n### Answer:", "output": "Usually encoded in the trnK tRNA gene intron.  Probably assists in splicing its own and other chloroplast group II introns.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01421", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMVTRFLGPRYRELVKNWVPTAYTWGAVGAVGLVWATDWRLILDWVPYINGKFKKDN\n\n### Sequence 2:\nMLTRFLGPRYRQLARNWVPTASLWGAVGAVGLVWATDWRLILDWVPYINGKFKKDD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01422", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nKLF1:low GATA1:low HBB:medium HBA1:low\n</cell>", "output": "This cell is most consistent with erythroid progenitor, an erythroid progenitor cell. Key markers such as GATA1, HBB, HBA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "erythroid progenitor", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["KLF1", "GATA1", "HBB", "HBA1"]}}
{"id": "eval_01423", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTNTPSDQPLKQVNQRGAARLAAVQALYQMDVGGTGVLEIVAEYEEHRLGKELDGDTYLRADASWFRSIVAGVVRDQRKLDPLIGSALQDDWALSRLDSTVRAILRAGTFEILERKDVPVPVIVTEYVEIAKAFFQDEEPKLVNAVLDRIAKQVRGDQRK\n\n### Sequence 2:\nMKLSDIDKGTYAAVKFDDSTLDMFQALQQIMELFNPVPRDKLHSTICFSRVKIPYIPLTEKMPIGSTHKLEVFEHNGKRALVVLLDSPYLESRHEYANILGATFDFPTYNPHVTLAYDIGAMEIPKHGVTGNPVVITHEYTEDLDLNWKP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01424", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPRAPRTYSKTYSTPKRPYESSRLDAELKLAGEFGLKNKKEIYRISFQLSKIRRAARDLLTRDEKDPKRLFEGNALIRRLVRVGVLSEDKKKLDYVLALKVEDFLERRLQTQVYKLGLAKSVHHARVLITQRHIAVGKQIVNIPSFMVRLDSEKHIDFAPTSPFGGARPGRVARRNAARKAEASGEAADEADEADEE\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCSSTTCHHHHHHHHHHHHHTTCSSTHHHHHHHHHHHHHHHHHHHHHHSCTTCHHHHHHHHHHHHHHHHTTSSCTTCCSHHHHTTCCHHHHHTTCHHHHHHHSSTTSSHHHHHHHHTTTCCEETTEECCCTTCCCCGGGGGGEECCSSSGGGSSSCCTTHHHHTTTTTHHHHCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01425", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nMRC1 APOE SPP1 CD206\n</cell>", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as MRC1, SPP1, APOE support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "lymph node", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MRC1", "APOE", "SPP1", "CD206"]}}
{"id": "eval_01426", "category": "Mutation", "instruction": "Describe the effect of mutation Q9LSL5-D475N on protein L-type lectin-domain containing receptor kinase IX.2.", "input": "### Wild-type:\nMLYFIFCQNLSSSSSMSNSILFLSLFLFLPFVVDSLYFNFTSFRQGDPGDIFYHGDATPDEDGTVNFNNAEQTSQVGWITYSKKVPIWSHKTGKASDFSTSFSFKIDARNLSADGHGICFFLAPMGAQLPAYSVGGFLNLFTRKNNYSSSFPLVHVEFDTFNNPGWDPNDVGSHVGINNNSLVSSNYTSWNASSHSQDICHAKISYDSVTKNLSVTWAYELTATSDPKESSSLSYIIDLAKVLPSDVMFGFIAAAGTNTEEHRLLSWELSSSLDSDKADSRIGLVIGISASGFVFLTFMV\n\n### Mutant:\nMLYFIFCQNLSSSSSMSNSILFLSLFLFLPFVVDSLYFNFTSFRQGDPGDIFYHGDATPDEDGTVNFNNAEQTSQVGWITYSKKVPIWSHKTGKASDFSTSFSFKIDARNLSADGHGICFFLAPMGAQLPAYSVGGFLNLFTRKNNYSSSFPLVHVEFDTFNNPGWDPNDVGSHVGINNNSLVSSNYTSWNASSHSQDICHAKISYDSVTKNLSVTWAYELTATSDPKESSSLSYIIDLAKVLPSDVMFGFIAAAGTNTEEHRLLSWELSSSLDSDKADSRIGLVIGISASGFVFLTFMV\n\n### Answer:", "output": "In LecRK-IX.2-RN; loss of kinase activity and impaired ability to mediate resistance to Phytophthora. The mutation in LecRK-IX. 1 is associated with compromised Phytophthora resistance.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01427", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nTPSAB1 CPA3 HPGDS KIT\n</cell>", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as HPGDS, KIT, TPSAB1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "colon", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TPSAB1", "CPA3", "HPGDS", "KIT"]}}
{"id": "eval_01428", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTHIE_ROSCS\n\n### Answer:", "output": "Thiamine-phosphate synthase, TP synthase, TPS, Thiamine-phosphate pyrophosphorylase, TMP pyrophosphorylase, TMP-PPase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01429", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: Cc1onc(-c2ccccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)(C)[C@@H]2C(=O)[O-]\nDescriptors: MW=434.88, LogP=1.21, TPSA=115.57, HBD=1, HBA=7", "output": "Descriptors: MW=434.88, LogP=1.21, TPSA=115.57, HBD=1, HBA=7, QED=0.72, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "ClinTox", "canonical_smiles": "Cc1onc(-c2ccccc2Cl)c1C(=O)N[C@@H]1C(=O)N2[C@@H]1SC(C)(C)[C@@H]2C(=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 434.88, "LogP": 1.21, "TPSA": 115.57, "QED": 0.72, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"FDA_APPROVED": 1, "CT_TOX": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01430", "category": "Mutation", "instruction": "Describe the effect of mutation O17185-S235F on protein Two pore potassium channel protein sup-9.", "input": "### Wild-type:\nMKRQNIRTLSLIVCTLTYLLVGAAVFDALETENEILQRKLVQRVREKLKTKYNMSNADYEILEATIVKSVPHKAGYQWKFSGAFYFATTVITTIGYGHSTPMTDAGKVFCMLYALAGIPLGLIMFQSIGERMNTFAAKLLRFIRRAAGKQPIVTSSDLIIFCTGWGGLLIFGGAFMFSSYENWTYFDAVYYCFVTLTTIGFGDYVALQKRGSLQTQPEYVFFSLVFILFGLTVISAAMNLLVLRFLTMNTEDERRDEQEAILAAQGLVRVGDPTADDDFGRLPLSDNVSLASCSCYQLPD\n\n### Mutant:\nMKRQNIRTLSLIVCTLTYLLVGAAVFDALETENEILQRKLVQRVREKLKTKYNMSNADYEILEATIVKSVPHKAGYQWKFSGAFYFATTVITTIGYGHSTPMTDAGKVFCMLYALAGIPLGLIMFQSIGERMNTFAAKLLRFIRRAAGKQPIVTSSDLIIFCTGWGGLLIFGGAFMFSSYENWTYFDAVYYCFVTLTTIGFGDYVALQKRGSLQTQPEYVFFSLVFILFGLTVIFAAMNLLVLRFLTMNTEDERRDEQEAILAAQGLVRVGDPTADDDFGRLPLSDNVSLASCSCYQLPD\n\n### Answer:", "output": "In n189 and n2358; loss of function.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01431", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "PRF1, GZMB, CD3D, NKG7", "output": "This cell is most consistent with cytotoxic T cell, an effector cytotoxic CD8+ T cell. Key markers such as NKG7, GZMB, PRF1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "brain", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PRF1", "GZMB", "CD3D", "NKG7"]}}
{"id": "eval_01432", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nKLPTNLAYERSIDPSDVCFFVVWPRKTPLTYNSRTLLGQMEAASLAYDVSGQPIKSATAEALAQGNPHQVDFCHVPYGASHIECSFSVSFSSELRQPYKCNSSKVKQTLVQLVELYETKIGWTELATRYLMNICNGKWLWKNTRKAYCWNIVLTPWPWNGEKVGFEDIRTNYTSRQDFKNNKNWSAIVEMIKTAFSSTDGLAIFEVRATLHLPTNAMVRPSQVFTNSRVFQSTTIDGERSPILGAFKTGAAIATIDDWYPEATEPLRVGRFGVHREDVTCYRHPSTGKDFFSILQQAEHYIEVLSANKTPAQETINDMHFLMANLIKGGMFQH\n\n### Answer:", "output": "<SEQ_3Di>DFALWFWFDAQKFKAKKFKWFAACDIDGFAKDKDKDKDFQPDPCVFAPPPHRTDPPPDPPNRGDIDMFIAIFTWHDFRTFWMKIKIKMKGAQRLQAIPDDPDPVLRVLSNLLVVLCCVFVRCLVVLLLLVCLVFLCLLQPVVVPQFPFKGKWKDKPPDDDDIQTRGRSVPQVPDSVSVVVGPCPVVLSVLNRCRNPDHGGMMMMMIMMITGHDGGGTRQEDADPDSPHHFDWDDDPNHTHGIHGRVSSVQRSQQRQQLFPDGDGTGRQDQFQADDVDRDTGNDVVVQLGLVSLSVCSVVLSVQSPDPDRDDPNSVSSSVSNVSCSNVIGIPSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01433", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nMYH7:high MYH6:high TNNI3:medium\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as TNNI3, MYH6, MYH7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["MYH7", "MYH6", "TNNI3"]}}
{"id": "eval_01434", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: ALK\nLigand SMILES: O=C(NCCCNc1nc(Nc2cccc(NC(=O)N3CCCC3)c2)ncc1Br)c1cccs1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "O=C(NCCCNc1nc(Nc2cccc(NC(=O)N3CCCC3)c2)ncc1Br)c1cccs1", "compound_name": null, "pubchem_cid": null, "MW": 544.48, "LogP": 4.9, "TPSA": 111.28, "QED": 0.29, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": 4.13}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01435", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nKDSB_SHESH\n\n### Answer:", "output": "8-amino-3,8-dideoxy-manno-octulosonate cytidylyltransferase, CMP-8-amino-3,8-dideoxy-manno-octulosonate synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01436", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMIASKFGIGQQVRHSLHGYLGVVIDIDPEYSLAPPEPDEVANNKTLRSSPWYHVVIEDDDGQPVHTYLAEAQLTYEDVDAHPEQPSLDELAASIRHQLQAPHLRN\n\n### Sequence 2:\nMNSILGLIDTVTNTVGKGQQIELDKAALGQQRELALQRMSLDRQALNNQVEQFNKLLEQRVHGPIQSVRLARAAGFRVDPYSYTNQNFYDDQLNAIRLSYRNLFKN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01437", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSNX2B_ARATH\n\n### Answer:", "output": "Plays a role in vesicular protein sorting.  Acts at the crossroads between the secretory and endocytic pathways.  Is involved in the endosome to vacuole protein transport and, as component of the membrane-associated retromer complex, is also involved in endosome-to-Golgi retrograde transport.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01438", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "GZMB, KLRD1, NKG7", "output": "The expression pattern is consistent with a healthy natural killer cell without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "natural killer cell", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["GZMB", "KLRD1", "NKG7"]}}
{"id": "eval_01439", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTVPKEIPEKWARAGAPPSWSQKKPSWGTEEERRARANDREYNEKFQYASNCIKTSKYNIVTFLPVNLFEQFQEVANTYFLFLLILQLIPQISSLSWFTTIVPLVLVLTITAVKDATDDYFRHKSDNQVNNRHSQVLINGVLQQEQWMNVCVGDIIKLENNQFVAADLLLLSSSEPHGLCYIETAELDGETNMKVRQAIPVTSELGDVSQLARFDGEVICEPPNNKLDKFSGTLYWKENKFPLSNQNMLLRGCVLRNTEWCFGLVIFAGPDTKLMQNSGRTKFKRTSIDRLMNTLVLWIFGFLVCMGVILAIGNAIWEHEVGTRFQVYLPWDEAVDSAFFSGFLSFWSYIIILNTVVPISLYVSVEVIRLGHSYFINWDKKMFCMKKRTPAEARTTTLNEELGQVEYIFSDKTGTLTQNIMVFNKCSINGHSYGDVFDVLGHKAELGERPEPVDFSFNPLADKKFLFWDSSLLEAVKMGDPHTHEFFRLLSLCHTVMSEEKNEGELYYKAQSPDEGALVTAARNFGFVFRSRTPKTITVHELGTAITYQLLAILDFNNIRKRMSVIVRNPEGKIRLYCKGADTILLDRLHPPTQELLSSTTDHLNEYAGDGLRTLVLAYKDLDEEYYEEWARRRLQASLAQDSREDRLASIYEEVESDMMLLGATAIEDKLQQGVPETIALLTLANIKIWVLTGDKQETAVNIGYSCKMLTDDMTEVFVVTGHTVLEVREELRKARKKMVDSSHAVGNGFTYQGNLSSSKLTSVLEAVAGEYALVINGHSLAHALEADMELEFLETACACKAVICCRVTPLQKAQVVELVKKYKKAVTLAIGDGANDVSMIKTAHIGVGISGQEGIQAVLASDYSFSQFKFLQRLLLVHGRWSYLRMCKFLCYFFYKNFAFTMVHFWFGFFCGFSAQTVYDQYFITLYNIVYTSLPVLAMGVFDQDVPEQRSMEYPKLYEPGQLNLLFNKREFFICIAQGIYTSVLMFFIPYGVFAEATRDDGTQLADYQSFAVTVATSLVIVVSVQIGLDTGYWTAINHFFIWGSLAVYFAILFAMHSNGLFDMFPNQFRFVGNAQNTLAQPTVWLTIALTTAVCIMPVVAFRFLRLSLKPDLSDTVRYTQLVRKKQKAQHRCMRRVGRTGSRRSGYAFSHQEGFGELIMSGKNMRLSSLALSSFSTRSSSSWIESLRRKKSDSANSPSGGAEKPLKG\n\n### Sequence 2:\nMTVPKEIPEKWARAGAPPSWSQKKPSWGTEEERRARANDREYNEKFQYASNCIKTSKYNIVTFLPVNLFEQFQEVANTYFLFLLILQLIPQISSLSWFTTIVPLVLVLTITAVKDATDDYFRHKSDNQVNNRHSQVLINGVLQQEQWMNVCVGDIIKLENNQFVAADLLLLSSSEPHGLCYIETAELDGETNMKVRQAIPVTSELGDVSQLARFDGEVICEPPNNKLDKFSGTLYWKENKFPLSNQNMLLRGCVLRNTEWCFGLVIFAGPDTKLMQNSGRTKFKRTSIDRLMNTLVLWIFGFLVCMGVILAIGNAIWEHEVGTRFQVYLPWDEAVDSAFFSGFLSFWSYIIILNTVVPISLYVSVEVIRLGHSYFINWDKKMFCMKKRTPAEARTTTLNEELGQVEYIFSDKTGTLTQNIMVFNKCSINGHSYGDVFDVLGHKAELGERPEPVDFSFNPLADKKFLFWDSSLLEAVKMGDPHTHEFFRLLSLCHTVMSEEKNEGELYYKAQSPDEGALVTAARNFGFVFRSRTPKTITVHELGTAITYQLLAILDFNNIRKRMSVIVRNPEGKIRLYCKGADTILLDRLHPPTQELLSSTTDHLNEYAGDGLRTLVLAYKDLDEEYYEEWARRRLQASLAQDSREDRLASIYEEVESDMMLLGATAIEDKLQQGVPETIALLTLANIKIWVLTGDKQETAVNIGYSCKMLTDDMTEVFVVTGHTVLEVREELRKARKKMVDSSHAVGNGFTYQGNLSSSKLTSVLEAVAGEYALVINGHSLAHALEADMELEFLETACACKAVICCRVTPLQKAQVVELVKKYKKAVTLAIGDGANDVSMIKTAHIGVGISGQEGIQAVLASDYSFSQFKFLQRLLLVHGRWSYLRMCKFLCYFFYKNFAFTMVHFWFGFFCGFSAQTVYDQYFITLYNIVYTSLPVLAMGVFDQDVPEQRSMEYPKLYEPGQLNLLFNKREFFICIAQGIYTSVLMFFIPYGVFAEATRDDGTQLADYQSFAVTVATSLVIVVSVQIGLDTGYWTAINHFFIWGSLAVYFAILFAMHSNGLFDMFPNQFRFVGNAQNTLAQPTVWLTIALTTAVCIMPVVAFRFLRLSLKPDLSDTVRYTQLVRKKQKAQHRCMRRVGRTGSRRSGYAFSHQEGFGELIMSGKNMRLSSLALSSFSTRSSSSWIESLRRKKSDSANSPSGGAEKPLKG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01440", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFLIW2_HELPH\n\n### Answer:", "output": "Flagellar assembly factor FliW 2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01441", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL19_LEPCP\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bL19 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01442", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "S100A8, FCGR3B, CXCR2, CSF3R, S100A9", "output": "The marker profile suggests functional enrichment consistent with the typical biology of neutrophil.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "blood", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_pathway", "top_genes": ["S100A8", "FCGR3B", "CXCR2", "CSF3R", "S100A9"]}}
{"id": "eval_01443", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C1Nc2ccc(Cl)cc2C(c2ccccc2Cl)=NC1O", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, hydroxyl, halide. Its descriptor profile: MW=321.16, LogP=3.1, TPSA=61.69, HBD=2, HBA=3, QED=0.85.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "O=C1Nc2ccc(Cl)cc2C(c2ccccc2Cl)=NC1O", "compound_name": null, "pubchem_cid": null, "MW": 321.16, "LogP": 3.1, "TPSA": 61.69, "QED": 0.85, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01444", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\n3SI1A_DENAN\n\n### Answer:", "output": "Belongs to the snake three-finger toxin family. Short-chain subfamily. Aminergic toxin sub-subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01445", "category": "Mutation", "instruction": "Describe the effect of mutation P00915-G254R on protein Carbonic anhydrase 1.", "input": "### Wild-type:\nMASPDWGYDDKNGPEQWSKLYPIANGNNQSPVDIKTSETKHDTSLKPISVSYNPATAKEIINVGHSFHVNFEDNDNRSVLKGGPFSDSYRLFQFHFHWGSTNEHGSEHTVDGVKYSAELHVAHWNSAKYSSLAEAASKADGLAVIGVLMKVGEANPKLQKVLDALQAIKTKGKRAPFTNFDPSTLLPSSLDFWTYPGSLTHPPLYESVTWIICKESISVSSEQLAQFRSLLSNVEGDNAVPMQHNNRPTQPLKGRTVRASF\n\n### Mutant:\nMASPDWGYDDKNGPEQWSKLYPIANGNNQSPVDIKTSETKHDTSLKPISVSYNPATAKEIINVGHSFHVNFEDNDNRSVLKGGPFSDSYRLFQFHFHWGSTNEHGSEHTVDGVKYSAELHVAHWNSAKYSSLAEAASKADGLAVIGVLMKVGEANPKLQKVLDALQAIKTKGKRAPFTNFDPSTLLPSSLDFWTYPGSLTHPPLYESVTWIICKESISVSSEQLAQFRSLLSNVEGDNAVPMQHNNRPTQPLKRRTVRASF\n\n### Answer:", "output": "In Guam. The amino acid analysis of the tryptic peptides indicated that the mutation leads to the substitution of glycine (G) with arginine (R) in the carbonic anhydrase-1 (CA1) protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01446", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMDCC_ACICA\n\n### Answer:", "output": "Malonate decarboxylase acyl carrier protein, Malonate decarboxylase subunit delta.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01447", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMETRRIRKVLVANRGEIAIRVFRACTELGIRTVAIYSKEDVGSYHRYKADEAYLVGEGKKPIEAYLDIEGIIEIAKAHDVDAIHPGYGFLSENIQFAKRCREEGIIFIGPNENHLDMFGDKVKARHAAVNAGIPVIPGSDGPVDGLEDVVAFAEAHGYPIIIKAALGGGGRGMRIVRSKSEVKEAFERAKSEAKAAFGSDEVYVEKLIENPKHIEVQILGDYEGNIVHLYERDCSVQRRHQKVVEVAPSVSLSDELRQRICEAAVQLMRSVGYVNAGTVEFLVSGDEFYFIEVNPRIQVEHTITEMITGIDIVQSQILIADGCSLHSHEVGIPKQEDIRINGYAIQSRVTTEDPLNNFMPDTGKIMAYRSGGGFGVRLDAGNGFQGAVITPYYDSLLVKLSTWALTFEQAARKMLRNLREFRIRGIKTNIPFLENVVQHPKFLSGEYDTSFIDTTPELFVF\n\n### Answer:", "output": "<SEQ_2D>CCCCCCSEEEECSCHHHHHHHHHHHHHHTCEEEEEECGGGTTCTHHHHSSSEEECSTTSCTTGGGTCHHHHHHHHHHTTCCEEECCSSSSTTCHHHHHHHHHTTCEESSCCHHHHHHTTSHHHHHHHHHHTTCCBCCBCSSCCSSHHHHHHHHHHHCSCEEEEESTTCSSSSEEEECCGGGHHHHHHHHHHHHHHHTSCCCEEEEECCCSCEEEEEEEEECTTCCEEEEEEEEEEEEETTEEEEEEESCTTSCHHHHHHHHHHHHHHHHHTTCCEEEEEEEEEETTEEEEEEEESSCCGGGHHHHHHHCCCHHHHHHHHHTTCCTTSTTTCCCCGGGCCCCSEEEEEEEESEEGGGTTEECCEECSEEECCCCTTEEEEESSCSTTCEECSSSCCEEEEEEEEESSHHHHHHHHHHHHHTCEEESSCCSHHHHHHHHHSHHHHTSCCCTTHHHHCGGGGCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01448", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "TTR, IRX2, GCG", "output": "This profile is most consistent with pancreatic alpha cell, which is commonly found in prostate under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "prostate", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TTR", "IRX2", "GCG"]}}
{"id": "eval_01449", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "PF4, GP1BA, PPBP, ITGA2B", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as PPBP, PF4, GP1BA support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "lung", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PF4", "GP1BA", "PPBP", "ITGA2B"]}}
{"id": "eval_01450", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "FC(F)Cl", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains halide. Its descriptor profile: MW=86.47, LogP=1.45, TPSA=0.0, HBD=0, HBA=0, QED=0.39.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "FreeSolv", "canonical_smiles": "FC(F)Cl", "compound_name": null, "pubchem_cid": null, "MW": 86.47, "LogP": 1.45, "TPSA": 0.0, "QED": 0.39, "functional_groups": ["halide"], "moleculenet_labels": {"label": -0.5}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01451", "category": "Mutation", "instruction": "Describe the effect of mutation Q9UUJ6-D402G on protein Nuclear elongation and deformation protein 1.", "input": "### Wild-type:\nMQYVGRAFDSVTKTWNAINPSTLSGAIDVIVVEQEDKTLACSPFHVRFGKFSLLLPSDKKVEFSVNGQLTGFNMKLGDGGEAFFVFATENAVPRELQTSPIVSPTTSPKQTPSINVTEPQDLELDKVSQDHEKDQSNTYLMEDGYEFPLTRDLIRRSKSDADQTPPTGFKHLRHSSCLEMAGSDRTPSMPATTLADLRLLQKAKELGKRLSGKELPTRVGDNGDVMLDMTGYKSSAANINIAELARETFKDEFPMIEKLLREDEEGNLWFHASEDAKKFAEVYGHSPPASPSRTPASPKS\n\n### Mutant:\nMQYVGRAFDSVTKTWNAINPSTLSGAIDVIVVEQEDKTLACSPFHVRFGKFSLLLPSDKKVEFSVNGQLTGFNMKLGDGGEAFFVFATENAVPRELQTSPIVSPTTSPKQTPSINVTEPQDLELDKVSQDHEKDQSNTYLMEDGYEFPLTRDLIRRSKSDADQTPPTGFKHLRHSSCLEMAGSDRTPSMPATTLADLRLLQKAKELGKRLSGKELPTRVGDNGDVMLDMTGYKSSAANINIAELARETFKDEFPMIEKLLREDEEGNLWFHASEDAKKFAEVYGHSPPASPSRTPASPKS\n\n### Answer:", "output": "10-fold decrease in minichromosome loss. The substitution mutation affected the two-hybrid interaction with Dis3 and decreased chromosome stability, suggesting the functional unimportance of the interaction.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01452", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTKKKAHKPGSATIALNKRARHEYFIEEEFEAGLALQGWEVKSLRAGKANIGDSYVILKDGEAWLFGANFTPMAVASTHVVCDPTRTRKLLLNQRELDSLYGRINREGYTVVALSLYWKNAWCKVKIGVAKGKKQHDKRSDLKEREWQLDKARIMKNAGR\n\n### Sequence 2:\nMTKKKAHKPGSATIALNKRARHEYFIEEEFEAGLALQGWEVKSLRAGKANIGDSYVILKDGEAWLFGANFTPMAVASTHVVCDPTRTRKLLLNQRELDSLYGRINREGYTVVALSLYWKNAWCKVKIGVAKGKKQHDKRSDLKEREWQLDKARIMKHAGR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01453", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nARAA_ECOLI\n\n### Answer:", "output": "Catalyzes the conversion of L-arabinose to L-ribulose.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01454", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "TPSB2, TPSAB1, HPGDS", "output": "This profile is most consistent with mast cell, which is commonly found in spleen under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "spleen", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["TPSB2", "TPSAB1", "HPGDS"]}}
{"id": "eval_01455", "category": "Mutation", "instruction": "Describe the effect of mutation Q6XQ97-S22C on protein U-spanin.", "input": "### Wild-type:\nMKLKKTCIAITVAVGVISLSGSSTASALSGLLSDSPDVTAQVGAENTKQLAGVTAKADDKREVKVSDSNIGKIDSSVKKSVEVSTIQANTVNAESITVTKSGSWYDPVVCWILVFIVLLLFYFLIRKHEKKEA\n\n### Mutant:\nMKLKKTCIAITVAVGVISLSGCSTASALSGLLSDSPDVTAQVGAENTKQLAGVTAKADDKREVKVSDSNIGKIDSSVKKSVEVSTIQANTVNAESITVTKSGSWYDPVVCWILVFIVLLLFYFLIRKHEKKEA\n\n### Answer:", "output": "Complete gain of lysis. The gp11 mutants lacking either membrane localization signal were functional and conferred a partially dominant phenotype.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01456", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMKVQYSFEREFEELMSDLLSKYGYEMFQMDGLGDQLDVVKFTEDFVRRGVTNISTYFIEISKPHTYLYSLYRIWQKMKEMFGKGVADEFVEAQINGAVYLHDRHHAALMPYCFAYTLKPIVEKGLPFIKTIKSEPAKHLSTFIQHVIQFVMFASNQSSGAVGLPDFFVWMWYFVKKDLKEGIIPRDKLDWYIEQHFQILTYSLNQPIRTTQSPYTNFTYLDRNYIKAIFEGERYPDGSLITDHVEDIIALQKHYWEWVSRERERQMFTFPVLTASLLYKDGKFLDEDSARFINKINMKWQDTNWYISDSIDAVASCCEKLKGRMNSIGGSDLNIGSFKVITVNLPRIALESGGDREKYLQILRHRVQLIKKALAAVREIIKERISEGLLPLYENGLMLLNRQYGTIGVTGVWESASIMGLTTEDIDGLKYTEEGEVFVDNVLDTIREEAEKGYHEYGFTFNIEQVPAEKAAVTLAQKDRFLFGEKQPFEIYSNQWVPLMANTDVLNRIRYSGKWDKKVSGGAILHINLGESFKTEEESFNMVKMIADMGVMYFAFNTKISVCEDGHAFYGERCPVCGKAKVDEYMRIVGYLVPVSAFNKER\n\n### Answer:", "output": "<SEQ_3Di>DDDDDDDDPVVVVLVVVVCVVQNQVLCVLLLNHCLLPLVVVVVVVVVPDLWDPLVVVVPSCLSVLLVVQLRLLLVLLCVVPNNVLSSVVVVCDSLQQKDWPPSSCLLPAFADAAWECPLCQCQNCVLPPVDGDHHDDALVVVLVSVLVVLVVVVVRHPAEYEYACPLLVNVVRLVNCDVVVVADPVCSVVVLVVSLVVNVVSQADDPGTSGGHAYAYEDEDLVLQCLQQVPDDDPVRDGSNVCVVVSNVSVLSSLVVVLVCQQVFDDFPDQAEYEFAAPDLGTPRVVVVLSVLVSCQRFLRYKYFYDHDCPLNSPQCQAHDFDDARQDDGHYLAAAGTEMEGELLVLLVVCVLPLVSSVVVLLVVLVSLLSSLLSSVVSVVVCVVVVNRSCPVSVTRDPSRHAHEYEYWQLPLSCVLSVFWDADPQWTDGHPVSVVSVVVSLVSSLVVQVVSCVVRVGHYAYALALCLPSQQSSLVSVCVVPPVSRPDRGIASANDHPPTRHHQVVQLLVLLVPCVSRPNRGEDEGDDDDRDPDSVSSVVVVSVSVVSRHRMYTYWDKWFDDPSRRIDDDFFSPPPTGGTDFIWTDPSNTTDTPPVSVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01457", "category": "Mutation", "instruction": "Describe the effect of mutation P01241-R42C on protein Somatotropin.", "input": "### Wild-type:\nMATGSRTSLLLAFGLLCLPWLQEGSAFPTIPLSRLFDNAMLRAHRLHQLAFDTYQEFEEAYIPKEQKYSFLQNPQTSLCFSESIPTPSNREETQQKSNLELLRISLLLIQSWLEPVQFLRSVFANSLVYGASDSNVYDLLKDLEEGIQTLMGRLEDGSPRTGQIFKQTYSKFDTNSHNDDALLKNYGLLYCFRKDMDKVETFLRIVQCRSVEGSCGF\n\n### Mutant:\nMATGSRTSLLLAFGLLCLPWLQEGSAFPTIPLSRLFDNAMLCAHRLHQLAFDTYQEFEEAYIPKEQKYSFLQNPQTSLCFSESIPTPSNREETQQKSNLELLRISLLLIQSWLEPVQFLRSVFANSLVYGASDSNVYDLLKDLEEGIQTLMGRLEDGSPRTGQIFKQTYSKFDTNSHNDDALLKNYGLLYCFRKDMDKVETFLRIVQCRSVEGSCGF\n\n### Answer:", "output": "In IGHD1B; reduced secretion.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01458", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nFRERLKELVVPKHVMDVVDEELSKLGLLDNHSSEFNVTRNYLDWLTSIPWGKYSNENLDLARAQAVLEEDHYGMEDVKKRILEFIAVSQLRGSTQGKILCFYGPPGVGKTSIARSIARALNREYFRFSVGGMTDVAEIKGHRRTYVGAMPGKIIQCLKKTKTENPLILIDEVDKIGRDPSSALLELLDPEQNANFLDHYLDVPVDLSKVLFICTANVTDTIPEPLRDRMEMINVSGYVAQEKLAIAERYLVPQARALCGLDESKAKLSSDVLTLLIKQYCRESGVRNLQKQVEKVLRKSAYKIVSGEAESVEVTPENLQDFVGKPVFTVERMYDVTPPGVVMGLAWTAMGGSTLFVETSLRRPQDKDGSLEVTGQLGEVMKESARIAYTFARAFLMQHAPANDYLVTSHIHLHVPEGATPKDGPSAGCTIVTALLSLAMGRPVRQNLAMTGEVSLTGKILPVGGIKEKTIAAKRAGVTCIVLPAENKKDFYDLAAFITEGLEVHFVEHYREIFDIAF\n\n### Answer:", "output": "<SEQ_3Di>DVVLPPVADADVVLVVVLVVLVVVLVPDDPPDPVNVLSVVVNVLSSLFRAHDADPWDLDQVQLLVLLPQQAPDPPQVSVVLSVQSVVCVVVVHQFAAAAEEAAAPALCLPVSLVSSCVSRVAAEAEDEQAADADLCQALWDASPDPDGDQHPVSVVSRVRVHQARAYEYEAQLHHHHDNVVLVVLCLDLVNQCFNQRRHNRGGHHNRNYYYYYYHLDDPSHDPSVVPSYDYDYRYFDALVRLLVCCVSPLLCVLLVVLPHDPFQEDEDSVNLSLCVLFVDDFRHCPVVSVLSNQQNNVSSCCCVVPVDVHDYDDPVCVCVRVNDGHPPPLDDDQFDFLQKAKEKDADSRFIAIWMKGKDFPDPDLCWEDEAEAQQEDPLVVVLLVLLLLQLVVVCCVPPVPLCHSTNTHMYMYIPPSPDHDYDQKCSLVSNSSSNSNSVRHAWDRQEYEIFHAGNNFFTHADDDPPRRLSSSVSSPGQEYEGAQVCPVVVVVDDPVSCPRHHYDHDTGCVVVVVRGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01459", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL30_STAHJ\n\n### Answer:", "output": "50S ribosomal protein L30.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01460", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPR20E_HUMAN\n\n### Answer:", "output": "Belongs to the PRR20 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01461", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CC(C)(C)c1cc(CCC(=O)NN)cc(C(C)(C)C)c1O", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, amide, amine, hydroxyl, phenol. Its descriptor profile: MW=292.42, LogP=2.91, TPSA=75.35, HBD=3, HBA=3, QED=0.46.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CC(C)(C)c1cc(CCC(=O)NN)cc(C(C)(C)C)c1O", "compound_name": null, "pubchem_cid": null, "MW": 292.42, "LogP": 2.91, "TPSA": 75.35, "QED": 0.46, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": null, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01462", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nNKX6-1:medium INS:medium IAPP:high\n</cell>", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as INS, IAPP, NKX6-1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "tumor", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["NKX6-1", "INS", "IAPP"]}}
{"id": "eval_01463", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMHKRAIYPGTFDPVTNGHADLIERAAKLFQHVVIGIAANPSKQPRFTLDERVQLLKLVTAHLDNVEVVGFSGLLVDFAKDQQASVLVRGLRAVSDFEYEFQLANMNRRLDPDLESVFLTPSEENSFISSTLVKEVALHGGDVSQFVHPEVSKALLSKG\n\n### Sequence 2:\nMHKRAIYPGTFDPVTNGHADLIERAANLFEHVIIGIAANPSKQPRFTLAERVELLKTVTAHLDNVEVVGFSGLLVDFAKEQNASVLVRGLRAVSDFEYEFQLANMNRRLSPDLESVFLTPAEENSFISSTLVKEVALHGGDVSQFVHAEVANALTKKA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01464", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CN1C(=O)C2CN(c3ncccc3C#N)CC2(c2cc(-c3cccc(C#N)c3)cs2)N=C1N", "output": "This molecule is medium-sized and polar, with moderately lipophilic character. It contains aromatic ring, amide, amine, nitrile, pyridine. Its descriptor profile: MW=453.53, LogP=2.67, TPSA=122.4, HBD=1, HBA=8, QED=0.65.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BACE", "canonical_smiles": "CN1C(=O)C2CN(c3ncccc3C#N)CC2(c2cc(-c3cccc(C#N)c3)cs2)N=C1N", "compound_name": null, "pubchem_cid": null, "MW": 453.53, "LogP": 2.67, "TPSA": 122.4, "QED": 0.65, "functional_groups": ["aromatic ring", "amide", "amine", "nitrile", "pyridine"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01465", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "COL1A2, LUM, COL3A1, DCN, COL1A1, PDGFRA", "output": "This profile is most consistent with fibroblast, which is commonly found in colon under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "colon", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["COL1A2", "LUM", "COL3A1", "DCN", "COL1A1", "PDGFRA"]}}
{"id": "eval_01466", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAVELKKVFSTILRFLALAATVVAVIVMIRSHDSAIVLNLTFSAKYNNTPAFKYFVIAEGIASVYTIIVIFLWSKGLLGRLIVILDMVTTVLLTSSISAALAIAQVGKKGNSHAGWLPVCGQVPKFCDQAIIALVAGFVAAIVYFMLLLCSLHAVLTPIFAVKP\n\n### Sequence 2:\nMRNKRVIPNRTFRVSDQIQRDVAELIRDLKDPRIGMVTINSVDITPDYAHAKVYFSVLIGDAADSELALNEAAGFLRNGLFKRLQIHTVPTLHFHFDRTTERAAELSALITQANARRALDSDQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01467", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCCCCCCCC(C)=O", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains ketone. Its descriptor profile: MW=156.27, LogP=3.33, TPSA=17.07, HBD=0, HBA=1, QED=0.52.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "ESOL", "canonical_smiles": "CCCCCCCCC(C)=O", "compound_name": null, "pubchem_cid": null, "MW": 156.27, "LogP": 3.33, "TPSA": 17.07, "QED": 0.52, "functional_groups": ["ketone"], "moleculenet_labels": {"label": -3.3}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01468", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTNHSVILCILDGWGNGENNQFNAIAQAYKPCWDNITLQYPKSHLITHGSSVGLPDGQIGNSEVGHVNLGSGRIVLQDLCKINNEIKYIKDNVYLLEFINKIKKNNGICHIAGLLSDGGVHSSYTHILDIINTLSDFQIQVAVHIFLDGRDTPPISATKYISILCEHIKHLNNINIATLSGRYYAMDRDNRLDRTTKAYNAIAFANAKRYEDPLTAVQDSYNLGITDEFILPCIIGNYQGMKPQDGFIMTNFRSDRVIQIIKMIIGDIKTDNHITLDNTIGMVKYSDKINIPCLFPNNNITNTLGEVIANNQLHQLRIAETEKYAHVTFFFNGGKEDMFENEDRIIIPSPSVATYNLTPEMSAEKVTDTIIEKIKLQKYSLIIVNYANADMVGHTGNIDATKQAITTIDQCLSKILNCIQNTNYVLVITSDHGNAEEMFDVKNNMPYTAHTLNPVPFIICNYNKKIRLKNGRLCDVAPTILEILNLAKPKEMTGVSLIEQV\n\n### Sequence 2:\nMTNHSVILCILDGWGNGENNQFNAIAQAYKPCWDNITLQYPKSHLITHGSSVGLPDGQIGNSEVGHVNLGSGRIVLQDLCKINNEIKYIKDNVYLLEFINKIKKNNGICHIAGLLSDGGVHSSYTHILDIINTLSDFQIQVAVHIFLDGRDTPPISATKYISILCEHIKHLNNINIATLSGRYYAMDRDNRLDRTTKAYNAIAFANAKRYEDPLTAVQDSYNLGITDEFILPCIIGNYQGMKPQDGFIMTNFRSDRVIQIIKMIIGDIKTDNHITLDNTIGMVKYSDKINIPCLFPNNNITNTLGEVIANNQLHQLRIAETEKYAHVTFFFNGGKEDMFENEDRIIIPSPSVATYNLTPEMSAEKVTDTIIEKIKLQKYSLIIVNYANADMVGHTGNIDATKQAITTIDQCLSKILNCIQNTNYVLVITSDHGNAEEMFDVKNNMPYTAHTLNPVPFIICNYNKKIRLKNGRLCDVAPTILEILNLAKPKEMTGVSLIEQV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01469", "category": "Mutation", "instruction": "Describe the effect of mutation P00966-N237I on protein Argininosuccinate synthase.", "input": "### Wild-type:\nMSSKGSVVLAYSGGLDTSCILVWLKEQGYDVIAYLANIGQKEDFEEARKKALKLGAKKVFIEDVSREFVEEFIWPAIQSSALYEDRYLLGTSLARPCIARKQVEIAQREGAKYVSHGATGKGNDQVRFELSCYSLAPQIKVIAPWRMPEFYNRFKGRNDLMEYAKQHGIPIPVTPKNPWSMDENLMHISYEAGILENPKNQAPPGLYTKTQDPAKAPNTPDILEIEFKKGVPVKVTNVKDGTTHQTSLELFMYLNEVAGKHGVGRIDIVENRFIGMKSRGIYETPAGTILYHAHLDIEAF\n\n### Mutant:\nMSSKGSVVLAYSGGLDTSCILVWLKEQGYDVIAYLANIGQKEDFEEARKKALKLGAKKVFIEDVSREFVEEFIWPAIQSSALYEDRYLLGTSLARPCIARKQVEIAQREGAKYVSHGATGKGNDQVRFELSCYSLAPQIKVIAPWRMPEFYNRFKGRNDLMEYAKQHGIPIPVTPKNPWSMDENLMHISYEAGILENPKNQAPPGLYTKTQDPAKAPNTPDILEIEFKKGVPVKVTIVKDGTTHQTSLELFMYLNEVAGKHGVGRIDIVENRFIGMKSRGIYETPAGTILYHAHLDIEAF\n\n### Answer:", "output": "In CTLN1. The mutation in the ASS1 gene leads to an impairment of the argininosuccinate synthetase function.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01470", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nACTC_HALRO\n\n### Answer:", "output": "Belongs to the actin family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01471", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nDLWQWGQMILKETGKLPFSYYTAYGCYCGWGGRGGKPKADTDRCCFVHDC\n\n### Sequence 2:\nDLWQFGKMILKVAGKLPFPYYGAYGCYCGWGGR-GKPKDPTDRCCFVHDC\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01472", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPCKA_TRYCR\n\n### Answer:", "output": "Belongs to the phosphoenolpyruvate carboxykinase (ATP) family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01473", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMTMTDPIADMLSRVRNASNAFHDSVTMPSSKLKAHIAEILKQEGYIEDFAVNDRKDGKAGKELEITLKYGPTRERALAGVRRVSKPGLRVYTKSTNLPKVLGGLGVAIISTSHGLLTDREASNKGVGGEVLAYVW\n\n### Sequence 2:\nMTMTDPIADMLSRVRNANNAHHDAVSMPSSKLKANIAEILKSEGYIADYKV---EDAKVGKTLTLDLKYGPNRQRSIEGVRRVSKPGLRVYAKSTNLPQVLGGLGVAIISTSHGLLTDRQATEKGVGGEVLAYVW\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01474", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCYSNC_MYCTO\n\n### Answer:", "output": "In the N-terminal section; belongs to the TRAFAC class translation factor GTPase superfamily. Classic translation factor GTPase family. CysN/NodQ subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01475", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMCPRPRRAVLLGLGVAMSAIAGCRETAPSTTQTSDGGPEPTGRSGDTQTDGGGETTRQATVAEGGESATRADETETSELDLREANVVDVTLQSAGGRAVEFSVTLYHDDDGEDGYADWWQVETLAGDRLGRRELLHAHSTAPFTRSETIEVPEGTTCVVVRGHDQTHGYGGQAMVVDTESGATRAVQQGPEPAEFDDGDCP\n\n### Sequence 2:\nMAEKTQKSVKIAPGAVVCVESEIRGDVTIGPRTVIHPKARIIAEAGPIVIGEGNLIEEQALIINAHPDNITPDAEDPEPKPMIIGTNNVFEVGCYCQAMKIGDNNVIESKAYVGRNVILTSGCIIGACCNLNTFEVIPENTVIYGGDCLRRVQTERPQPQTLQLDFLMKILPNYHHLKKTMKGSSTPVKN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01476", "category": "Mutation", "instruction": "Describe the effect of mutation O95786-R788K on protein Antiviral innate immune response receptor RIG-I.", "input": "### Wild-type:\nMTTEQRRSLQAFQDYIRKTLDPTYILSYMAPWFREEEVQYIQAEKNNKGPMEAATLFLKFLLELQEEGWFRGFLDALDHAGYSGLYEAIESWDFKKIEKLEEYRLLLKRLQPEFKTRIIPTDIISDLSECLINQECEEILQICSTKGMMAGAEKLVECLLRSDKENWPKTLKLALEKERNKFSELWIVEKGIKDVETEDLEDKMETSDIQIFYQEDPECQNLSENSCPPSEVSDTNLYSPFKPRNYQLELALPAMKGKNTIICAPTGCGKTFVSLLICEHHLKKFPQGQKGKVVFFANQI\n\n### Mutant:\nMTTEQRRSLQAFQDYIRKTLDPTYILSYMAPWFREEEVQYIQAEKNNKGPMEAATLFLKFLLELQEEGWFRGFLDALDHAGYSGLYEAIESWDFKKIEKLEEYRLLLKRLQPEFKTRIIPTDIISDLSECLINQECEEILQICSTKGMMAGAEKLVECLLRSDKENWPKTLKLALEKERNKFSELWIVEKGIKDVETEDLEDKMETSDIQIFYQEDPECQNLSENSCPPSEVSDTNLYSPFKPRNYQLELALPAMKGKNTIICAPTGCGKTFVSLLICEHHLKKFPQGQKGKVVFFANQI\n\n### Answer:", "output": "Increased polyubiquitination. Gain of function in RIG-I signaling pathway. Increased ubiquitination and function in RIG-I signaling pathway without effect on RNA-binding; when associated with R-849, R-851, R-888, R-907 and R-909. K788 within the RIG-I repressor domain was critical for Riplet-mediated K63-linked polyubiquitination and that Riplet was required for the release of RIG-I autorepression of its N-terminal CARDs, which leads to the association of RIG-I with TRIM25 ubiquitin ligase and TBK1 protein kinase.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01477", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: COc1cccc(C[NH2+]CC(O)C(Cc2ccccc2)NC(=O)c2cc(C(=O)NC(C)c3ccccc3)cc(N(C)S(C)(=O)=O)c2)c1\nDescriptors: MW=659.83, LogP=3.05, TPSA=141.65, HBD=4, HBA=6", "output": "Descriptors: MW=659.83, LogP=3.05, TPSA=141.65, HBD=4, HBA=6, QED=0.15, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "BACE", "canonical_smiles": "COc1cccc(C[NH2+]CC(O)C(Cc2ccccc2)NC(=O)c2cc(C(=O)NC(C)c3ccccc3)cc(N(C)S(C)(=O)=O)c2)c1", "compound_name": null, "pubchem_cid": null, "MW": 659.83, "LogP": 3.05, "TPSA": 141.65, "QED": 0.15, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether", "sulfonamide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01478", "category": "Mutation", "instruction": "Describe the effect of mutation P14929-A307G on protein Signal recognition particle receptor FtsY.", "input": "### Wild-type:\nMFSFFRRKKKQETPALEEAQVQETAAKVESEVAQIVGNIKEDVESLAESVKGRAESAVETVSGAVEQVKETVAEMPSEAGEAAERVESAKEAVAETVGEAVGQVQEAVATTEEHKLGWAARLKQGLAKSRDKMAKSLAGVFGGGQIGEDLYEELETVLITGDMGMEATEYLMKDVRGRVSLKGLKDGNELRGALKEALYDLIKPLEKPLVLPETKEPFVIMLAGINGAGKTTSIGKLAKYFQAQGKSVLLAAGDTFRAAAREQLQAWGGRNNVTVISQTTGDSAAVCFDAVQAAKARIDI\n\n### Mutant:\nMFSFFRRKKKQETPALEEAQVQETAAKVESEVAQIVGNIKEDVESLAESVKGRAESAVETVSGAVEQVKETVAEMPSEAGEAAERVESAKEAVAETVGEAVGQVQEAVATTEEHKLGWAARLKQGLAKSRDKMAKSLAGVFGGGQIGEDLYEELETVLITGDMGMEATEYLMKDVRGRVSLKGLKDGNELRGALKEALYDLIKPLEKPLVLPETKEPFVIMLAGINGAGKTTSIGKLAKYFQAQGKSVLLAAGDTFRAAAREQLQAWGGRNNVTVISQTTGDSAAVCFDAVQAAKARIDI\n\n### Answer:", "output": "Increase in GTPase activity; able to complement the lack of FtsY in E.coli; translocation activity. The mutation in the PilA gene of Neisseria gonorrhoeae enhances the GTPase activity of the protein. Overexpression of the mutant prevents the accumulation of pre-beta-lactamase, suggesting that the mutant protein may not interact with the SRP apparatus to affect protein maturation. The GTPase activity is unnecessary for the function of PilA.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01479", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRK14_PHATC\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uL14 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01480", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "TTR, HNF4A, APOB, ALB", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as TTR, ALB, HNF4A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "blood", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TTR", "HNF4A", "APOB", "ALB"]}}
{"id": "eval_01481", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "C#CCN(N=O)C(=O)NC(CCCCNC(=O)OCc1ccccc1)C(=O)OCc1ccccc1", "output": "The molecule contains the following functional groups: aromatic ring, ester, amide, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "HIV", "canonical_smiles": "C#CCN(N=O)C(=O)NC(CCCCNC(=O)OCc1ccccc1)C(=O)OCc1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 480.52, "LogP": 3.52, "TPSA": 126.4, "QED": 0.15, "functional_groups": ["aromatic ring", "ester", "amide", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "ester", "amide", "ether"]}}
{"id": "eval_01482", "category": "Mutation", "instruction": "Describe the effect of mutation Q8I4S1-S108A on protein Thymidylate kinase.", "input": "### Wild-type:\nMTDDKKKGKFIVFEGLDRSGKSTQSKLLVEYLKNNNVEVKHLYFPNRETGIGQIISKYLKMENSMSNETIHLLFSANRWEHMNEIKSLLLKGIWVVCDRYAYSGVAYSSGALNLNKTWCMNPDQGLIKPDVVFYLNVPPNYAQNRSDYGEEIYEKVETQKKIYETYKHFAHEDYWINIDATRKIEDIHNDIVKEVTKIKVEPEEFNFLWS\n\n### Mutant:\nMTDDKKKGKFIVFEGLDRSGKSTQSKLLVEYLKNNNVEVKHLYFPNRETGIGQIISKYLKMENSMSNETIHLLFSANRWEHMNEIKSLLLKGIWVVCDRYAYSGVAYASGALNLNKTWCMNPDQGLIKPDVVFYLNVPPNYAQNRSDYGEEIYEKVETQKKIYETYKHFAHEDYWINIDATRKIEDIHNDIVKEVTKIKVEPEEFNFLWS\n\n### Answer:", "output": "No defect in thymidylate kinase activity. 1.3-fold reduction in affinity for dGMP. The mutation in PfTMK contributes to the unique ability of PfTMK to utilize dGMP as a substrate.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01483", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMTAFKQGDPTTLNRLYGRSIGKPLRTHQQHLVDNLLPQISPPAEGPVTAERLFGEDCPLHFEIGFGGGEHLVYRADLLPNHGFIGAEPFLNGVASCLSQVEEQRLANIRIHNGDALEVLDRVPDGALTMLYLLHPDPWPKNKHAKRRMMNDGPVKMFADKLKPGGEFRFGTDHPVYLRHALMVMRRHTDAFEWVVEGPSDWQNRPSGWPETRYEHKARTKFGHEVWYFRFRRR\n\n### Sequence 2:\nMTAFKQGDPTTLNRLYGRSIGKPLRTHQQHLVDNLLPQISPPAEGPVTAERLFGEDCPLHFEIGFGGGEHLVYRADLLPNHGFIGAEPFLNGVASCLSQVEEQRLANIRIHNGDALEVLDRVPDGALTMLYLLHPDPWPKNKHAKRRMMNDGPVKMFADKLKPGGEFRFGTDHPVYLRHALMVMRRHTDAFEWVVEGPSDWQNRPSGWPETRYEHKARTKFGHEVWYFRFRRR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01484", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMVMFMVFRILGRMTKIEKEIKEEEAKYDLIIKNEAKIEPIVAEEDMEFKQGDIKPIRIKKIKIPPMSVLLICPYGRHRVGHVVAVGEEVPMPIDVEREVDMAMFACGFEGEVKKGDLIGMLLILAAEKRE\n\n### Sequence 2:\nMSLVYMNIMTAFTVSLVGLLMYRSHLMSSLLCLEGMMLSLFVMATLTILNSHFTLASMMPIILLVFAACEAALGLSLLVMVSNTYGTDYVQNLNLLQC\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01485", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nCD79B:medium CD79A:high TCL1A:medium\n</cell>", "output": "This cell is most consistent with naive B cell, a naive B cell. Key markers such as CD79A, TCL1A, CD79B support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive B cell", "tissue": "ovary", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD79B", "CD79A", "TCL1A"]}}
{"id": "eval_01486", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMENKLSFEEAISQLEHLVSKLEQGDVPLEEAISYFKEGMELSKLCDEKLKNVQEQMAVILGEDGELEPFTAVGDEA\n\n### Sequence 2:\nMKPSGITFAFLVVFMMAIMYNSVQAAAIADADADAEAKAFADAFAEAGWGSIFKTVGKMIAKAAVKAAPEAISAMASQNEK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01487", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCOC(=O)C1CC2OC2CC1C", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains ester, ether. Its descriptor profile: MW=184.23, LogP=1.36, TPSA=38.83, HBD=0, HBA=3, QED=0.48.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "CCOC(=O)C1CC2OC2CC1C", "compound_name": null, "pubchem_cid": null, "MW": 184.23, "LogP": 1.36, "TPSA": 38.83, "QED": 0.48, "functional_groups": ["ester", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01488", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CC(C)C1COC2CCC(=O)N21\nDescriptors: MW=169.22, LogP=0.99, TPSA=29.54, HBD=0, HBA=2", "output": "Descriptors: MW=169.22, LogP=0.99, TPSA=29.54, HBD=0, HBA=2, QED=0.59, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "CC(C)C1COC2CCC(=O)N21", "compound_name": null, "pubchem_cid": null, "MW": 169.22, "LogP": 0.99, "TPSA": 29.54, "QED": 0.59, "functional_groups": ["amide", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01489", "category": "Mutation", "instruction": "Describe the effect of mutation Q96K58-R556Q on protein Zinc finger protein 668.", "input": "### Wild-type:\nMEVEAAEARSPAPGYKRSGRRYKCLSCTKTFPNAPRAARHAATHGPADCSEEVAEVKPKPETEAKAEEASGEKVSGSAAKPRPYACPLCPKAYKTAPELRSHGRSHTGEKPFPCPECGRRFMQPVCLRVHLASHAGELPFRCAHCPKAYGALSKLKIHQRGHTGERPYACADCGKSFADPSVFRKHRRTHAGLRPYSCERCGKAYAELKDLRNHERSHTGERPFLCSECGKSFSRSSSLTCHQRIHAAQKPYRCPACGKGFTQLSSYQSHERTHSGEKPFLCPRCGRMFSDPSSFRRHQR\n\n### Mutant:\nMEVEAAEARSPAPGYKRSGRRYKCLSCTKTFPNAPRAARHAATHGPADCSEEVAEVKPKPETEAKAEEASGEKVSGSAAKPRPYACPLCPKAYKTAPELRSHGRSHTGEKPFPCPECGRRFMQPVCLRVHLASHAGELPFRCAHCPKAYGALSKLKIHQRGHTGERPYACADCGKSFADPSVFRKHRRTHAGLRPYSCERCGKAYAELKDLRNHERSHTGERPFLCSECGKSFSRSSSLTCHQRIHAAQKPYRCPACGKGFTQLSSYQSHERTHSGEKPFLCPRCGRMFSDPSSFRRHQR\n\n### Answer:", "output": "In a breast cancer sample; somatic mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01490", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nUNG_YEAST\n\n### Answer:", "output": "Uracil-DNA glycosylase, UDG.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01491", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQEDVLAKELEDVNKWGLHVFRIAELSGNRPLTVIMHTIFQERDLLKTFKIPVDTLITYLMTLEDHYHADVAYHNNIHAADVVQSTHVLLSTPALEAVFTDLEILAAIFASAIHDVDHPGVSNQFLINTNSELALMYNDSSVLENHHLAVGFKLLQEENCDIFQNLTKKQRQSLRKMVIDIVLATDMSKHMNLLADLKTMVETKKVTVLLLDNYSDRIQVLQNMVHCADLSNPTKPLQLYRQWTDRIMEEFFRQGDRERERGMEISPMCDKHNASVEKSQVGFIDYIVHPLWETWADLVHPDAQDILDTLEDNREWYQSTIPGNQVSEFISNTFLD\n\n### Answer:", "output": "<SEQ_3Di>DVVVLVVLCVCLLPFFRPLVVNCVVVVNLLLLVSVVSLCVVVVVCVVQVFDPVLSSVQSVVQQVQFDPPQLFRHQSLLSQLLSQLSNVCPFPLCPPVDDPLLVVLLNLLSNQLQGNPPLDDLVVCVVVVHPQCVVVVSQLSRRVVSLVVSVVSCPPPSHVRNPRPDPVSVVVSSVLSSLQSSLLQLQCQVVLLVVLLVCLVPWDGVSTDDDDPVVSSSVSSLSSNLSSVLLLLGDLVRSVSSLVRSLVSLLVVQVVCVVVVHDRAFSNHPVDHFSLVLVLLCCVPGNVSSVVSVCSNSPPRCVSSVVSSVVSSVVSVVPRCDVRSVVVCVVGNHD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01492", "category": "Mutation", "instruction": "Describe the effect of mutation Q12158-K252R on protein Sister chromatid cohesion protein 1.", "input": "### Wild-type:\nMVTENPQRLTVLRLATNKGPLAQIWLASNMSNIPRGSVIQTHIAESAKEIAKASGCDDESGDNEYITLRTSGELLQGIVRVYSKQATFLLTDIKDTLTKISMLFKTSQKMTSTVNRLNTVTRVHQLMLEDAVTEREVLVTPGLEFLDDTTIPVGLMAQENSMERKVQGAAPWDTSLEVGRRFSPDEDFEHNNLSSMNLDFDIEEGPITSKSWEEGTRQSSRNFDTHENYIQDDDFPLDDAGTIGWDLGITEKNDQNNDDDDNSVEQGRRLGESIMSEEPTDFGFDLDIEKEAPAGNIDTI\n\n### Mutant:\nMVTENPQRLTVLRLATNKGPLAQIWLASNMSNIPRGSVIQTHIAESAKEIAKASGCDDESGDNEYITLRTSGELLQGIVRVYSKQATFLLTDIKDTLTKISMLFKTSQKMTSTVNRLNTVTRVHQLMLEDAVTEREVLVTPGLEFLDDTTIPVGLMAQENSMERKVQGAAPWDTSLEVGRRFSPDEDFEHNNLSSMNLDFDIEEGPITSKSWEEGTRQSSRNFDTHENYIQDDDFPLDDAGTIGWDLGITERNDQNNDDDDNSVEQGRRLGESIMSEEPTDFGFDLDIEKEAPAGNIDTI\n\n### Answer:", "output": "No effect on acetylation by ECO1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01493", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSAAMTAEVCEDHTKPMLMLLDGNSLAFRAFYALPTENFKTRGGLTTNAVYGFTAMLINLLREEAPTHIAAAFDVSRKTFRSECYAGYKANRSSIPAEFHGQIDITKEVLGALGITVFAEAGFEADDLIATLATQAENEGYRVLVVTGDRDALQLVSNDVTVLYPRKGVSELTRFTPEAVIEKYGVTPAQYPDLAALRGDPSDNLPGIPGVGEKTAAKWIVDYGSLQGLVDNVESVRGKVGEALRTHLASVVRNRELTELVKDVPLVQTSDTLRLQPWDRDRIHRLFDNLEFRVLRDRLFEALAAAGERVPEVDEGFDVRGGLLESGTVGRWLAKHADDGRRSGLAIVGTHLPHGGDATALAVAAADGNGGYIDTAMLTPDDDDALAAWLADPDNPKALHEAKLAMHDLAGRGWTLGGITSDTALAAYLVRPGQRSFTLDDLSLRYLRRELRAETPEQEQFSLLDNVDEVDKQAIQTLILRARAVVDLAAALDAELDLIDSTSLLGEMELPVQQVLADMEKAGIAADLRLLTELQSQFGDQIRDAAEAAYAVIGKQINLSSPKQLQVVLFEELGMPKTKRTKTGYTTDADALQSLFCKTEHPFLQHLLTHRDVTRLKVTVDGLLNAVAADGRIHTTFNQTIATTGRLSSTEPNLQNIPIRTNAGRQIRDAFVVGSENNGYTELMTADYSQIEMRIMAHLSRDEGLIEAFHTGEDLHSFVASRAFGIPIEDITPELRRRVKAMSYGLAYGLSAYGLATQLKISTEEAKLQMEQYFARFGGVRDYLMDVVEQARKDGYTSTVLGRRRYLPELDSSNRQIREAAERAALNAPIQGSAADIIKVAMIAVDKSLKQAKLASRMLLQVHDELLFEVAIGEREQIEAMVREQMGSAYPLDVPLEVSVGFGRSWGAAAH\n\n### Sequence 2:\nLVLVDGNSLAYRAFFALPL--LSNDKGVHTNAVYGFAMILMKMLEDEKPTHMLVAFDAGKTTFRHGTFKEYKGGRQKTPPELSEQMPFIRELLDAYQISRYELEQYEADDIIGTLAKSAEKDGFEVKVFSGDKDLTQLATDKTTVAITRKGITDVEFYTPEHVKEKYGLTPEQIIDMKGLMGDSSDNIPGVPGVGEKTAIKLLKQFDSVEKLLESIDEVSGKKLKEKLEEFKDQALMSKELATIMTDAPIEVSVSGLEYQGFNREQVIAIFKDLGFNTLLERLGEDSAEA-----EQDQS-------LEDINV-KTVTDVTSDILVSPSAFVVEQIGDNYHEEPILGFSIVNETGAYFIPKDIAV----ESEVFKEWVENDEQKKWVFDSKRAVVALRWQGIELKGAEFDTLLAAYIINPGN---SYDDVASVAKDYGLHIVSSDESVYG------KGAKRAVPSEDVLSEHLGRKALAIQSLREKLVQELENNDQLELFEELEMPLALILGEMESTGVKVDVDRLKRMGEELGAKLKEYEEKIHEIAGEPFNINSPKQLGVILFEKIGLPVVKKTKTGYSTSADVLEKLADK--HDIVDYILQYRQIGKLQSTYIEGLLKVTRPDSHKVHTRFNQALTQTGRLSSTDPNLQNIPIRLEEGRKIRQAFV-PSEKDWL--IFAADYSQIELRVLAHISKDENLIEAFTNDMDIHTKTAMDVFHVAKDEVTSAMRRQAKAVNFGIVYGISDYGLSQNLGITRKEAGAFIDRYLESFQGVKAYMEDSVQEAKQKGYVTTLMHRRRYIPELTSRNFNIRSFAERTAMNTPIQGSAADIIKKAMIDMAAKLKEKQLKARLLLQVHDELIFEAPKEEIEILEKLVPEVMEHALALDVPLKVDFASGPSW\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01494", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nEPCAM KRT18 CDH1 KRT8\n</cell>", "output": "This cell is most consistent with epithelial cell, an epithelial cell. Key markers such as EPCAM, CDH1, KRT18 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "epithelial cell", "tissue": "heart", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["EPCAM", "KRT18", "CDH1", "KRT8"]}}
{"id": "eval_01495", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: COC1=C(OC)C2C(=O)OCC2C=C1\nDescriptors: MW=196.2, LogP=0.85, TPSA=44.76, HBD=0, HBA=4", "output": "Descriptors: MW=196.2, LogP=0.85, TPSA=44.76, HBD=0, HBA=4, QED=0.61, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "ESOL", "canonical_smiles": "COC1=C(OC)C2C(=O)OCC2C=C1", "compound_name": null, "pubchem_cid": null, "MW": 196.2, "LogP": 0.85, "TPSA": 44.76, "QED": 0.61, "functional_groups": ["ester", "ether"], "moleculenet_labels": {"label": -1.899}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01496", "category": "Mutation", "instruction": "Describe the effect of mutation Q9VFS8-R10A on protein 26S proteasome non-ATPase regulatory subunit 9.", "input": "### Wild-type:\nMAAGTTTKERLERLINAKKQLEAQINRNGQILAANDNVGMSGPLVDAEGFPRNDIDVYQVRLARQTIICLQNDHKELMNQIQTLLNQYHSEIATTDPELVNRASALDLDSDRSPGGANITDLAPARAIVVVNLVSPDSPAERAGLCAGDAILRFGSINSGNFKGDLAQIGELVRNMQSQNVQLKVKRGEQQLDLILVPKTWSGRGLLGCNIVLPPEAMDH\n\n### Mutant:\nMAAGTTTKEALERLINAKKQLEAQINRNGQILAANDNVGMSGPLVDAEGFPRNDIDVYQVRLARQTIICLQNDHKELMNQIQTLLNQYHSEIATTDPELVNRASALDLDSDRSPGGANITDLAPARAIVVVNLVSPDSPAERAGLCAGDAILRFGSINSGNFKGDLAQIGELVRNMQSQNVQLKVKRGEQQLDLILVPKTWSGRGLLGCNIVLPPEAMDH\n\n### Answer:", "output": "Does not affect interaction with PI31.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01497", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "CD8A, LEF1, CCR7, CD3D, CD3E, CD8B", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as CD8B, CD8A, CCR7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "skin", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD8A", "LEF1", "CCR7", "CD3D", "CD3E", "CD8B"]}}
{"id": "eval_01498", "category": "Mutation", "instruction": "Describe the effect of mutation A8B1Z2-G238A on protein Di-trans,poly-cis-undecaprenyl-diphosphate synthase.", "input": "### Wild-type:\nMIPMHVAVIMDGNGRWARKQLQERTFGHEQGVSVLESIVDECINCGIRFLTVYAFSTENWSRPPTEVSFLFELLSAAIQRVRTTYRERNVKVQFCGERTTQIPETVIAAMNCIEQDTAACTGLILSVCFNYGGHTEIAQACRSVLADCLEGDAVENIKTRLQMPIEQFIQQIDTHLYANLPPVDLLIRTGCEKRLSNFLPWHLAYAEIIFSDLLWPEFSVRAFKDCLDEFASRTRRFGGVQLSPMTGVYSDTHPHSSTNALSNHD\n\n### Mutant:\nMIPMHVAVIMDGNGRWARKQLQERTFGHEQGVSVLESIVDECINCGIRFLTVYAFSTENWSRPPTEVSFLFELLSAAIQRVRTTYRERNVKVQFCGERTTQIPETVIAAMNCIEQDTAACTGLILSVCFNYGGHTEIAQACRSVLADCLEGDAVENIKTRLQMPIEQFIQQIDTHLYANLPPVDLLIRTGCEKRLSNFLPWHLAYAEIIFSDLLWPEFSVRAFKDCLDEFASRTRRFAGVQLSPMTGVYSDTHPHSSTNALSNHD\n\n### Answer:", "output": "Loss of catalytic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01499", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEQADEDLLSQYENGEGLFSQKSFLVLGFSNENESNIANIIKENAGKIMSLLSRTVADYAVVPLLGCEVEATVGEVVTNTWLVTCIDYQTLFDPKSNPLFTPVPVMTGMTPLEDCVISFSQCAGAEKESLTFLANLLGASVQEYFVRKSNAKKGMFASTHLILKERGGSKYEAAKKWNLPAVTIAWLLETARTGKRADESHFLIENS\n\n### Answer:", "output": "<SEQ_3Di>DVVVVVVVVVPDDDVAQQAPFEEEEDDDDPVVSVVVVVVCVVRNYDYADPVGPDEGQEYEAAQQGDDDPGDYNYYFYVLQVVVCVVVVHNDDCVVDPSGHHQDFDPPFQQCLVFEEEEPPDDDSRVVVVCVLNVRSNYDYDPDAAQDDDVVVRDHHHQEYEYAAPDDDVNVVCVVVVHAYFYCVQVSVCRNVSHRDDSVVTGNVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01500", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: O=S1(=O)OC(c2ccc([O-])cc2)(c2ccc(O)cc2)c2ccccc21\nDescriptors: MW=353.38, LogP=2.48, TPSA=86.66, HBD=1, HBA=5", "output": "Descriptors: MW=353.38, LogP=2.48, TPSA=86.66, HBD=1, HBA=5, QED=0.72, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "O=S1(=O)OC(c2ccc([O-])cc2)(c2ccc(O)cc2)c2ccccc21", "compound_name": null, "pubchem_cid": null, "MW": 353.38, "LogP": 2.48, "TPSA": 86.66, "QED": 0.72, "functional_groups": ["aromatic ring", "hydroxyl", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01501", "category": "Mutation", "instruction": "Describe the effect of mutation Q14684-V684A on protein Ribosomal RNA processing protein 1 homolog B.", "input": "### Wild-type:\nMAPAMQPAEIQFAQRLASSEKGIRDRAVKKLRQYISVKTQRETGGFSQEELLKIWKGLFYCMWVQDEPLLQEELANTIAQLVHAVNNSAAQHLFIQTFWQTMNREWKGIDRLRLDKYYMLIRLVLRQSFEVLKRNGWEESRIKVFLDVLMKEVLCPESQSPNGVRFHFIDIYLDELSKVGGKELLADQNLKFIDPFCKIAAKTKDHTLVQTIARGVFEAIVDQSPFVPEETMEEQKTKVGDGDLSAEEIPENEVSLRRAVSKKKTALGKNHSRKDGLSDERGRDDCGTFEDTGPLLQFDY\n\n### Mutant:\nMAPAMQPAEIQFAQRLASSEKGIRDRAVKKLRQYISVKTQRETGGFSQEELLKIWKGLFYCMWVQDEPLLQEELANTIAQLVHAVNNSAAQHLFIQTFWQTMNREWKGIDRLRLDKYYMLIRLVLRQSFEVLKRNGWEESRIKVFLDVLMKEVLCPESQSPNGVRFHFIDIYLDELSKVGGKELLADQNLKFIDPFCKIAAKTKDHTLVQTIARGVFEAIVDQSPFVPEETMEEQKTKVGDGDLSAEEIPENEVSLRRAVSKKKTALGKNHSRKDGLSDERGRDDCGTFEDTGPLLQFDY\n\n### Answer:", "output": "Abolishes interaction with protein phosphatase PP1 subunits PPP1CB and PPP1CC; when associated with A-686.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01502", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMWVRGSGPSVLSRLQDAAVVRPGFLSTAEEETLSRELEPELRRRRYEYDHWDAAIHGFRETEKSRWSEASRAILRRVQAAAFGTLLSSVHVXDLEARGYIKPHVDSIKFCGATIAGLSLLSPSVMRLVHTQEPGEWLELLLEPGSLYILRGSARYDFSHEILRDEESFFGERRIPRGRRISVICRSLP\n\n### Answer:", "output": "<SEQ_3Di>DFKDWLDVVLCVVCVVQKDKDPQLDDPVLLVQVVVVCPVVQVVDDQDDDDPPAFKFQKDKDWAQDDDPSNVVVVVSQCVVAHCLADNTKIKMKHALQIKGHWDAPDPQFFDFKKKKAKADFKWKKKKAFPVHRSTIMIGIHGHNMMIMHGDCSRPGITIMIDHQVGTDDVPHHGRGGIMMMIMGGHGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01503", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGAQLSLVVQASPSIAIFSYIDVLEEVHYVSQLNSSRFLKTCKALDPNGEIVIKVFIKPKDQYSLRPFLQRIRAQSFKLGQLPHVLNYSKLIETNRAGYMIRQHLKNNLYDRLSLRPYLQDIELKFIAFQLLNALKDIHNLNIVHGDIKTENILVTSWNWCILTDFAAFIKPVYLPEDNPGEFLFYFDTSKRRTCYLAPERFNSKLYQDGKSNNGRLTKEMDIFSLGCVIAEIFAEGRPIFNLSQLFKYKSNSYDVNREFLMEEMNSTDLRNLVLDMIQLDPSKRLSCDELLNKYRGIFFPDYFYTFIYDYFRNLVTMTTSTPISDNTCTNSTLEDNVKLLDETTEKIYRDFSQICHCLDFPLIKDGGEIGSDPPILESYKIEIEISRFLNTNLYFPQNYHLVLQQFTKVSEKIKSVKEECALLFISYLSHSIRSIVSTATKLKNLELLAVFAQFVSDENKIDRVVPYFVCCFEDSDQDVQALSLLTLIQVLTSVRKLNQLNENIFVDYLLPRLKRLLISNRQNTNYLRIVFANCLSDLAIIINRFQEFTFAQHCNDNSMDNNTEIMESSTKYSAKLIQSVEDLTVSFLTDNDTYVKMALLQNILPLCKFFGRERTNDIILSHLITYLNDKDPALRVSLIQTISGISILLGTVTLEQYILPLLIQTITDSEELVVISVLQSLKSLFKTGLIRKKYYIDISKTTSPLLLHPNNWIRQFTLMIIIEIINKLSKAEVYCILYPIIRPFFEFDVEFNFKSMISCCKQPVSRSVYNLLCSWSVRASKSLFWKKIITNHVDSFGNNRIEFITKNYSSKNYGFNKRDTKSSSSLKGIKTSSTVYSHDNKEIPLTAEDRNWIDKFHIIGLTEKDIWKIVALRGYVIRTARVMAANPDFPYNNSNYRPLVQNSPPNLNLTNIMPRNIFFDVEFAEESTSEGQDSNLENQQIYKYDESEKDSNKLNINGSKQLSTVMDINGSLIFKNKSIATTTSNLKNVFVQLEPTSYHMHSPNHGLKDNANVKPERKVVVSNSYEGDVESIEKFLSTFKILPPLRDYKEFGPIQEIVRSPNMGNLRGKLIATLMENEPNSITSSAVSPGETPYLITGSDQGVIKIWNLKEIIVGEVYSSSLTYDCSSTVTQITMIPNFDAFAVSSKDGQIIVLKVNHYQQESEVKFLNCECIRKINLKNFGKNEYAVRMRAFVNEEKSLLVALTNLSRVIIFDIRTLERLQIIENSPRHGAVSSICIDEECCVLILGTTRGIIDIWDIRFNVLIRSWSFGDHAPITHVEVCQFYGKNSVIVVGGSSKTFLTIWNFVKGHCQYAFINSDEQPSMEHFLPIEKGLEELNFCGIRSLNALSTISVSNDKILLTDEATSSIVMFSLNELSSSKAVISPSRFSDVFIPTQVTANLTMLLRKMKRTSTHSVDDSLYHHDIINSISTCEVDETPLLVACDNSGLIGIFQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCSTTCCSSSSCCSSCCCCCCEECCSSSCCEEEECCCSSSCCCEEECCCSSSGGGCCCCHHHHHHHHHHHHSCCSSCCEEEECCCSSCCCEEECCCCCCHHHHHHHCTTTCCHHHHHHHHHHHHHHHHHHHHTTCCCCCCCTTSCBCCTTCCCBCCSCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHTCSHHHHHCCCCCCHHHHHHHHHHHHHHHHHSSCCCCHHHHHHHHTTCCCCCCCCCCCCCCCHHHHHHHHHHHHCTTTTTCCHHHHHSSCSCSHHHHHHHHHHHHHHHHHHHHCCCCCCCCCCSSCSHHHHHHHHHHHHHHHHHHHHHHHHSSCCSCCSSHHHHHHSSTTTCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHTTCCCCSHHHHHHHHTTSCSSSHHHHHHHHHHHHTTSCCHHHHHHHHHHHHHHHHHCCSCTHHHHHHHHHHHHHHSSCCCCSSTTHHHHHHHHHHHHHHTTCCCTTCCCHHHHHHHHHTHHHHHHHHHHHHHHHHHHTSSTTTTTTTTTTTTCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCCCCCCCCCCCTHHHHHHHHHHHSGGGCCCCTHHHHHHHHHHHHHHHTCHHHHHHHHHHHTTCCCSCTHHHHHHHHHHHTTTSSCHHHHHHHHHHHHHHHHSCCCHHHHHHHHHHHHHGGGCCSHHHHHHHTHHHHHHHCCCCCCCCCHHHHHHHHHHCCCSSCCSHHHHHHHHHTTCCCSHHHHHHHHHHHHHHSSCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHSTTCSHHHHHHHHSSCCCSGGGCCCTHHHHHHHHHHHHTTTSSCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCSCCCSCCCCCCSTHHHHHHHHHHHHTCCCCSSCCCCCCSSTTHHHHHHHTTSCCCTTHHHHHSSSCCCCCCCCCCCTTSCCCEEEEECSCSSSCCSCEEEECSSSCEEEECCTTSEEEEEEHHHHHTTCCCSCSEEEECSSCEEEEEECTTSSCEEEEETTTEEEEEEEEEEEETTEEEEEEEEEEEEEEGGGGSSCCCCCCEEEEECSSCEEEEEECTTCEEEEEETTTTEEEEEEECCGGGCSBCCCEECSSSSEEEEEBTTSCEEEEETTTTEEEEEEBCTTCCCCCEEECCSTTCTTEEEEECSSSSEEEEEEETTTTEEEEEEEEBSSCCCHHHHSCBCSCCTTCCCCCCCSTTSCCCEEEETTEEEEEETTTTEEEEEESSCTTTCEEEECCSSCCCEEEEEEEETTEEEEEEECCCCCCCCCCCCCCCCCCEEEEEEEECSSSEEEEEEETTSCEEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01504", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLNEIKAKTKERMLKTIQSFHDDIKGVRTGRASASLLDGIVVNIYGGHQKLNQVAGVSVIDNKTLSIKVWDISVVGEVKNAILNANLNLNPVVEGSTIRIALPDLTQETREKLVKLLHQFAENARIAIRNIRRDIMEETEKMKENKEISEDDFHGAKKEIQNITDDNIKKIDGELSIKEKDILNH\n\n### Sequence 2:\nMALQRTHSLLLLLLLTLLGLGLVQPSYGQDGMYQRFLRQHVHPEETGGSDRYCNLMMQRRKMTLYHCKRFNTFIHEDIWNIRSICSTTNIQCKNGKMNCHEGVVKVTDCRDTGSSRAPNCRYRAMASTRRVVIACEGNPQVPVHFDG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01505", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTDAANPPADISALSFEEALSQLERIVQELESGQAALERSIDIYERGAALKAHCEKKLEAARLKVEKIVLGQGGAVAAEPAEFN\n\n### Sequence 2:\nPADIAAMTFEQALAELEQIVARLESGQAPLEDSIRMYERGAALKAHCETRLEAARLRVEKIVMGAGGAPASEPAEFG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01506", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "XBP1, JCHAIN, SDC1, CD38, MZB1", "output": "This cell is most consistent with plasma cell, an antibody-secreting plasma cell. Key markers such as MZB1, XBP1, SDC1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasma cell", "tissue": "colon", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["XBP1", "JCHAIN", "SDC1", "CD38", "MZB1"]}}
{"id": "eval_01507", "category": "Mutation", "instruction": "Describe the effect of mutation P49768-V271L on protein Presenilin-1.", "input": "### Wild-type:\nMTELPAPLSYFQNAQMSEDNHLSNTVRSQNDNRERQEHNDRRSLGHPEPLSNGRPQGNSRQVVEQDEEEDEELTLKYGAKHVIMLFVPVTLCMVVVVATIKSVSFYTRKDGQLIYTPFTEDTETVGQRALHSILNAAIMISVIVVMTILLVVLYKYRCYKVIHAWLIISSLLLLFFFSFIYLGEVFKTYNVAVDYITVALLIWNFGVVGMISIHWKGPLRLQQAYLIMISALMALVFIKYLPEWTAWLILAVISVYDLVAVLCPKGPLRMVVETAQERNETLFPALIYSSTMVWLVNMAE\n\n### Mutant:\nMTELPAPLSYFQNAQMSEDNHLSNTVRSQNDNRERQEHNDRRSLGHPEPLSNGRPQGNSRQVVEQDEEEDEELTLKYGAKHVIMLFVPVTLCMVVVVATIKSVSFYTRKDGQLIYTPFTEDTETVGQRALHSILNAAIMISVIVVMTILLVVLYKYRCYKVIHAWLIISSLLLLFFFSFIYLGEVFKTYNVAVDYITVALLIWNFGVVGMISIHWKGPLRLQQAYLIMISALMALVFIKYLPEWTAWLILAVISVYDLVAVLCPKGPLRMLVETAQERNETLFPALIYSSTMVWLVNMAE\n\n### Answer:", "output": "Enhances protease activity with APP. The mutation in exon 8 of the PS-1 gene was detected in an Alzheimer's disease pedigree. It decreased secretion of the 42-amino acid amyloid-beta peptide, suggesting a benign mutation. The mutation also reduced the deletion of exon 8, leading to underexpression of the PS-1deltaexon8 splice isoform. PS-1deltaexon8 interacts with wild-type PS-1 to inhibit amyloid-beta production. PS-1deltaexon8 does not interact directly with Tau or GSK-3beta, potential enhancers of neuritic dystrophy.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01508", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nHSL53_DICDI\n\n### Answer:", "output": "HssA/B-like protein 53.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01509", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMVSVEELMGQVIDSEMEVISKELSKALEEALNLVKQKRDSVERTYTAKMQEMVTKAKEEIEGERARLDIEVKRAVLGEKNYWLNKVYEGTIKSLGTVKNSQGYKQGLESVLKRELRDGSIVYCSEDEVDQVQKMIKGLKAKAEVRSDPKIMGGVKIQYSDVGLVRDYSLNLILDQVFESLKPKIAEILFGEM\n\n### Sequence 2:\nMHKLFYLLSLLMAPFVANANFMIYPISKDLKNGNSELVRVYSKSKEIQYIKIYTKKIINPGTTEEYKVDIPNWDGGLVVTPQKVILPAGASKSIRLTQFKIPKKEEVYRVYFEAVKPDSKENVIDNKKLTTELSVNIIYAALIRSLPSEQNISLNISRNAKKNIIIYNNGNVRAGVKDIYFCKSSNIDDNCVKKAYNKNIYPEKSFDTLVNNNFSYVFIKLNHEGIEKEQGLIQLKVP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01510", "category": "Mutation", "instruction": "Describe the effect of mutation P02462-G525L on protein Collagen alpha-1.", "input": "### Wild-type:\nMGPRLSVWLLLLPAALLLHEEHSRAAAKGGCAGSGCGKCDCHGVKGQKGERGLPGLQGVIGFPGMQGPEGPQGPPGQKGDTGEPGLPGTKGTRGPPGASGYPGNPGLPGIPGQDGPPGPPGIPGCNGTKGERGPLGPPGLPGFAGNPGPPGLPGMKGDPGEILGHVPGMLLKGERGFPGIPGTPGPPGLPGLQGPVGPPGFTGPPGPPGPPGPPGEKGQMGLSFQGPKGDKGDQGVSGPPGVPGQAQVQEKGDFATKGEKGQKGEPGFQGMPGVGEKGEPGKPGPRGKPGKDGDKGEKGS\n\n### Mutant:\nMGPRLSVWLLLLPAALLLHEEHSRAAAKGGCAGSGCGKCDCHGVKGQKGERGLPGLQGVIGFPGMQGPEGPQGPPGQKGDTGEPGLPGTKGTRGPPGASGYPGNPGLPGIPGQDGPPGPPGIPGCNGTKGERGPLGPPGLPGFAGNPGPPGLPGMKGDPGEILGHVPGMLLKGERGFPGIPGTPGPPGLPGLQGPVGPPGFTGPPGPPGPPGPPGEKGQMGLSFQGPKGDKGDQGVSGPPGVPGQAQVQEKGDFATKGEKGQKGEPGFQGMPGVGEKGEPGKPGPRGKPGKDGDKGEKGS\n\n### Answer:", "output": "In HANAC; requires 2 nucleotide substitutions. The mutation affects a highly conserved glycine residue within the collagenous domain of the COL4A1 protein. This mutation is associated with the HANAC syndrome, which is characterized by systemic involvement and usually asymptomatic brain disease. Abnormal cell-type IV collagen interactions may underlie the systemic defects observed in this syndrome.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01511", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CC(C=O)Cc1ccc(C(C)(C)C)cc1", "output": "This molecule is small and relatively nonpolar, with lipophilic character. It contains aromatic ring, aldehyde. Its descriptor profile: MW=204.31, LogP=3.36, TPSA=17.07, HBD=0, HBA=1, QED=0.69.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CC(C=O)Cc1ccc(C(C)(C)C)cc1", "compound_name": null, "pubchem_cid": null, "MW": 204.31, "LogP": 3.36, "TPSA": 17.07, "QED": 0.69, "functional_groups": ["aromatic ring", "aldehyde"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01512", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nRDDGDEDYYKQRLRRWNKLRLQDKEESDAEFDEGFKVPGFLFKKLFKYQQTGVRWLWELHCQQAGGILGDEMGLGKTIQIIAFLAGLSYSKIRTRGSNYRFEGLGPTVIVCPTTVMHQWVKEFHTWWPPFRVAILHETGSYTHKKEKLIRDVAHCHGILITSYSYIRLMQDDISRYDWHYVILDEGHKIRNPNAAVTLACKQFRTPHRIILSGSPMQNNLRELWSLFDFIFPGKLGTLPVFMEQFSVPITMGGYSNASPVQVKTAYKCACVLRDTINPYLLRRMKSDVKMSLSLPDKNEQVLFCRLTDEQHKVYQNFVDSKEVYRILNGEMQIFSGLIALRKICNHPDLFSGGPKNLKGLPDDELEEDQFGYWKRSGKMIVVESLLKIWHKQGQRVLLFSQSRQMLDILEVFLRAQKYTYLKMDGTTTIASRQPLITRYNEDTSIFVFLLTTRVGGLGVNLTGANRVVIYDPDWNPSTDTQARERAWRIGQKKQVTVYRLLTAGTIEEKIYHRQIFKQFLTNRVLKDPKQRRFFKSNDLYELFTLTSPDASQSTETSAIFAGTGSDSNDDYVLEKLFKKSVGVHSVMKHDAIMDGASRFGKKRNPLASSSLLAKMRARNHL\n\n### Answer:", "output": "<SEQ_3Di>DAQLDPVVLVVLVVVVVVVVVVPPDFDWDDDPPFFTFDRVLVVVDDPVLVVLLVLLVVCVVVLFEFEAQAAPQNCPPSSVLRNLNRQQVRQDQDPDDPPRGRGFAAEEEEDDPVCPVVSRVVCCPRPVPAREFEPDPPTPHDDDSLVSLVVCRVVRHYYYYYLVCLLVVLCSNQVGQHAAYEYEQVLVLQDVPDSSLVSLQSRPGSRYYYYHNPCDWAFCSSVVSVCCSRPNCLCHDPVVNCVQQGVLLLLLLAPPRDPVSVLSNLLSLVVLLVSRVSRYDYDYCVRDCPSPDFAAEFEKEFAFAADPVLVVVLVCLLVDVVVVCCVVVVDPCVVVVVLLQVSQLPVCVNVVFDDDDPPDDPVVPAVRDRLRQVNHRLSLLVVLVVVLCVVVLAAEEEEEQDVVSLVSVVVVCVVVPAAEAEDDPVPDPVCPVVVLVCQLPPSSHRYYYYYLPPPPPDLQNQSGQEYEYSAADLTLVSVVRSCCSHGDPVRDDHHYYYYYDHFQALRVLSVVNSLSSVVSSCCNVPNVDFDHFHHPVVSPPHSDGDHDDLPDDGVVCVRPPPQPFQVSVVVSVVCCPVVVSRRSGMTGPVSSSCRDPPPQPDDPPDRPVVVVVVVVVSPDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01513", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAVGKNKRLSKGKKGLKKRTQDPFSRKDEYSVKAPSTFAVRDVGKTLVNRTTGLKNANDSLKGRIFEVSLADLQNDEDHAFRKVKLRVDEVQGKNCLTNFHGLDFTSDKLRSLVRKWQTLIEANVTVKTTDDYLLRLFAIAFTKRRPNQIKKTTYARSSQIRAIRKKITEIIQREASTRTLAQLTKLIPEVIGREIEKSTHGIYPLQNVHIRKVKLLKSPKFDLGALLALHGESSTDDKGQKVEREFKEQVLESV\n\n### Sequence 2:\nMDAIKKKMLAMKMEKENAIDRAEQMEQKLRDVEETKNKLEEEFNNLQNKFSNLQNDFDTANEGLTEAQTKLEASEKHVAELESDTAGLNRRIQLLEEDLERSEERLQSATEKLEEASKAADESERGRKVLESRSLADDERLDGLEAQLKEAKYIAEDAERKYDEAARKLAITEVDLERAEARLEAAEAKVWELDEELHIVGNNIKTLSIQNDQASQREDSYQETIRDLTQRLKDAENRATEAERTVSKLQKEVDRLEDELLAEKERYKSISDELDSTFAELAGY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01514", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "Cc1cc(CO)c(O)c(C(=O)CC2CC(=O)NC(=O)C2)c1", "output": "The SMILES represents an unnamed molecule with MW=291.3, LogP=0.82, TPSA=103.7. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "Cc1cc(CO)c(O)c(C(=O)CC2CC(=O)NC(=O)C2)c1", "compound_name": null, "pubchem_cid": null, "MW": 291.3, "LogP": 0.82, "TPSA": 103.7, "QED": 0.56, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ketone", "phenol"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01515", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMPFTLGQRWISDTESELGLGTVVALDARMVTIMFPATGENRLYARNDSPVTRVMFNPGDTVTSHEGWQLKIEDVKEENGLLAYTGTRLDTDEADVMLREVMLDSRLVFSKPQDRLFAGQIDRMDRFSLRYRARKFQSEQYRMPWSGLRGQRTSLIPHQLNIAHDVGRRHAPRVLLADEVGLGKTIEAGMILHQQLLSGAAERVLIVVPETLQHQWLVEMLRRFNLRFSLFDDERYAEAQHDADNPFETEQLVICSLDFVRRSKQRLEHLCDAEWDLLVVDEAHHLVWSENAPSREYMAIEQLAERVPGILLLTATPEQLGLESHFARLRLLDPNRFHDFDVFVEEQQNYRPVADAVAMLLAGKHLSNDELNTLSDLIGEQDIEPLLHTANSDRDGADAARQELVSMLMDRHGTSRVLFRNTRNGVKGFPKRELHTIKLPLPTQYQTAIKVSGIMGARKSAEERARDMLYPEQIYQEFEGDTGTWWNFDPRVEWLMGYLTAHRSQKVLVICAKAATALQLEQVLREREGIRAAVFHEGMSIIERDRAAAWFGEEDSGAQVLLCSEIGSEGRNFQFASKLVMFDLPFNPDLLEQRIGRLDRIGQAHDIQIHVPYLENTAQSVLVRWFHEGLDAFEHTCPTGRTIYDQVHSDLIGYLASPENTDGFDDLIKTCREKHDALKIQLEQGRDRLLEIHSNGGEKAQALAESIEEQDDDTSLISFAMNLFDIVGINQDDRGENMIVLTPSDHMLVPDFPGLPEDGCTITFERDVALSREDAQFITWEHPLIRNGLDLILSGDTGSSTISLLKNKALPVGTLLLELIYVVEAQAPKQLQLNRFLPATPVRLMLDKNGTNLAAQVEFESFNRQLSAVNRHTGSKLVNAVQQDVHAILQQGEAQIEKAARALIDSARREADEKLSAELSRLEALRAVNPNIRDDELAAIESNRQQVLESLDQASWRLDALRLIVVTHQ\n\n### Sequence 2:\nMPFTLGQRWISDTESELGLGTVVALDARMVTIMFPATGENRLYARNDSPVTRVMFNPGDTVTSHEGWQLKIEDVKEENGLLAYTGTRLDTDEADVMLREVMLDSRLVFSKPQDRLFAGQIDRMDRFSLRYRARKFQSEQYRMPWSGLRGQRTSLIPHQLNIAHDVGRRHAPRVLLADEVGLGKTIEAGMILHQQLLSGAAERVLIVVPETLQHQWLVEMLRRFNLRFSLFDDERYAEAQHDADNPFETEQLVICSLDFVRRSKQRLEHLCDAEWDLLVVDEAHHLVWSENAPSREYMAIEQLAERVPGILLLTATPEQLGLESHFARLRLLDPNRFHDFDVFVEEQQNYRPVADAVAMLLAGKHLSNDELNTLSDLIGEQDIEPLLHTANSDRDGADAARQELVSMLMDRHGTSRVLFRNTRNGVKGFPKRELHTIKLPLPTQYQTAIKVSGIMGARKSAEERARDMLYPEQIYQEFEGDTGTWWNFDPRVEWLMGYLTAHRSQKVLVICAKAATALQLEQVLREREGIRAAVFHEGMSIIERDRAAAWFGEEDSGAQVLLCSEIGSEGRNFQFASKLVMFDLPFNPDLLEQRIGRLDRIGQAHDIQIHVPYLENTAQSVLVRWFHEGLDAFEHTCPTGRTIYDQVHSDLIGYLASPENTDGFDDLIKTCREKHDALKIQLEQGRDRLLEIHSNGGEKAQALAESIEEQDDDTSLISFAMNLFDIVGINQDDRGENMIVLTPSDHMLVPDFPGLPEDGCTITFERDVALSREDAQFITWEHPLIRNGLDLILSGDTGSSTISLLKNKALPVGTLLLELIYVVEAQAPKQLQLNRFLPATPVRLMLDKNGTNLAAQVEFESFNRQLSAVNRHTGSKLVNAVQQDVHAILQQGEAQIEKAARALIDSARREADEKLSAELSRLEALRAVNPNIRDDELAAIESNRQQVLESLDQASWRLDALRLIVVTHQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01516", "category": "Mutation", "instruction": "Describe the effect of mutation E9Q9A9-L538P on protein 2'-5'-oligoadenylate synthase 2.", "input": "### Wild-type:\nMGNWLTGNWSSDRSSGYSSGWSPGGSSGVPSGPVHKLEKSIQANLTPNENCLKQIAVSSVPSQKLEGYIQENLKPNRESLKQIDQAVDAIWDLLRSQIPVKEVAKGGSYGRETALRGCSDGTLVLFMDCFQQFQDQIKYQDAYLDVIELWLKIHEKKSVKHEHALVVQVSVPGQRILLQLLPVFNPLRSNENPSSCVYVDLKKSMDQVRASPGEFSDCFTTLQQRFFKKYPQRLKDLILLVKHWYEQCQEKWKTPPPQPLLYALELLTVYAWEQGCQAEDFDMAQGVRTVLRLIQRPTEL\n\n### Mutant:\nMGNWLTGNWSSDRSSGYSSGWSPGGSSGVPSGPVHKLEKSIQANLTPNENCLKQIAVSSVPSQKLEGYIQENLKPNRESLKQIDQAVDAIWDLLRSQIPVKEVAKGGSYGRETALRGCSDGTLVLFMDCFQQFQDQIKYQDAYLDVIELWLKIHEKKSVKHEHALVVQVSVPGQRILLQLLPVFNPLRSNENPSSCVYVDLKKSMDQVRASPGEFSDCFTTLQQRFFKKYPQRLKDLILLVKHWYEQCQEKWKTPPPQPLLYALELLTVYAWEQGCQAEDFDMAQGVRTVLRLIQRPTEL\n\n### Answer:", "output": "Strongly increased 2'-5'-oligoadenylate synthase activity. The mutation renders the protein enzymatically active.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01517", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nY938_CAUVC\n\n### Answer:", "output": "Belongs to the UPF0337 (CsbD) family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01518", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVDRSNFKTCEESSFCKRQRSIRPGLSPYRALLDSLQLGPDSLTVHLIHEVTKVLLVLELQGLQKNMTRFRIDELEPRRPRYRVPDVLVADPPIARLSVSGRDENSVELTMAEGPYKIILTARPFRLDLLEDRSLLLSVNARGLLEFEHQRAPRVDEPGAWEETFKTHSDSKPYGPMSVGLDFSLPGMEHVYGIPEHADNLRLKVTEGGEPYRLYNLDVFQYELYNPMALYGSVPVLLAHNPHRDLGIFWLNAAETWVDISSNTGETPQTDVRWMSETGIIDVFLLLGPSISDVFRQYASLTGTQALPPLFSLGYHQSRWNYRDEADVLEVDQGFDDHNLPCDVIWLDIEHADGKRYFTWDPSRFPQPRTMLERLASKRRKLVAIVDPHIKVDSGYRVHEELRNLGLYVKTRDGSDYEGWCWPGSAGYPDFTNPTMRAWWANMFSYDNYEGSAPNLFVWNDMNEPSVFNGPEVTMLKDAQHYGGWEHRDVHNIYGLYVHMATADGLRQRSGGMERPFVLARAFFAGSQRFGAVWTGDNTAEWDHLKISIPMCLSLGLVGLSFCGADVGGFFKNPEPELLVRWYQMGAYQPFFRAHAHLDTGRREPWLLPSQHNDIIRDALGQRYSLLPFWYTLLYQAHREGIPVMRPLWVQYPQDVTTFNIDDQYLLGDALLVHPVSDSGAHGVQVYLPGQGEVWYDIQSYQKHHGPQTLYLPVTLSSIPVFQRGGTIVPRWMRVRRSSECMKDDPITLFVALSPQGTAQGELFLDDGHTFNYQTRQEFLLRRFSFSGNTLVSSSADPEGHFETPIWIERVVIIGAGKPAAVVLQTKGSPESRLSFQHDPETSVLVLRKPGINVASDWSIHLR\n\n### Answer:", "output": "<SEQ_3Di>DPPLQDDFLVLQVVSVVLVVDFFDFAQWFWDVPQFDFDFFWTWTWTARNVVRFIWIKIWFAADQQKIKIWIGGPDAVDAFDDDPLQFPDDTHTFGWDFPDDDPFWTWTDGNPAQWIWIFGGRHTKIFIDRNPHTFKIKCPRRSWDFDGQDDQDVPDPNQAWDDSPHFIFGDRRGFGKTKIKMKGWQFPFKAFQFLAQEFRTDDAQPPHTFAKAAQAQDDADDHRDSDHYRATAQKMWTADLFAIKMKGFRFQFIKGKGWDWDDDPTTMIMIMIMGRHGIGIIMMGGGRALLSSLLVLCNTRHFQAQAFLQQQAEEEEEQFDQALVRVVVVLVVLVVVQFAHAEYEGEQVQPPQSAPLEGDCPRYVQLQVSVVVCCVVRHAYEYEDEQWHADDPPRPVNVVCVVPPFFWAAQVGHFQQFQDSRGRTTGGQLVDPVVLVVLLQCLAPVNRDSDDLRYAYEYECQLVQGCPDRSRDHHQQTAHPPRDTCSRNRLVRQLSVLVSRQSSSCPNVDNWWGHEYEYAYHYNNSSNRHAYESRAHAQDLSVLQSQQSSQQNCLSRNRQQYGYAQQGAHDGYDLLSSLLSLQVLLVRSGHYHYYYPVDDDPRLVVGDPVSNVLSSVSSLVSQFLSQLSSVLSVCSNPSVRHQWHFPCNVVSNPPVRSSDRQWTDRPQWKIWRGDRDDPAQWDWDADADQQKWKAFPPLLDIDGDRDTDIGTDDSNDTRMITIAAEKGKTQNDDGNYCQVSLQGAIEIAHHHGPQRKHKYKGKHDNSTTPCCPPVVQIFIWMWMTHDQKTKIDGPDVSRHHDGNYHHFKYKYFPAAQLVWKWKDFDPPDTDTWDWDADPVSRIIITGGPVGHNRGIMMIGSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01519", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMQVSVETTQGLGRRVTITIAADSIETAVKSELVNVAKKVRIDGFRKGKVPMNIVAQRYGASVRQDVLGDLMSRNFVDAIIKEKINPAGAPNYVPGEYKVGEDFTYSVEFEVYPEVELTGLESIEVEKPVVEVTDADVDVMLDTLRKQQATWKEKDGAADAEDRVTIDFTGSVDGEEFEGGKATDFVLAMGQGRMIPGFEDGVKGHKAGEEFTIDVTFPEEYHAENLKGKAAKFVINLKKVEERELPELTEEFIKRFGVEDGSVAGLRAEVRKNMERELKGAVRNRVKSQAIEGLVKANDIDVPAALIDSEIDVLRRQAAQRFGGNEKQALELPRELFEEQAKRRVVVGLLLGEVIRTNELKADEERVKGLIEEMASAYEDPKEVIEFYSKNKELMDNMRNVALEEQAVEAVLAKAKVSEKATSFNELMNQQA\n\n### Sequence 2:\nMQVSVETTQGLGRRVTITIAADSIETAVKSELVNVAKKVRIDGFRKGKVPMNIVAQRYGASVRQDVLGDLMSRNFVDAIIKEKINPAGAPNYVPGEYKVGEDFTYSVEFEVYPEVELTGLESIEVEKPVVEVTDADVDVMLDTLRKQQATWKEKDGAADAEDRVTLDFTGTVDGEEFEGGKATDFVLAMGQGRMIPGFEDGVKGHKVGEEFTIDVTFPEEYHAENLKGKAAKFAINLKKVEERELPELTEEFIKRFGVEDGSVAGLRAEVRKNMERELKGAVRNRVKSQAIEGLVKANDIDVPSALIDSEIDVLRRQAAQRFGGNEKQALELPRELFEEQAKRRVVVGLLLGEVIRTNELKADEVRVKGLIEEMASAYEDPKEVIEFYSKNKELMDNMRNVALEEQAIEAVLAKAKVSEKATSFNELMNQQA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01520", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNTPPB_COLP3\n\n### Answer:", "output": "7-methyl-GTP pyrophosphatase, m(7)GTP pyrophosphatase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01521", "category": "Mutation", "instruction": "Describe the effect of mutation Q8NBS3-G818S on protein Solute carrier family 4 member 11.", "input": "### Wild-type:\nMAAATRRVFHLQPCENSPTMSQNGYFEDSSYYKCDTDDTFEAREEILGDEAFDTANSSIVSGESIRFFVNVNLEMQATNTENEATSGGCVLLHTSRKYLKLKNFKEEIRAHRDLDGFLAQASIVLNETATSLDNVLRTMLRRFARDPDNNEPNCNLDLLMAMLFTDAGAPMRGKVHLLSDTIQGVTATVTGVRYQQSWLCIICTMKALQKRHVCISRLVRPQNWGENSCEVRFVILVLAPPKMKSTKTAMEVARTFATMFSDIAFRQKLLETRTEEEFKEALVHQRQLLTMVSHGPVAPR\n\n### Mutant:\nMAAATRRVFHLQPCENSPTMSQNGYFEDSSYYKCDTDDTFEAREEILGDEAFDTANSSIVSGESIRFFVNVNLEMQATNTENEATSGGCVLLHTSRKYLKLKNFKEEIRAHRDLDGFLAQASIVLNETATSLDNVLRTMLRRFARDPDNNEPNCNLDLLMAMLFTDAGAPMRGKVHLLSDTIQGVTATVTGVRYQQSWLCIICTMKALQKRHVCISRLVRPQNWGENSCEVRFVILVLAPPKMKSTKTAMEVARTFATMFSDIAFRQKLLETRTEEEFKEALVHQRQLLTMVSHGPVAPR\n\n### Answer:", "output": "In FECD4; does not interfere with post-translational processing; the mutant protein partially localizes to the cytoplasm.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01522", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTNAA_HALSA\n\n### Answer:", "output": "Probable tryptophanase, L-tryptophan indole-lyase, TNase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01523", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CCI", "output": "The molecule contains the following functional groups: halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "ESOL", "canonical_smiles": "CCI", "compound_name": null, "pubchem_cid": null, "MW": 155.97, "LogP": 1.44, "TPSA": 0.0, "QED": 0.37, "functional_groups": ["halide"], "moleculenet_labels": {"label": -1.6}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["halide"]}}
{"id": "eval_01524", "category": "Mutation", "instruction": "Describe the effect of mutation Q99176-F95D on protein Protein SRN2.", "input": "### Wild-type:\nMKVKATKLRIKQRRKNKGLNISRLDIIRAEMDVVPSPGLPEKVNEKSKNIPLPEGINLLSSKEIIDLIQTHRHQLELYVTKFNPLTDFAGKIHAFRDQFKQLEENFEDLHEQKDKVQALLENCRILESKYVASWQDYHSEFSKKYGDIALKKKLEQNTKKLDEESSQLETTTRSIDSADDLDQFIKNYLDIRTQYHLRREKLATWDKQGNLKY\n\n### Mutant:\nMKVKATKLRIKQRRKNKGLNISRLDIIRAEMDVVPSPGLPEKVNEKSKNIPLPEGINLLSSKEIIDLIQTHRHQLELYVTKFNPLTDFAGKIHADRDQFKQLEENFEDLHEQKDKVQALLENCRILESKYVASWQDYHSEFSKKYGDIALKKKLEQNTKKLDEESSQLETTTRSIDSADDLDQFIKNYLDIRTQYHLRREKLATWDKQGNLKY\n\n### Answer:", "output": "Defective in ESCRT-I cargo sorting; when associated with D-99.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01525", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSIVTKSIVNADAEARYLSPGELDRIKAFVTGGAARLRIAETLTGSRETIVKQAGDRLFQKRPDIVSPGGNAYGEEMTATCLRDMDYYLRLVTYGVVSGDVTPIEEIGLVGVREMYRSLGTPIEAVAQSVREMKEVASGLMSSDDAAEASAYFDFVIGKMS\n\n### Answer:", "output": "<SEQ_3Di>DLVVVQVVVCVVVVHDRDPVSVVVVVVCVVCLVVLVVLLVLLVVCLVPLLVQLLVVLCVVPVQCCPPPHLQPDPVSSVVLSVVSSVLSVLLSVCSNVVHCPSVCVPDQPCNVVVCVVSVHDLLSVLSSLVSSLVSSVVSDDDPVSVVSVVSSVVSSVSRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01526", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMPKDAPAGLAPGFLVAAPALADPNFNGSLVLMAEHHAQGALGFVVNRPGPITVADVLGGLDERLRERAEGAGRADDPVLVGGPVQPERLWILFRPGPTAPEEGAVALGSGLALGGSRELLEALVRSRDPGPYLLLLGYAGWAPLQIEHEVGEGAWVPLPLQGDLVFDVPMEKRWETAVRRLGLDPAGFLVGGGGAEA\n\n### Sequence 2:\nSLRDHFLIAMPGLQDPIFSRSLTYICDHTAQGAMGIVVNQPMNLTLGDIFEQLE--LQDKAQQAGRA---VLAGGPVNTERGFVLHRD--SGAWESTMHIAPDVNLTASRDIVHAIANNTGPKSSLFALGYAGWSAGQLEEEISANSWLTIPADSSIIFDIPVEDRWAAAARQLGID\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01527", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nUVRC_ALIF1\n\n### Answer:", "output": "Belongs to the UvrC family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01528", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSCD2_CAEEL\n\n### Answer:", "output": "ALK tyrosine kinase receptor homolog scd-2, Suppressor of constitutive dauer formation protein 2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01529", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "CCn1cc[n+](C)c1.COS(=O)(=O)[O-]", "output": "This molecule is small and moderately polar, with hydrophilic character. It contains no standard functional groups detected. Its descriptor profile: MW=222.27, LogP=-0.57, TPSA=75.24, HBD=0, HBA=4, QED=0.38.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CCn1cc[n+](C)c1.COS(=O)(=O)[O-]", "compound_name": null, "pubchem_cid": null, "MW": 222.27, "LogP": -0.57, "TPSA": 75.24, "QED": 0.38, "functional_groups": [], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01530", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CC#Cc1cncc(-c2cccc(C3(C4CC4)N=C(C)C(N)=N3)c2)c1\nDescriptors: MW=328.42, LogP=3.51, TPSA=63.63, HBD=1, HBA=4", "output": "Descriptors: MW=328.42, LogP=3.51, TPSA=63.63, HBD=1, HBA=4, QED=0.88, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Lipophilicity", "canonical_smiles": "CC#Cc1cncc(-c2cccc(C3(C4CC4)N=C(C)C(N)=N3)c2)c1", "compound_name": null, "pubchem_cid": null, "MW": 328.42, "LogP": 3.51, "TPSA": 63.63, "QED": 0.88, "functional_groups": ["aromatic ring", "amine", "pyridine"], "moleculenet_labels": {"label": 3.3}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01531", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAPNLRKSHPLLKLINNSLIDLPTPSNISAWWNFGSLLGICLLTQILTGLLLATHYTADTTLAFSSVAHTCRNVQYGWLIRNLHANGASFFFICIYLHIGRGFYYGSYLNKETWNTGIILLLTLMATAFVGYVLPWGQMSFWGATVITNLFSAIPYIGQTLVEWAWGGFSVDNPTLTRFFALHFLLPFMIAGLALIHLTFLHESGSNNPLGILSNCDKIPFHPYFSLKDILGFIIMFLPLTTLALFSPNLLGDPENFTPANPLVTPPHIKPEWYFLFAYAILRSIPNKLGGVLALAASVLVLFLVPLLHKSKQRAMTFRPLSQFLFWTLVANLLILTWVGSQPVEHPFIIIGQLASLTYFTILLLLFPIIGALENKMLNY\n\n### Sequence 2:\nMLDIKLIREQPDEVKRRLARCGVDGAVVDQVLAFDEQRRRLIYEVETRKAERNTVSKQIGAMKDPAERQAKIDAMRQLGDEIAALDRQLAEVEEQQRAVMLEIRNLPHPDVPDGVDDRDNVVIYQEGEERQLPFPARPHWELGEALGIIDFERGVKLAGSRFYVMRGAGARLQRAVIQWLIDLHLEQGYQEVYTPFVVKESVLWASGQLPKFRDNLYRDEESGLWLVPTAEVPLTSLYADEILDASQLPIYHVAYTPCFRKEQLSAGRDVRGIKRGHQFDKVEMYMFVTPDQSYQALEKLRRDAEECARRLGLPFRTKLLCTGDLGFGSTKTYDIEVWAPGVGEWLEVSSCSNVEAFQARRANLRYRPEPGAKPEFLHTLNGSGLGLPRTIIAIMENYQQEDGSILIPEVLRPYMGGMERIGP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01532", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSGSTRTGSASVLLVGLSFRSAPVTMLEQATVADADLPKMQLSLVDNDVISESLVLSTCNRMEFYTVANAFHSGLDHVVDTIAQFSGLQTAELEPHLYVHYADSAAEHMLKVASGLDSMVIGEQQIIGQLRSAYQSANETGTVGRTLHDLTQRALRTGKRVHSETAIDSAGASMVSFALDQALRYIEPARALSAVVDDAPLSQPLAGHRALIIGAGAMASLASTHLGKLGIDHVTVANRTLSRAENLVNHARQAGVDASAVPLDGVTDCLSAVDIVVSATGAVGNVVTEQDVRAAVGAAGGVASVAGRRGTKVMIDLSMPADIEHSVAEIDGVKLLNIEELTTMAGDRVQDESPARAIVADELQSFLEQQRAQSVVPTVKALRQKAGEVMAEELMALERLTPDMSEADRAAVVKSMKRVVDKLLHTPTVQAKKLSAGGQQVSYPDALAALFNLPNGMVDSVTQPGQADSSAAQTAGTSARADQIPSAARVGRVVREA\n\n### Sequence 2:\nMLVVGMSHRSAPVALLERLSMDDSVRGQTTSALVERPSLSEALIVSTCNRLEVYTVTSSFHTGVNDVVEVLHEISGVDIETLRGYLYVRYADAAAEHMLVVASGLDSMVVGEQQIIGQVRTAYQQATEAGTVGPALHALAQTALHTGKRVHTETDIDEAGASMVSFAIDRALTQMGIDPAS-----------DKPLAGKTALVLGAGAMSSLAATHLGREGVDKLIMANRTRERAERLASHSLEAGVPAEVVDFADRASVLDRVDMVVSATGADDFTVKPADIPAGV--------------QLMLVDLSMPRDIDDACAEVPGVDLVNIERL-----HRAKREDPTGAAAEGADALAIVREELEAFTSEQRIRDVVPAVSALRKRANELLLDELDRLQARTPDISQEDWKEVTRTVRRVMDKFLHEPTVRVKKLAARSGSVSYESALQELFGL------------EAVQTTAPPAITSVNASDLPDAGIVAIV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01533", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "CCN1C(=O)c2ccccc2C1Nc1ccc(OCCN2CCCCC2)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "CCN1C(=O)c2ccccc2C1Nc1ccc(OCCN2CCCCC2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 379.5, "LogP": 4.14, "TPSA": 44.81, "QED": 0.79, "functional_groups": ["aromatic ring", "amide", "amine", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "ether"]}}
{"id": "eval_01534", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nACCD_CRYJA\n\n### Answer:", "output": "Component of the acetyl coenzyme A carboxylase (ACC) complex.  Biotin carboxylase (BC) catalyzes the carboxylation of biotin on its carrier protein (BCCP) and then the CO(2) group is transferred by the transcarboxylase to acetyl-CoA to form malonyl-CoA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01535", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nAPRG1_HUMAN\n\n### Answer:", "output": "APRG1 tumor suppressor candidate, AP20 region protein 1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01536", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRPOC1_MORIN\n\n### Answer:", "output": "DNA-directed RNA polymerase subunit beta', PEP, Plastid-encoded RNA polymerase subunit beta', RNA polymerase subunit beta'.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01537", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nS100A4 FAP ACTA2 PDGFRB COL1A1\n</cell>", "output": "This cell is most consistent with cancer-associated fibroblast, a cancer-associated fibroblast (CAF). Key markers such as ACTA2, PDGFRB, FAP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "bone marrow", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["S100A4", "FAP", "ACTA2", "PDGFRB", "COL1A1"]}}
{"id": "eval_01538", "category": "Mutation", "instruction": "Describe the effect of mutation P80192-A308S on protein Mitogen-activated protein kinase kinase kinase 9.", "input": "### Wild-type:\nMEPSRALLGCLASAAAAAPPGEDGAGAGAEEEEEEEEEAAAAVGPGELGCDAPLPYWTAVFEYEAAGEDELTLRLGDVVEVLSKDSQVSGDEGWWTGQLNQRVGIFPSNYVTPRSAFSSRCQPGGEDPSCYPPIQLLEIDFAELTLEEIIGIGGFGKVYRAFWIGDEVAVKAARHDPDEDISQTIENVRQEAKLFAMLKHPNIIALRGVCLKEPNLCLVMEFARGGPLNRVLSGKRIPPDILVNWAVQIARGMNYLHDEAIVPIIHRDLKSSNILILQKVENGDLSNKILKITDFGLARE\n\n### Mutant:\nMEPSRALLGCLASAAAAAPPGEDGAGAGAEEEEEEEEEAAAAVGPGELGCDAPLPYWTAVFEYEAAGEDELTLRLGDVVEVLSKDSQVSGDEGWWTGQLNQRVGIFPSNYVTPRSAFSSRCQPGGEDPSCYPPIQLLEIDFAELTLEEIIGIGGFGKVYRAFWIGDEVAVKAARHDPDEDISQTIENVRQEAKLFAMLKHPNIIALRGVCLKEPNLCLVMEFARGGPLNRVLSGKRIPPDILVNWAVQIARGMNYLHDEAIVPIIHRDLKSSNILILQKVENGDLSNKILKITDFGLARE\n\n### Answer:", "output": "Enhances JNK activation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01539", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nCD4:low CD45RO:medium S100A4:medium IL7R:medium\n</cell>", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as S100A4, CD4, CD45RO support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "blood", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CD4", "CD45RO", "S100A4", "IL7R"]}}
{"id": "eval_01540", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKRIAFVGAVGAGKTTLFNALRGNYSLARKTQAVEFNDHGDIDTPGEYFSHPRWYHALITTLQDVDTLIYVHAANDKESRLPAGLLDVGTRKRHIAVISKTDMPDADVAATRQLLCEIGFREPIFELNGHDPQSVRQLVDYLAALSEQEEEAGEKTYHS\n\n### Sequence 2:\nMKRIAFVGAVGAGKTTLFNALRGNYSLARKTQAVEFNDHGDIDTPGEYFSHPRWYHALITTLQDVDTLIYVHAANDKESRLPAGLLDVGTRKRHIAVISKTDMPDADVAATRQLLCEIGFREPIFELNGHDPQSVRQLVDYLAALSEQEEEAGEKTYHS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01541", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nSLC17A7:low SATB2:high RBFOX3:medium\n</cell>", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as RBFOX3, SLC17A7, SATB2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "breast", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["SLC17A7", "SATB2", "RBFOX3"]}}
{"id": "eval_01542", "category": "Mutation", "instruction": "Describe the effect of mutation T1WGN1-A712W on protein Endo-beta-N-acetylglucosaminidase EndoS2.", "input": "### Wild-type:\nMDKHLLVKRTLGCVCAATLMGAALATHHDSLNTVKAEEKTVQTGKTDQQVGAKLVQEIREGKRGPLYAGYFRTWHDRASTGIDGKQQHPENTMAEVPKEVDILFVFHDHTASDSPFWSELKDSYVHKLHQQGTALVQTIGVNELNGRTGLSKDYPDTPEGNKALAAAIVKAFVTDRGVDGLDIDIEHEFTNKRTPEEDARALNVFKEIAQLIGKNGSDKSKLLIMDTTLSVENNPIFKGIAEDLDYLLRQYYGSQGGEAEVDTINSDWNQYQNYIDASQFMIGFSFFEESASKGNLWFDV\n\n### Mutant:\nMDKHLLVKRTLGCVCAATLMGAALATHHDSLNTVKAEEKTVQTGKTDQQVGAKLVQEIREGKRGPLYAGYFRTWHDRASTGIDGKQQHPENTMAEVPKEVDILFVFHDHTASDSPFWSELKDSYVHKLHQQGTALVQTIGVNELNGRTGLSKDYPDTPEGNKALAAAIVKAFVTDRGVDGLDIDIEHEFTNKRTPEEDARALNVFKEIAQLIGKNGSDKSKLLIMDTTLSVENNPIFKGIAEDLDYLLRQYYGSQGGEAEVDTINSDWNQYQNYIDASQFMIGFSFFEESASKGNLWFDV\n\n### Answer:", "output": "Enhanced endoglucosidase activity; enhanced activity on both high-mannose and complex-type glycans.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01543", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGLO2_ACTPJ\n\n### Answer:", "output": "Thiolesterase that catalyzes the hydrolysis of S-D-lactoyl-glutathione to form glutathione and D-lactic acid.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01544", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCYSC_SHISS\n\n### Answer:", "output": "Adenylyl-sulfate kinase, APS kinase, ATP adenosine-5'-phosphosulfate 3'-phosphotransferase, Adenosine-5'-phosphosulfate kinase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01545", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSYDND_PSEMY\n\n### Answer:", "output": "Aspartate--tRNA(Asp/Asn) ligase, Aspartyl-tRNA synthetase, AspRS, Non-discriminating aspartyl-tRNA synthetase, ND-AspRS.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01546", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGMNHNYVEVKYPAIFRDEGTYWDVRFPDVPAAQTFGASVQVAADNAANALAIALFEQSLPPASDPQYWRLASTEFVVWITMADVQFGPGADTPEPMN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCEEEEEEEECSSSEEEECTTCGGGCEEESSHHHHHHHHHHHHHHHHTTSCCCCCCCGGGCCCCTTEEEEEEECCCCCCSTTCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01547", "category": "Mutation", "instruction": "Describe the effect of mutation P84507-R27N on protein Beta-theraphotoxin-Cm1a.", "input": "### Wild-type:\nDCLGWFKSCDPKNDKCCKNYTCSRRDRWCKYDL\n\n### Mutant:\nDCLGWFKSCDPKNDKCCKNYTCSRRDNWCKYDL\n\n### Answer:", "output": "Important increase in potency toward Nav1.7/SCN9A, and low potency on Nav1.4/SCN4A and Nav1.5/SCN5A; synthetic variant D1Z/M5I/R27Na.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01548", "category": "Mutation", "instruction": "Describe the effect of mutation Q06520-Y238W on protein Sulfotransferase 2A1.", "input": "### Wild-type:\nMSDDFLWFEGIAFPTMGFRSETLRKVRDEFVIRDEDVIILTYPKSGTNWLAEILCLMHSKGDAKWIQSVPIWERSPWVESEIGYTALSETESPRLFSSHLPIQLFPKSFFSSKAKVIYLMRNPRDVLVSGYFFWKNMKFIKKPKSWEEYFEWFCQGTVLYGSWFDHIHGWMPMREEKNFLLLSYEELKQDTGRTIEKICQFLGKTLEPEELNLILKNSSFQSMKENKMSNYSLLSVDYVVDKAQLLRKGVSGDWKNHFTVAQAEDFDKLFQEKMADLPRELFPWE\n\n### Mutant:\nMSDDFLWFEGIAFPTMGFRSETLRKVRDEFVIRDEDVIILTYPKSGTNWLAEILCLMHSKGDAKWIQSVPIWERSPWVESEIGYTALSETESPRLFSSHLPIQLFPKSFFSSKAKVIYLMRNPRDVLVSGYFFWKNMKFIKKPKSWEEYFEWFCQGTVLYGSWFDHIHGWMPMREEKNFLLLSYEELKQDTGRTIEKICQFLGKTLEPEELNLILKNSSFQSMKENKMSNYSLLSVDWVVDKAQLLRKGVSGDWKNHFTVAQAEDFDKLFQEKMADLPRELFPWE\n\n### Answer:", "output": "Strongly reduces substrate inhibition when ADT or DHEA are used as substrates.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01549", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRHGD_ASPNC\n\n### Answer:", "output": "Belongs to the glycosyl hydrolase 28 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01550", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMCM7_YEAST\n\n### Answer:", "output": "Acts as component of the MCM2-7 complex (MCM complex) which is the putative replicative helicase essential for 'once per cell cycle' DNA replication initiation and elongation in eukaryotic cells.  Core component of CDC45-MCM-GINS (CMG) helicase, the molecular machine that unwinds template DNA during replication, and around which the replisome is built.  The active ATPase sites in the MCM2-7 ring are formed through the interaction surfaces of two neighboring subunits such that a critical structure of a conserved arginine finger motif is provided in trans relative to the ATP-binding site of the Walker A box of the adjacent subunit.  The six ATPase active sites, however, are likely to contribute differentially to the complex helicase activity.  Once loaded onto DNA, double hexamers can slide on dsDNA in the absence of ATPase activity.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01551", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nLDRADILYNIRQTSRPDVIPTQRDRPVAVSVSLKFINILEVNEITNEVDVVFWQQTTWSDRTLAWNHSPDQVSVPISSLWVPDLAAYNAISKPEVLTPQLARVVSDGEVLYMPSIRQRFSCDVSGVDTESGATCRIKIGSWTHHSREISVDPTTEDDSEYFSQYSRFEILDVTQKKNSVTYSCCPEAYEDVEVSLNFRKKG\n\n### Answer:", "output": "<SEQ_3Di>DAPVVLQVQLLVPADQPDQQDDPLAAFEKEWEWAWAAWADQDQVQQKTKTKTWIKMKTFRQSSADDVYDFKDKAASVSHHDFQKDFPFFPDDWDFPDHRIWMAGNRGMIIGITITITMGRFDQNQQAPDQGGKGKTKMDGPPDFNSHYDYHYDDDPDNPNYDPPYQKDWPDKDKDKDWDDDPVDDTIHIIIIIMTGMHGDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01552", "category": "Mutation", "instruction": "Describe the effect of mutation O18191-S828F on protein Anaphase-promoting complex subunit 1.", "input": "### Wild-type:\nMRKYVLFFISGNNDSQIWATTPNTPRVIARGGLERNIHTRTLARMVNEDAPGTSTPAAQSRLQTTASPFHRTHLTQMCRRGDTNASLLRDFTRMIRDTPRNFSKNTQHGNDRDFGDLERDPDVDLLLSKVCLECVYVEPKEGAIPKANKIFISNFLSDMYINLVSVTGEVMKIIPIWKNAETTRKNLLEKGKHEPCVVDCVDAAFVMKSGITVVLGSDFTTAMFGGNERIAPIFIKEMSNQRVGRKFRLFSFAENRIFAVNEMRCIVVEIPETVTCKSATELMRTCFLHLDRDLSRKLLI\n\n### Mutant:\nMRKYVLFFISGNNDSQIWATTPNTPRVIARGGLERNIHTRTLARMVNEDAPGTSTPAAQSRLQTTASPFHRTHLTQMCRRGDTNASLLRDFTRMIRDTPRNFSKNTQHGNDRDFGDLERDPDVDLLLSKVCLECVYVEPKEGAIPKANKIFISNFLSDMYINLVSVTGEVMKIIPIWKNAETTRKNLLEKGKHEPCVVDCVDAAFVMKSGITVVLGSDFTTAMFGGNERIAPIFIKEMSNQRVGRKFRLFSFAENRIFAVNEMRCIVVEIPETVTCKSATELMRTCFLHLDRDLSRKLLI\n\n### Answer:", "output": "In ax102; results in sterility.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01553", "category": "Mutation", "instruction": "Describe the effect of mutation Q9LIQ6-W203A on protein PHD finger protein ING1.", "input": "### Wild-type:\nMSFAEEFEANLVSLAHVLQKKYALLRDLDKSLQENQRQNEQRCEKEIEDIRRGRAGNITPNTSLTKFSEEALDEQKHSVRIADEKVTLAMQAYDLVDMHVQQLDQYMKKSDEVIRKEKEAAAATLELENNGKAGNAGEGGRGGRKKTRLATAASTAAASTGMTSSNMDLDLPVDPNEPTYCICNQVSFGEMVACDNNACKIEWFHFGCVGLKEQPKGKWYCPECATVKKSRKGR\n\n### Mutant:\nMSFAEEFEANLVSLAHVLQKKYALLRDLDKSLQENQRQNEQRCEKEIEDIRRGRAGNITPNTSLTKFSEEALDEQKHSVRIADEKVTLAMQAYDLVDMHVQQLDQYMKKSDEVIRKEKEAAAATLELENNGKAGNAGEGGRGGRKKTRLATAASTAAASTGMTSSNMDLDLPVDPNEPTYCICNQVSFGEMVACDNNACKIEAFHFGCVGLKEQPKGKWYCPECATVKKSRKGR\n\n### Answer:", "output": "Abolishes binding to H3K4me2/3. The mutational analysis confirmed that the mutation in AtING protein is crucial for binding to H3K4me3/2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01554", "category": "Mutation", "instruction": "Describe the effect of mutation Q96RP9-M496R on protein Elongation factor G, mitochondrial.", "input": "### Wild-type:\nMRLLGAAAVAALGRGRAPASLGWQRKQVNWKACRWSSSGVIPNEKIRNIGISAHIDSGKTTLTERVLYYTGRIAKMHEVKGKDGVGAVMDSMELERQRGITIQSAATYTMWKDVNINIIDTPGHVDFTIEVERALRVLDGAVLVLCAVGGVQCQTMTVNRQMKRYNVPFLTFINKLDRMGSNPARALQQMRSKLNHNAAFMQIPMGLEGNFKGIVDLIEERAIYFDGDFGQIVRYGEIPAELRAAATDHRQELIECVANSDEQLGEMFLEEKIPSISDLKLAIRRATLKRSFTPVFLGSA\n\n### Mutant:\nMRLLGAAAVAALGRGRAPASLGWQRKQVNWKACRWSSSGVIPNEKIRNIGISAHIDSGKTTLTERVLYYTGRIAKMHEVKGKDGVGAVMDSMELERQRGITIQSAATYTMWKDVNINIIDTPGHVDFTIEVERALRVLDGAVLVLCAVGGVQCQTMTVNRQMKRYNVPFLTFINKLDRMGSNPARALQQMRSKLNHNAAFMQIPMGLEGNFKGIVDLIEERAIYFDGDFGQIVRYGEIPAELRAAATDHRQELIECVANSDEQLGEMFLEEKIPSISDLKLAIRRATLKRSFTPVFLGSA\n\n### Answer:", "output": "In COXPD1. The EFG1-mutant patient had early-onset Leigh syndrome.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01555", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nVSFRQKRTRIPLLAMTVTALAAAVCGVTTAPAATGAEVAVPLSVGAAAGNATPIPGYVIQSSAQVSDDSAVSKPGFPTSGWYPVSSRSTVYAGLLQNGKYADPFYSTNMQNVPAAQFSVPWWYRTDLNVDDTSSRTYLDFSGVLSKADVWVNGTKVATKDQVNGAYTRHDLDITAQVHTGVNSVAFKVYPNDPNRDLSMGWIDWAQTPPDQNMGIVRDVLVRRSGAVALRSAHVIQKLNSALDHADLTVKADVRNDSANAVQTTVAGTVAGKPISQTVSLAAKERKTVTFPLVGLDRPNVWWPAGMGGQHRYDLDLTASVGGTPSDAAKSKFGVRDVKATLNSSGGRQYSVNGKPLLIRGGGYTPDLFLRWNETAAADKLKYVLNLGLNTVRLEGHIEPDEFFDIADDLGVLTMPGWECCDKWEGQVNGEEKGEPWVESDYPIAKASMFSEAERLRDHPSVISFHIGSDFAPDRRIEQGYLDAMKAADFLLPVIPAASARPSPITGASGMKMNGPYDYVPPVYWYDKSQKDRGGAWSFNSETSAGVDIPTMDTLKRMMSASELDTMWKNPSAKQYHRSSSDTFGNLKLFGDALTKRYGASANLNDFVRKAQLSQYENVRAEFESHSRNYTDSTNPSTGLIYWMLNSPWTSLHWQLFDAYMDQNGAYYGAKKANEPLHIQYSHDNRSVVVINQTSNAVSGLTATTKLYNLDGTEKYSNTKTGLSVGALGAKATAVTVPAVSGLSTTYLAKNVLTDSSGKEVSRNVYWLSTKADTLNWGGSDWYYTPQSAFADLSGLNNLGQSAVGATANSVAGADGTTTTTVTLKNTSGGRLPAFYVDSKVVDSAGKPVLPVEWNDNAVSLWPGETTTLTAKYRTADLKGSKPSVRISGWNTGTQTVPADGSGPGPSDPVDYQAEDATIVQGAVESNHAGYTGTGFVNYDNVAGSSVEWTVTVPSAGTYDVVVRYANGTTTSRPLDFSVNGSISASGVAFGSTGTWPAWTTKTVRVTLAAGVNKIKAVATTANGGPNVDKITL\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCEEEECCEEEEEETTSCSCGGGTTSTTCCCTTCEEEETTCBHHHHHHHTTCSCCTTSTTGGGGSCSGGGSSCEEEEEEEEESCSSSEEEEEECCEESEEEEEETTEEEEEEEEEESTTCCEEEECTTTCCSEEEEEEEEECCCCTTTSSSCCCTTTSCCCTTTTCEECSCEEEEEESSEEEEEEEEEEEECTTSSEEEEEEEEEEEECSSSCEEEEEEEEETTEEEEEEEEECTTCEEEEECCCEEEESCCCCCCTTSSCCCCEEEEEEEEETTEEEEEEEEEECCCCEEEEECTTSCEEEEETTEECCEEEEEECCCTTCCCCHHHHHHHHHHHHHHTCCEEEEESCCCCHHHHHHHHHHTCEEEEECCSSSGGGTTTSTTSSSCCCCTTHHHHHHHHHHHHHHHHTTCTTBCCEESCSSSCCCHHHHHHHHHHHHHTTCCSCEESCSSSCCBTTTBCCCCBCCCCCSCCCGGGGGCCSCTTSCSSSEEEEEECCSBCCCCHHHHHHHSCHHHHHHHHHCTTSCCTTSCSSSTTSSCHHHHHHHHHHHCCCSSHHHHHHHHHHHHHHHHHHHHHHHHHGGGCSSSCEEEEEESCSCCSSSCSBSCSBCTTCCBCHHHHHHHHHTCSEEEEECTTTCEEEEEECSSSCEEEEEEEEEEEETTCCEEEEEEEEEEEECSTTEEEEEEECCCCTTCCSEEEEEEEEECTTSCEEEEEEEEEESSCCEECGGGCBTTBCCEEECCCCGGGGGCCCCCEEEEEEEEECTTSEEEEEEEEEECCSSSCCEEEEEEEEECTTCCBCSSCEESSCSEEECTTCEEEEEEEEEGGGGTTCCCEEEEEETTTEEEEEECCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01556", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nGAD2 GAD1 SLC32A1\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as GAD2, SLC32A1, GAD1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "pancreas", "disease": "normal", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["GAD2", "GAD1", "SLC32A1"]}}
{"id": "eval_01557", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCLC1A_MOUSE\n\n### Answer:", "output": "C-type lectin domain family 1 member A.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01558", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nYSV4_CAEEL\n\n### Answer:", "output": "Membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01559", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMIGRLRGILLEKQPPEVLIEVNGIGYEVQMPMSCFYELPNIGEEAIIYTHFVVREDAQLLYGFNTVKERALFREVIKANGVGPKLGLGILSGMTASQFVSCVEREDVSTLVKLPGVGKKTAERLVVEMKDRLKGWGAGDLFTPFTDAAPTDSAAASSNSAEEEAVSALLALGYKPTQASKVVSQIAKPDMSSEQLIREALKSMV\n\n### Sequence 2:\nMKNNTKIIETKCSMEQINVSPKYYDIISFGYFPNYFNINNSTNSIENQSNNIKEYKISILEDFQENDNQTNKLCTTITNNIKIIKKISKSKKLNSKSHIFKKRNFKSNCTSSNQNSNNVIATLILPKIDDNNNNNNNDQIENNSINCINNNNINNNINENNKFTWVYYHYDCNGKRKSIYDDNEMSELSTTPFPKNHICSKCSSSQRSVFKLNKFGKLDCSFCLNNNNNTTTTTTTTPQPNSDFN\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01560", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NVV9-P47A on protein THAP domain-containing protein 1.", "input": "### Wild-type:\nMVQSCSAYGCKNRYDKDKPVSFHKFPLTRPSLCKEWEAAVRRKNFKPTKYSSICSEHFTPDCFKRECNNKLLKENAVPTIFLCTEPHDKKEDLLEPQEQLPPPPLPPPVSQVDAAIGLLMPPLQTPVNLSVFCDHNYTVEDTMHQRKRIHQLEQQVEKLRKKLKTAQQRCRRQERQLEKLKEVVHFQKEKDDVSERGYVILPNDYFEIVEVPA\n\n### Mutant:\nMVQSCSAYGCKNRYDKDKPVSFHKFPLTRPSLCKEWEAAVRRKNFKATKYSSICSEHFTPDCFKRECNNKLLKENAVPTIFLCTEPHDKKEDLLEPQEQLPPPPLPPPVSQVDAAIGLLMPPLQTPVNLSVFCDHNYTVEDTMHQRKRIHQLEQQVEKLRKKLKTAQQRCRRQERQLEKLKEVVHFQKEKDDVSERGYVILPNDYFEIVEVPA\n\n### Answer:", "output": "Does not affect DNA-binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01561", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMDCG_KLEP3\n\n### Answer:", "output": "Phosphoribosyl-dephospho-CoA transferase, Malonate decarboxylase holo-[acyl-carrier-protein] synthase, Holo-ACP synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01562", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMIDQLQGTWKSISCENSEDYMKELGIGRASRKLGRLAKPTVTISTDGDVITIKTKSIFKNNEISFKLGEEFEEITPGGHKTKSKVTLDKESLIQVQDWDGKETTITRKLVDGKMVVESTVNSVICTRTYEKVSSNSVSNS\n\n### Sequence 2:\nMVDQLQGTWKSVSCENFENYMKELGAGRAIRKLGCLARPVVTISTDGDRITIKTKSIFKNKEISFKLGEEFEEITPGGRKSKSTVVLDNDSLVQVQDWDGKEATIRRRLVDGKMVVESAVNNVTCTRTYQRV\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01563", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVSEARVLEDRPLSDKGSGDSSQVTQVSPQRIALRLRPDDSKNFSIQVRQVEDYPVDIYYLMDLSYSMKDDLWSIQNLGTKLATQMRKLTSNLRIGFGAFVDKPVSPYMYISPPEALENPCYDMKTTCLPMFGYKHVLTLTDQVTRFNEEVKKQSVSRNRDAPEGGFDAIMQATVCDEKIGWRNDASHLLVFTTDAKTHIALDGRLAGIVQPNDGQCHVGSDNHYSASTTMDYPSLGLMTEKLSQKNINLIFAVTENVVNLYQNYSELIPGTTVGVLSMDSSNVLQLIVDAYGKIRSKVELEVRDLPEELSLSFNATCLNNEVIPGLKSCMGLKIGDTVSFSIEAKVRGCPQEKEKSFTIKPVGFKDSLIVQVTFD\n\n### Answer:", "output": "<SEQ_3Di>DKDKAQDAFDDADAAQDDDHPPHFQKGPQEIETEDAFQDKDKDKMKGAQHFQFEAEEEEEEEQAPLQQLQLVQQLPPLVLLCVLCVRRHPHYWYFYKYAFAQLFPPFQDDDDPVCQQAVCVVVVDGFGGHAGMFGFGAGDNPSVVSNVRSVPDHGHHTDDFAGNQQVRLLQCLQVCVVNHDDQRHFAEYEREDAGHHDEFPPCVNRVANAEDPSHHQQDPRRGRPCRNHYTRYHLVSSLVSCQVRLHQYEYAYAPVCQVVVVVSCVVHPNYDYHHADSRNNCVSVSVVVSVFQSQQKWAKDKFPDDPQKDKWKQKQFPPPPRRGRGRMHGRHGNRGMIMIIMMITGNGADPDQWDWMWMHTRRHPRIYIYTYGYD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01564", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "IRF7, LILRA4, TCF4", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as IRF7, TCF4, LILRA4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "spleen", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IRF7", "LILRA4", "TCF4"]}}
{"id": "eval_01565", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKFLLLFLVVLPIMGVFGKKNGYAVDSSGKAPECLLSNYCNNECTKVHYADKGYCCLLSCYCFGLNDDKKVLEISDTRKSYCDTTIIN\n\n### Sequence 2:\nMKFLLLFLVVLPIMGVLGKKNGYAVDSSGKAPECLLSNYCYNECTKVHYADKGYCCLLSCYCFGLNDDKKVLEISDTRKSYCDTPIIN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01566", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: HIV protease\nLigand SMILES: CCCCCCCCSCC1OC2OC3C(CSCCCCCCCC)OC(OC4C(CSCCCCCCCC)OC(OC5C(CSCCCCCCCC)OC(OC6C(CSCCCCCCCC)OC(OC7C(CSCCCCCCCC)OC(OC8C(CSCCCCCCCC)OC(OC1C(OS(=O)(=O)O)C2OS(=O)(=O)O)C(OS(=O)(=O)O)C8OS(=O)(=O)O)C(OS(=O)(=O)O)C7OS(=O)(=O)O)C(OS(=O)(=O)O)C6OS(=O)(=O)O)C(OS(=O)(=O)O)C5OS(=O)(=O)O)C(OS(=O)(=O)O)C4OS(=O)(=O)O)C(OS(=O)(=O)O)C3OS(=O)(=O)O\nEvidence: MoleculeNet HIV reports anti-HIV activity.", "output": "The ligand is relevant to HIV-1 protease (HIV protease). MoleculeNet HIV reports anti-HIV activity. The relationship is best described as target-associated activity evidence rather than an unsupported binding prediction.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "CCCCCCCCSCC1OC2OC3C(CSCCCCCCCC)OC(OC4C(CSCCCCCCCC)OC(OC5C(CSCCCCCCCC)OC(OC6C(CSCCCCCCCC)OC(OC7C(CSCCCCCCCC)OC(OC8C(CSCCCCCCCC)OC(OC1C(OS(=O)(=O)O)C2OS(=O)(=O)O)C(OS(=O)(=O)O)C8OS(=O)(=O)O)C(OS(=O)(=O)O)C7OS(=O)(=O)O)C(OS(=O)(=O)O)C6OS(=O)(=O)O)C(OS(=O)(=O)O)C5OS(=O)(=O)O)C(OS(=O)(=O)O)C4OS(=O)(=O)O)C(OS(=O)(=O)O)C3OS(=O)(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 3153.87, "LogP": 11.06, "TPSA": 1019.62, "QED": 0.02, "functional_groups": ["hydroxyl", "ether"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "HIV protease", "has_evidence": true}}
{"id": "eval_01567", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: CN(C)[C@@H]1C(=O)/C(=C(/O)NCN2CCCC(C(=O)O)C2)C(=O)[C@@]2(O)C(=O)C3=C(O)c4c(O)cccc4[C@@](C)(O)[C@H]3C[C@@H]12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "CN(C)[C@@H]1C(=O)/C(=C(/O)NCN2CCCC(C(=O)O)C2)C(=O)[C@@]2(O)C(=O)C3=C(O)c4c(O)cccc4[C@@](C)(O)[C@H]3C[C@@H]12", "compound_name": null, "pubchem_cid": null, "MW": 585.61, "LogP": 0.01, "TPSA": 208.17, "QED": 0.11, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "ketone", "phenol"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01568", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nKRT18 CDH1 KRT8 EPCAM\n</cell>", "output": "This cell is most consistent with epithelial cell, an epithelial cell. Key markers such as CDH1, KRT8, EPCAM support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "epithelial cell", "tissue": "lymph node", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["KRT18", "CDH1", "KRT8", "EPCAM"]}}
{"id": "eval_01569", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRS5_AGRVS\n\n### Answer:", "output": "30S ribosomal protein S5.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01570", "category": "Mutation", "instruction": "Describe the effect of mutation P24772-Y35E on protein Protein OPG200.", "input": "### Wild-type:\nMTANFSTHVFSPQHCGCDRLTSIDDVRQCLTEYIYWSSYAYRNRQCAGQLYSTLLSFRDDAELVFIDIRELVKNMPWDDVKDCAEIIRCYIPDEQKTIREISAIIGLCAYAATYWGGEDHPTSNSLNALFVMLEMLNYVDYNIIFRRMN\n\n### Mutant:\nMTANFSTHVFSPQHCGCDRLTSIDDVRQCLTEYIEWSSYAYRNRQCAGQLYSTLLSFRDDAELVFIDIRELVKNMPWDDVKDCAEIIRCYIPDEQKTIREISAIIGLCAYAATYWGGEDHPTSNSLNALFVMLEMLNYVDYNIIFRRMN\n\n### Answer:", "output": "Disrupts self-association. The mutation in the B14 protein results in a monomeric form of the protein. However, it still co-immunoprecipitates with IKKβ and blocks both NF-κB nuclear translocation and NF-κB-dependent gene expression.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01571", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "Cc1ccc(NC(=O)c2cccc(N(C)C)c2)cc1NC(=O)c1ccc(OC(C)C)cc1", "output": "The molecule contains the following functional groups: aromatic ring, amide, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "Cc1ccc(NC(=O)c2cccc(N(C)C)c2)cc1NC(=O)c1ccc(OC(C)C)cc1", "compound_name": null, "pubchem_cid": null, "MW": 431.54, "LogP": 5.35, "TPSA": 70.67, "QED": 0.53, "functional_groups": ["aromatic ring", "amide", "ether"], "moleculenet_labels": {"label": 3.1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "ether"]}}
{"id": "eval_01572", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "TP63, KRT14, KRT5", "output": "This cell is most consistent with basal epithelial cell, a basal epithelial cell. Key markers such as KRT5, TP63, KRT14 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "basal epithelial cell", "tissue": "lung", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TP63", "KRT14", "KRT5"]}}
{"id": "eval_01573", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGSCKGRCFELQCRCDNLCKSYTSCCHDFDELCLKTARGWECTKDRCGEVRNEENACHCSEDCLARGDCCTNYQVVCKGESHWVDDDCEEIKAAECPAGFVRPPLIIFSVDGFRASYMKKGSKVMPNIEKLRSCGTHSPYMRPVYPTKTFPNLYTLATGLYPESHGIVGNSMYDPVFDATFHLRGREKFNHRWWGGQPLWITATKQGVKAGTFFWSVVIPHERRILTILQWLTLPDHERPSVYAFYSEQPDFSGHKYGPFGPEMTNPLREIDKIVGQLMDGLKQLKLHRCVNVIFVGDHGMEDVTCDRTEFLSNYLTNVDDITLVPGTLGRIRSKFSNNAKYDPKAIIAALTCKKPDQHFKPYLKQHLPKRLHYANNRRIEDIHLLVERRWHVARKPFFQGDHGFDNKVNSMQTVFVGYGSTFKYKTKVPPFENIELYNVMCDLLGLKPAPNNGTHGSLNHLLRTNTFRPTMPEEVTRPNYPGIMYLQSDFDLGTEERHLLYGRPAVLYRTRYDILYHTDFESGYSEIFLMPLWTSYTVSKQACVRPDVRVSPSFSQNCLAYKNDKQMSYGFLFPPYLSSSPEAKYDAFLVTNMVPMYPAFKRVWNYFQRVLVKKYASERNGVNVISGPIFDYDYDGLHDTEDKIKQYVEGSSIPVPTHYYSIITSCLDFTQPADKCDGPLSVSSFILPHRPDNEESCNSSEDESKWVEELMKMHTARVRDIEHLTSLDFFRKTSRSYPEILTLKTYLHTYE\n\n### Answer:", "output": "<SEQ_3Di>DACVVPAQAFDCDHHQCCVVVVRPDPCNCQQQVAQPPHQEDALVQAQDDDDRNHRFHLHPVCLLLQTHRLCSCVHHVVDDFPLVDDQDDQQFQPDDPLPPALFEEEEEEEQFALVVCVVDCVQFSLVNCLQQWFEWESFAFFAFQLDPPQRVVLLFALAHCLAQQCQDCKFQQPVVRDIDHVPDDPVVPCVRGDGQTLCRLQVVVVQHCTPADGDPPRDLSSVLSVLSSLVSDDSVRRGRYYYDYDHPPVVVCFQAPQPGPVSSPRSSVVSVSSSSNLSSCVSNVRNQRHKYKYKYSWHKAWADPVQEAEVVVQDPPLVQWDWPKWQKTAIDGPDDPPPVDDVVVSQVSLFPPDPVDFKHKAQVLAAFLQNLNDNHSRRHRIMIHGHTRHMYDHDDRTTMDGRHDSLDSRTIHIIGIDHNFFHTSHYDYYFHSSQNQLVVCQSRVGDRDDHDHDRLLCNNGTPDGPDDDDDDDQPDDADAFDQADDQVQADACDCLFWAQQDDKGQQDDDYWHWDDADFWIFIFGLQQQETLKIKGKDALPFQAGEDRRDHNNSHHHPVLLVVFPFKDKDFPQHLLGDHDPVRSNVRRYSHRIAIAGPQVVLLVVCCNNPVQNVVCNVLRMKMKMKDAFADQSSSLAADDPVRQPDASPPDSNTHGQKMKMKIKAAPPSVDHRSPHADDMDIWMFIHGRDNHVSVFPPPVDRCVVTSVVNGLSNIHFLVSSCRRSVIGDSPGDNDRPSNVVNNRSDTDHPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01574", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCXA1_MACFA\n\n### Answer:", "output": "Belongs to the connexin family. Alpha-type (group II) subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01575", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: BACE1\nLigand SMILES: CC(=O)NC(Cc1cc(F)cc(F)c1)C(O)C[NH2+]C1(c2cccc(C(C)(C)C)c2)CCC(CO)CC1\nEvidence: MoleculeNet BACE reports activity against BACE1.", "output": "The ligand is relevant to beta-secretase (BACE1). MoleculeNet BACE reports activity against BACE1. The relationship is best described as target-associated activity evidence rather than an unsupported binding prediction.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BACE", "canonical_smiles": "CC(=O)NC(Cc1cc(F)cc(F)c1)C(O)C[NH2+]C1(c2cccc(C(C)(C)C)c2)CCC(CO)CC1", "compound_name": null, "pubchem_cid": null, "MW": 503.65, "LogP": 3.31, "TPSA": 86.17, "QED": 0.42, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "BACE1", "has_evidence": true}}
{"id": "eval_01576", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDAPB_COXB2\n\n### Answer:", "output": "Belongs to the DapB family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01577", "category": "Mutation", "instruction": "Describe the effect of mutation Q14749-N177H on protein Glycine N-methyltransferase.", "input": "### Wild-type:\nMVDSVYRTRSLGVAAEGLPDQYADGEAARVWQLYIGDTRSRTAEYKAWLLGLLRQHGCQRVLDVACGTGVDSIMLVEEGFSVTSVDASDKMLKYALKERWNRRHEPAFDKWVIEEANWMTLDKDVPQSAEGGFDAVICLGNSFAHLPDCKGDQSEHRLALKNIASMVRAGGLLVIDNRNYDHILSTGCAPPGKNIYYKSDLTKDVTTSVLIVNNKAHMVTLDYTVQVPGAGQDGSPGLSKFRLSYYPHCLASFTELLQAAFGGKCQHSVLGDFKPYKPGQTYIPCYFIHVLKRTD\n\n### Mutant:\nMVDSVYRTRSLGVAAEGLPDQYADGEAARVWQLYIGDTRSRTAEYKAWLLGLLRQHGCQRVLDVACGTGVDSIMLVEEGFSVTSVDASDKMLKYALKERWNRRHEPAFDKWVIEEANWMTLDKDVPQSAEGGFDAVICLGNSFAHLPDCKGDQSEHRLALKNIASMVRAGGLLVIDHRNYDHILSTGCAPPGKNIYYKSDLTKDVTTSVLIVNNKAHMVTLDYTVQVPGAGQDGSPGLSKFRLSYYPHCLASFTELLQAAFGGKCQHSVLGDFKPYKPGQTYIPCYFIHVLKRTD\n\n### Answer:", "output": "Significantly increases stability of the tetramer.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01578", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nDEFA5 LYZ DEFA6\n</cell>", "output": "This cell is most consistent with paneth cell, a paneth cell. Key markers such as DEFA6, DEFA5, LYZ support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "paneth cell", "tissue": "spleen", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["DEFA5", "LYZ", "DEFA6"]}}
{"id": "eval_01579", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCLPP_NITOC\n\n### Answer:", "output": "Cleaves peptides in various proteins in a process that requires ATP hydrolysis.  Has a chymotrypsin-like activity.  Plays a major role in the degradation of misfolded proteins.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01580", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nTNNI3 MYH6 MYH7 TNNT2\n</cell>", "output": "This cell is most consistent with cardiomyocyte, a cardiac muscle cell. Key markers such as TNNI3, MYH6, MYH7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "pancreas", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["TNNI3", "MYH6", "MYH7", "TNNT2"]}}
{"id": "eval_01581", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMSGDFAAAFLPTIFVPLVGLGLPAVLMSLLFTYIESEA\n\n### Answer:", "output": "<SEQ_3Di>DQDPDPPSCVCVVVVCCVVPVCVVVVVVVVVCVVVDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01582", "category": "Mutation", "instruction": "Describe the effect of mutation P64596-V83G on protein Outer membrane lipoprotein DolP.", "input": "### Wild-type:\nMKALSPIAVLISALLLQGCVAAAVVGTAAVGTKAATDPRSVGTQVDDGTLEVRVNSALSKDEQIKKEARINVTAYQGKVLLVVQSPNAELSARAKQIAMGVDGANEVYNEIRQGQPIGLGEASNDTWITTKVRSQLLTSDLVKSSNVKVTTENGEVFLMGLVTEREAKAAADIASRVSGVKRVTTAFTFIK\n\n### Mutant:\nMKALSPIAVLISALLLQGCVAAAVVGTAAVGTKAATDPRSVGTQVDDGTLEVRVNSALSKDEQIKKEARINVTAYQGKVLLVGQSPNAELSARAKQIAMGVDGANEVYNEIRQGQPIGLGEASNDTWITTKVRSQLLTSDLVKSSNVKVTTENGEVFLMGLVTEREAKAAADIASRVSGVKRVTTAFTFIK\n\n### Answer:", "output": "Gain of activity. Mutant can grow on 4.8% SDS; when associated with V-160.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01583", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRS3_BORT9\n\n### Answer:", "output": "Binds the lower part of the 30S subunit head.  Binds mRNA in the 70S ribosome, positioning it for translation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01584", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: PARP1\nLigand SMILES: CC(Oc1ccc(Oc2ncc(C(F)(F)F)cc2Cl)cc1)C(=O)O\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and poly(ADP-ribose) polymerase 1 (PARP1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CC(Oc1ccc(Oc2ncc(C(F)(F)F)cc2Cl)cc1)C(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 361.7, "LogP": 4.4, "TPSA": 68.65, "QED": 0.85, "functional_groups": ["aromatic ring", "hydroxyl", "ether", "halide", "pyridine"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": null, "NR-AhR": 0.0, "NR-Aromatase": 1.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": null, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "PARP1", "has_evidence": false}}
{"id": "eval_01585", "category": "Mutation", "instruction": "Describe the effect of mutation P07602-C388F on protein Prosaposin.", "input": "### Wild-type:\nMYALFLLASLLGAALAGPVLGLKECTRGSAVWCQNVKTASDCGAVKHCLQTVWNKPTVKSLPCDICKDVVTAAGDMLKDNATEEEILVYLEKTCDWLPKPNMSASCKEIVDSYLPVILDIIKGEMSRPGEVCSALNLCESLQKHLAELNHQKQLESNKIPELDMTEVVAPFMANIPLLLYPQDGPRSKPQPKDNGDVCQDCIQMVTDIQTAVRTNSTFVQALVEHVKEECDRLGPGMADICKNYISQYSEIAIQMMMHMQPKEICALVGFCDEVKEMPMQTLVPAKVASKNVIPALELVE\n\n### Mutant:\nMYALFLLASLLGAALAGPVLGLKECTRGSAVWCQNVKTASDCGAVKHCLQTVWNKPTVKSLPCDICKDVVTAAGDMLKDNATEEEILVYLEKTCDWLPKPNMSASCKEIVDSYLPVILDIIKGEMSRPGEVCSALNLCESLQKHLAELNHQKQLESNKIPELDMTEVVAPFMANIPLLLYPQDGPRSKPQPKDNGDVCQDCIQMVTDIQTAVRTNSTFVQALVEHVKEECDRLGPGMADICKNYISQYSEIAIQMMMHMQPKEICALVGFCDEVKEMPMQTLVPAKVASKNVIPALELVE\n\n### Answer:", "output": "In GDSAPC. A G1154----T transversion in the mRNA of the SAP-2 precursor results in the substitution of F for C385 in the mature SAP-2.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01586", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "TPSB2, TPSAB1, CPA3, HPGDS", "output": "This cell is most consistent with mast cell, a tissue mast cell. Key markers such as HPGDS, TPSB2, TPSAB1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "mast cell", "tissue": "heart", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TPSB2", "TPSAB1", "CPA3", "HPGDS"]}}
{"id": "eval_01587", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nMETN_STRPB\n\n### Answer:", "output": "Part of the ABC transporter complex MetNIQ involved in methionine import.  Responsible for energy coupling to the transport system.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01588", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "NC(=O)c1cccc(OC2CCN(Cc3ccccc3)CC2)c1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "NC(=O)c1cccc(OC2CCN(Cc3ccccc3)CC2)c1", "compound_name": null, "pubchem_cid": null, "MW": 310.4, "LogP": 2.83, "TPSA": 55.56, "QED": 0.92, "functional_groups": ["aromatic ring", "amide", "amine", "ether"], "moleculenet_labels": {"label": 1.92}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "ether"]}}
{"id": "eval_01589", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "LST1, CDKN1C, MS4A7", "output": "This profile is most consistent with CD16+ monocyte, which is commonly found in bone marrow under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "bone marrow", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["LST1", "CDKN1C", "MS4A7"]}}
{"id": "eval_01590", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nASTB_HAHCH\n\n### Answer:", "output": "Catalyzes the hydrolysis of N(2)-succinylarginine into N(2)-succinylornithine, ammonia and CO(2).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01591", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nURE1_CHESB\n\n### Answer:", "output": "Urease subunit alpha, Urea amidohydrolase subunit alpha.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01592", "category": "Mutation", "instruction": "Describe the effect of mutation P0AAM3-S2C on protein Hydrogenase maturation factor HypC.", "input": "### Wild-type:\nMSIGVPGQIRTIDGNQAKVDVCGIQRDVDLTLVGSCDENGQPRVGQWVLVHVGFAMSVINEAEARDTLDALQNMFDVEPDVGALLYGEEK\n\n### Mutant:\nMCIGVPGQIRTIDGNQAKVDVCGIQRDVDLTLVGSCDENGQPRVGQWVLVHVGFAMSVINEAEARDTLDALQNMFDVEPDVGALLYGEEK\n\n### Answer:", "output": "Decreased hydrogenase 3 activity. Can interact with pre-HycE or HypD. CP synthetase mutants (carAB) are able to generate active hydrogenases due to a block in enzyme maturation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01593", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDXR_METCA\n\n### Answer:", "output": "Catalyzes the NADPH-dependent rearrangement and reduction of 1-deoxy-D-xylulose-5-phosphate (DXP) to 2-C-methyl-D-erythritol 4-phosphate (MEP).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01594", "category": "Mutation", "instruction": "Describe the effect of mutation P53193-A48H on protein J-type co-chaperone JAC1, mitochondrial.", "input": "### Wild-type:\nMLKYLVQRRFTSTFYELFPKTFPKKLPIWTIDQSRLRKEYRQLQAQHAPDMAQQGSEQSSTLNQAYHTLKDPLRRSQYMLKLLRNIDLTQEQTSNEVTTSDPQLLLKVLDIHDELSQMDDEAGVKLLEKQNKERIQDIEAQLGQCYNDKDYAAAVKLTVELKYWYNLAKAFKDWAPGKQLEMNH\n\n### Mutant:\nMLKYLVQRRFTSTFYELFPKTFPKKLPIWTIDQSRLRKEYRQLQAQHHPDMAQQGSEQSSTLNQAYHTLKDPLRRSQYMLKLLRNIDLTQEQTSNEVTTSDPQLLLKVLDIHDELSQMDDEAGVKLLEKQNKERIQDIEAQLGQCYNDKDYAAAVKLTVELKYWYNLAKAFKDWAPGKQLEMNH\n\n### Answer:", "output": "Increases JAC1 ATPase activator activity. Increases vegetative cell population growth. Resistant to high temperature. Enables JAC1 ATPase activator activity; when associated with A-20; A-35; S-38 and L-41. It is viable in S. pombe, but not in S. cerevisiae. The in vivo similarities correlated with in vitro biochemical measurements.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01595", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGRFQGPKCIQCGQYLDDPDLKYGQHPPDAVDEPQMLTNELPQHKLTCFSVYCKHGHLCPIDTGLIEKNIELFFSGSAKPIYDDDEGGVNGKNLGPINEWWITKALIGFSTSFAEYILMDPSPEYAPIFGLMQEKIYISKIVVEFLQSNSDSTYEDLINKIETTVPPSGLNLNRFTEDSLLRHAQFVVEQVESYDEAGDQPIFLTPCMRDLIKLAGVTLGQRRA\n\n### Answer:", "output": "<SEQ_3Di>DQVPADADPLARDGPPDPLWDFDDFDPQAEEQVVLVPPPCAEWAWEQKFKAFPVGTTAEDQPCSLVVVGFIFMWTFTDGHPDDDHPGGTHGRHARFNDWAKPVTWIWTDHPSHIYTYDYHNPSNLVRQLCDDLLSNLLRLLLNCLVPCLADAQVVSQVSQQPNDHHPSNVDDRDGPCSCLVNVLVSLVSLVVVCVPDHVRRCVGNHNVNSCVVSVDDPVSNVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01596", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nCXCR2:high S100A9:low CSF3R:high FCGR3B:low S100A8:medium\n</cell>", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as CSF3R, S100A9, CXCR2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "breast", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["CXCR2", "S100A9", "CSF3R", "FCGR3B", "S100A8"]}}
{"id": "eval_01597", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMNMG_SHEON\n\n### Answer:", "output": "tRNA uridine 5-carboxymethylaminomethyl modification enzyme MnmG, Glucose-inhibited division protein A.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01598", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSEKPEKIVLASGNAGKLKEFGNLFEPYGITVLPQSEFSIPECPEPYSTFVENALAKARHAAGHSGLPALADDSGICAAALNGAPGIHSARYAGSNPKSDTANNLKLAAELAGKADKSCCYVCVLVFVRHKDDPRPIIAEGVWHGQWNGVPAGENGFGYDPYFYLPGHGKTAAELDSEVKNRESHRAQALAELLRKLAL\n\n### Sequence 2:\nMIRLVLLDIEGTTLPISFVRDVMFPYAAKALPALMQDHTNPIVVGARADIVMEHPGQDPLKVCQDWMKADVKAAPLKTLQGLTWRQGFEDGTLQADLYPDVPPALKAWSKGGLRLAVYSSGSIPSQKLLYGHTAQGDLTPLFEDFFDLSTGGKKDAASYEKITAAVGLPADEILFLSDIGAELDAAQRAGLSVCQLVREQDGTVPHPGVPQAPDLNAVSTQFGLPVA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01599", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "HLA-DQA1, CLEC10A, CD1C, FCER1A", "output": "The expression pattern is consistent with conventional dendritic cell in a context associated with rheumatoid arthritis, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "spleen", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["HLA-DQA1", "CLEC10A", "CD1C", "FCER1A"]}}
{"id": "eval_01600", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTSVVIVVSVAVLLGVLVITDSSWPCVFNSTAKWSLQHIESVTLERPRWSRVKKPSSSKWRHLRTILQWWQERRKQAHLLDTWQKQAMDMDECLTRYRDAKHDDICVISE\n\n### Sequence 2:\nMPRSLKKGPFIDLHLLKKVEKAVESGDKKPLRTWSRRSTIFPNMIGLTIAVHNGRQHVPVFVSDEMVGHKLGEFAPTRTYRGHAADKKAKKK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01601", "category": "Mutation", "instruction": "Describe the effect of mutation Q8ZMM8-Y346A on protein Secreted effector protein PipB2.", "input": "### Wild-type:\nMERSLDSLAGMAKSAFGAGTSAAMRQATSPKTILEYIINFFTCGGIRRRNETQYQELIETMAETLKSTMPDRGAPLPENIILDDMDGCRVEFNLPGENNEAGQVIVRVSKGDHSETREIPLASFEKICRALLFRCEFSLPQDSVILTAQGGMNLKGAVLTGANLTSENLCDADLSGANLEGAVLFMADCEGANFKGANLSGTSLGDSNFKNACLEDSIMCGATLDHANLTGANLQHASLLGCSMIECNCSGANMDHTNLSGATLIRADMSGATLQGATIMAAIMEGAVLTRANLRKASFI\n\n### Mutant:\nMERSLDSLAGMAKSAFGAGTSAAMRQATSPKTILEYIINFFTCGGIRRRNETQYQELIETMAETLKSTMPDRGAPLPENIILDDMDGCRVEFNLPGENNEAGQVIVRVSKGDHSETREIPLASFEKICRALLFRCEFSLPQDSVILTAQGGMNLKGAVLTGANLTSENLCDADLSGANLEGAVLFMADCEGANFKGANLSGTSLGDSNFKNACLEDSIMCGATLDHANLTGANLQHASLLGCSMIECNCSGANMDHTNLSGATLIRADMSGATLQGATIMAAIMEGAVLTRANLRKASFI\n\n### Answer:", "output": "No effect. Deletion of the Y346 residue in PipB2 leads to a loss of its ability to reorganize LE/K compartments and induce the peripheral accumulation of LE/K compartments.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01602", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "Oc1ccc(C2CNCCc3c2cc(O)c(O)c3Cl)cc1", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, amine, hydroxyl, halide, phenol. Its descriptor profile: MW=305.76, LogP=2.73, TPSA=72.72, HBD=4, HBA=4, QED=0.61.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "Oc1ccc(C2CNCCc3c2cc(O)c(O)c3Cl)cc1", "compound_name": null, "pubchem_cid": null, "MW": 305.76, "LogP": 2.73, "TPSA": 72.72, "QED": 0.61, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "halide", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 1.0, "NR-ER-LBD": 1.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01603", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nCYSN_SALAR\n\n### Answer:", "output": "Belongs to the TRAFAC class translation factor GTPase superfamily. Classic translation factor GTPase family. CysN/NodQ subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01604", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nFMO3_HUMAN\n\n### Answer:", "output": "Essential hepatic enzyme that catalyzes the oxygenation of a wide variety of nitrogen- and sulfur-containing compounds including drugs as well as dietary compounds (PubMed:10759686, PubMed:30381441, PubMed:32156684).  Plays an important role in the metabolism of trimethylamine (TMA), via the production of trimethylamine N-oxide (TMAO) metabolite (PubMed:9776311).  TMA is generated by the action of gut microbiota using dietary precursors such as choline, choline containing compounds, betaine or L-carnitine.  By regulating TMAO concentration, FMO3 directly impacts both platelet responsiveness and rate of thrombus formation (PubMed:29981269).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01605", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nYEGH_ECOLI\n\n### Answer:", "output": "Cell membrane; Multi-pass membrane protein.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01606", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMRFTFTSRCLALFLLLNHPTPILPAFSNQTYPTIEPKPFLYVVGRKKMMDAQYKCYDRMQQLPAYQGEGPYCNRTWDGWLCWDDTPAGVLSYQFCPDYFPDFDPSEKVTKYCDEKGVWFKHPENNRTWSNYTMCNAFTPEKLKNAYVLYYLAIVGHSLSIFTLVISLGIFVFFRSLGCQRVTLHKNMFLTYILNSMIIIIHLVEVVPNGELVRRDPVSCKILHFFHQYMMACNYFWMLCEGIYLHTLIVVAVFTEKQRLRWYYLLGWGFPLVPTTIHAITRAVYFNDNCWLSVETHLLYIIHGPVMAALVVNFFFLLNIVRVLVTKMRETHEAESHMYLKAVKATMILVPLLGIQFVVFPWRPSNKMLGKIYDYVMHSLIHFQGFFVATIYCFCNNEVQTTVKRQWAQFKIQWNQRWGRRPSNRSARAAAAAAEAGDIPIYICHQEPRNEPANNQGEESAEIIPLNIIEQESSA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCGGGCCHHHHHHHHHHHHHHHHHHHHHHHHHHCCCCCCCCCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTSCSCCCTTHHHHHHHHTTTTHHHHHHHHHHHHHHTCCSTTCCCCSGGGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCCCCCCCHHHHHHHHHHHHHHHHTCHHHHSCBCCSSHHHHHHHHHHHHHHHHTHHHHHHHHHHTSCHHHHHHHHHHHHHHHHTTCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01607", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: PARP1\nLigand SMILES: CC1(C)NC(=O)C(/C=C/c2ccccc2)O1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and poly(ADP-ribose) polymerase 1 (PARP1) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "CC1(C)NC(=O)C(/C=C/c2ccccc2)O1", "compound_name": null, "pubchem_cid": null, "MW": 217.27, "LogP": 1.95, "TPSA": 38.33, "QED": 0.82, "functional_groups": ["aromatic ring", "amide", "ether"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_protein_ligand", "target": "PARP1", "has_evidence": false}}
{"id": "eval_01608", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y3Q7-P170S on protein Disintegrin and metalloproteinase domain-containing protein 18.", "input": "### Wild-type:\nMFLLLALLTELGRLQAHEGSEGIFLHVTVPRKIKSNDSEVSERKMIYIITIDGQPYTLHLGKQSFLPQNFLVYTYNETGSLHSVSPYFMMHCHYQGYAAEFPNSFVTLSICSGLRGFLQFENISYGIEPVESSARFEHIIYQMKNNDPNVSILAVNYSHIWQKDQPYKVPLNSQIKNLSKLLPQYLEIYIIVEKALYDYMGSEMMAVTQKIVQVIGLVNTMFTQFKLTVILSSLELWSNENQISTSGDADDILQRFLAWKRDYLILRPHDIAYLLVYRKHPKYVGATFPGTVCNKSYDAG\n\n### Mutant:\nMFLLLALLTELGRLQAHEGSEGIFLHVTVPRKIKSNDSEVSERKMIYIITIDGQPYTLHLGKQSFLPQNFLVYTYNETGSLHSVSPYFMMHCHYQGYAAEFPNSFVTLSICSGLRGFLQFENISYGIEPVESSARFEHIIYQMKNNDPNVSILAVNYSHIWQKDQPYKVSLNSQIKNLSKLLPQYLEIYIIVEKALYDYMGSEMMAVTQKIVQVIGLVNTMFTQFKLTVILSSLELWSNENQISTSGDADDILQRFLAWKRDYLILRPHDIAYLLVYRKHPKYVGATFPGTVCNKSYDAG\n\n### Answer:", "output": "In a cutaneous metastatic melanoma sample; somatic mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01609", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPFKA_CLOBA\n\n### Answer:", "output": "Catalyzes the phosphorylation of D-fructose 6-phosphate to fructose 1,6-bisphosphate by ATP, the first committing step of glycolysis.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01610", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y3Z3-P593A on protein Deoxynucleoside triphosphate triphosphohydrolase SAMHD1.", "input": "### Wild-type:\nMQRADSEQPSKRPRCDDSPRTPSNTPSAEADWSPGLELHPDYKTWGPEQVCSFLRRGGFEEPVLLKNIRENEITGALLPCLDESRFENLGVSSLGERKKLLSYIQRLVQIHVDTMKVINDPIHGHIELHPLLVRIIDTPQFQRLRYIKQLGGGYYVFPGASHNRFEHSLGVGYLAGCLVHALGEKQPELQISERDVLCVQIAGLCHDLGHGPFSHMFDGRFIPLARPEVKWTHEQGSVMMFEHLINSNGIKPVMEQYGLIPEEDICFIKEQIVGPLESPVEDSLWPYKGRPENKSFLYEI\n\n### Mutant:\nMQRADSEQPSKRPRCDDSPRTPSNTPSAEADWSPGLELHPDYKTWGPEQVCSFLRRGGFEEPVLLKNIRENEITGALLPCLDESRFENLGVSSLGERKKLLSYIQRLVQIHVDTMKVINDPIHGHIELHPLLVRIIDTPQFQRLRYIKQLGGGYYVFPGASHNRFEHSLGVGYLAGCLVHALGEKQPELQISERDVLCVQIAGLCHDLGHGPFSHMFDGRFIPLARPEVKWTHEQGSVMMFEHLINSNGIKPVMEQYGLIPEEDICFIKEQIVGPLESPVEDSLWPYKGRPENKSFLYEI\n\n### Answer:", "output": "Promotes ability to restrict infection by viruses.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01611", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSFT2C_HUMAN\n\n### Answer:", "output": "May be involved in fusion of retrograde transport vesicles derived from an endocytic compartment with the Golgi complex.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01612", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSKESIYGLTMAQLTDWLMERGHKKFRATQVWDWLYRKRVTTFAEMTNVNKDCLQLLEDHFAIETMSEHVRQESKDGTIKFLFRLQDGNLIETVLMRHKYGFSVCVTTQVGCNIGCSFCASGLLTKNRDLSSGEIVEQIMKVQFHLDQVGKEERVSHVVVMGIGEPFDNFQNTVDFLEIIKDHKGLAIGARHITVSTSGLAHKIYEFADLKLQVNLAVSLHAPNNELRSRIMKINKAFPIEKLMDSINYYIEKTNRRVTYEYILIKDVNDHKEEALQLAELIGDKRHLSYVNLIPYNPVDEHSQYQRSEPEAISQFFDTLKKKGINCGVRLEHGTDIDAACGQLRSKQEKKKVKVN\n\n### Sequence 2:\nMEKSSIYGLTWTNLTEWLEAHGQKKFRATQVWDWLYRKRVKTFEEMSNVPKETIELLTANFVMNTLEEQVVQESTDGTTKYLFKLSDGNLIETVMMKQEYGLSVCVTTQVGCNIGCTFCASGLLKKSRDLTAGEIVEQIMNVQHYLDGRNLEERVSHVVVMGIGEPFDNYDNVMDFLRVINHDKGLAIGARHITVSTSGLAPRIIDFANEDFQVNLAISLHAPNNELRTSIMRINKTYSIEKLMEAIHYYVNKTNRRITFEYIMLKGVNDHKKEALELAALLGEHRHLAYVNLIPYNPVDEHIDYERSTKEDVLAFYDTLKKNGINCVIRREHGTDIDAACGQLRSKQIKR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01613", "category": "Mutation", "instruction": "Describe the effect of mutation Q925D8-Q371R on protein Taste receptor type 1 member 3.", "input": "### Wild-type:\nMPALAIMGLSLAAFLELGMGASLCLSQQFKAQGDYILGGLFPLGSTEEATLNQRTQPNSIPCNRFSPLGLFLAMAMKMAVEEINNGSALLPGLRLGYDLFDTCSEPVVTMKSSLMFLAKVGSQSIAAYCNYTQYQPRVLAVIGPHSSELALITGKFFSFFLMPQVSYSASMDRLSDRETFPSFFRTVPSDRVQLQAVVTLLQNFSWNWVAALGSDDDYGREGLSIFSSLANARGICIAHEGLVPQHDTSGQQLGKVLDVLRQVNQSKVQVVVLFASARAVYSLFSYSIHHGLSPKVWVAS\n\n### Mutant:\nMPALAIMGLSLAAFLELGMGASLCLSQQFKAQGDYILGGLFPLGSTEEATLNQRTQPNSIPCNRFSPLGLFLAMAMKMAVEEINNGSALLPGLRLGYDLFDTCSEPVVTMKSSLMFLAKVGSQSIAAYCNYTQYQPRVLAVIGPHSSELALITGKFFSFFLMPQVSYSASMDRLSDRETFPSFFRTVPSDRVQLQAVVTLLQNFSWNWVAALGSDDDYGREGLSIFSSLANARGICIAHEGLVPQHDTSGQQLGKVLDVLRQVNQSKVQVVVLFASARAVYSLFSYSIHHGLSPKVWVAS\n\n### Answer:", "output": "Altered sweet preference in mice has not been associated with the mutation in T1R3 in strain: 129/J, 129/SvEv, AKR/J, BALB/c, CAST/Ei, C3H/HeJ, DBA/2J, FVB/N, ST/bJ, and SWR/J.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01614", "category": "Mutation", "instruction": "Describe the effect of mutation Q9U9Y8-E402K on protein Serine/threonine kinase NLK.", "input": "### Wild-type:\nMVSWGRGKDAYYLYISREQEEDDDDSLSFYSSQKDTEDEFCGCLFPDPEIPGSSSSSGCSSSSTELYDLAAAHAALISRQQQILSQAIPIIPEHQLAAVAAHHQHHQQLHPSVQYQLVAAATHHNHHQPQAAQPHYSAVVPRSDVIQQPPHFALHHHLQNLVQQQQQQQAHHHHQQLVGEMALVSHTHPAAVGSTTCYEKNQQKQQQVQQIPTQPQVAHVSSNAILAAAQPFYPPPVQDSQPDRPIGYGAFGVVWSVTDPRSGKRVALKKMPNVFQNLASCKRVFREIKMLSSFRHDNVL\n\n### Mutant:\nMVSWGRGKDAYYLYISREQEEDDDDSLSFYSSQKDTEDEFCGCLFPDPEIPGSSSSSGCSSSSTELYDLAAAHAALISRQQQILSQAIPIIPEHQLAAVAAHHQHHQQLHPSVQYQLVAAATHHNHHQPQAAQPHYSAVVPRSDVIQQPPHFALHHHLQNLVQQQQQQQAHHHHQQLVGEMALVSHTHPAAVGSTTCYEKNQQKQQQVQQIPTQPQVAHVSSNAILAAAQPFYPPPVQDSQPDRPIGYGAFGVVWSVTDPRSGKRVALKKMPNVFQNLASCKRVFREIKMLSSFRHDNVL\n\n### Answer:", "output": "In t1534; reduced body-wall muscle.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01615", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMANIKASKKDALTSEKRRKKNSSRRSMIRTFVKKVRVAIMSGNKTTAEDAFKKMQPIIDSHVNKGLIHKNKAARYKSNLSLQIKKISKI\n\n### Sequence 2:\nMANNNSSNNNELLVYGAEQAIDQMKYEIASEFGVNLGADTTARANGSVGGEITKRLVQLAEQQLGGGRF\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01616", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nS17A9_HUMAN\n\n### Answer:", "output": "Involved in vesicular storage and exocytosis of ATP.  May accumulate ATP and other nucleotides in secretory vesicles such as adrenal chromaffin granules and synaptic vesicles.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01617", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "ITGA2B, GP1BA, PF4, PPBP", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as PPBP, ITGA2B, GP1BA support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "skin", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["ITGA2B", "GP1BA", "PF4", "PPBP"]}}
{"id": "eval_01618", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "COc1cc(C2(c3cccc(-c4cncnc4)c3)N=C(N)c3c(F)cccc32)ccn1", "output": "The molecule contains the following functional groups: aromatic ring, amine, ether, halide, pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BACE", "canonical_smiles": "COc1cc(C2(c3cccc(-c4cncnc4)c3)N=C(N)c3c(F)cccc32)ccn1", "compound_name": null, "pubchem_cid": null, "MW": 411.44, "LogP": 3.7, "TPSA": 86.28, "QED": 0.55, "functional_groups": ["aromatic ring", "amine", "ether", "halide", "pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amine", "ether", "halide", "pyridine"]}}
{"id": "eval_01619", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "SATB2, RBFOX3, SLC17A7", "output": "This cell is most consistent with excitatory neuron, an excitatory (glutamatergic) neuron. Key markers such as RBFOX3, SATB2, SLC17A7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "colon", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SATB2", "RBFOX3", "SLC17A7"]}}
{"id": "eval_01620", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTENRQLGALLAACHWIGEKGWCPATGGNMSLRLDLAHCLITESGKDKGSLAAEDFLLVETANNHVPSGRTPSAETGLHTLLYRLYPEIQAVLHTHSVNATVLSRVERSNALVLQGYEMQKSLSGQRSHLDAVVIPIFDNDQDIPALAQRVAAYADNRPLQYGFLVRGHGLYCWGNSVVEARRHLEGLEFLFQCELQRRLFDVNSNVDVKPNVDVNPNVEAK\n\n### Sequence 2:\nMTENQQLTALLAACHWIGEKGWCPATGGNMSVRLDDAQCLITESGKDKGSLQAEDFLLVEIATNHVPSGRTPSAETGLHTLLYRREPTIGAVLHTHSVNATVLSRVEKGAELVLHGYEMQKSLAGQTTHLDRVAIPIFDNDQDIPALAQRVTEYASHTPLRYGFLVRGHGLYCWGATVKEARRHLEGLEFLFQCELQRRLLEAKA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01621", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDCTA3_BRADU\n\n### Answer:", "output": "Responsible for the transport of dicarboxylates such as succinate, fumarate, and malate from the periplasm across the membrane.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01622", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CN1CCC23CCCCC2C1Cc1ccc(O)cc13", "output": "The SMILES represents an unnamed molecule with MW=257.38, LogP=3.08, TPSA=23.47. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "SIDER", "canonical_smiles": "CN1CCC23CCCCC2C1Cc1ccc(O)cc13", "compound_name": null, "pubchem_cid": null, "MW": 257.38, "LogP": 3.08, "TPSA": 23.47, "QED": 0.77, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "phenol"], "moleculenet_labels": {"Hepatobiliary disorders": 1, "Metabolism and nutrition disorders": 0, "Product issues": 0, "Eye disorders": 1, "Investigations": 0, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 1, "Immune system disorders": 0, "Reproductive system and breast disorders": 0, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 0, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 0, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 0, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01623", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nCSNB_ASPOR\n\n### Answer:", "output": "Endo-chitosanase B.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01624", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMADIEGLLHEVADAVISCKKEKVLEAVEKARAEVPPEEIIEKGLSAGMNQVGVLFERGKLFLPHVMMAADAMTTGVKLLEADLPAGAEKKLGVIVNGTVEGDVHDIGKSIVSTMLQSAGFEVHDIGRDVPIRDFIEKAKETDADMIGISALMTTTLQGQRDVIELLKEEGLRSRVKVMVGGAPATQAWADKIGADCYAENASEAVAKAKELLL\n\n### Sequence 2:\nMASKEELLQELSDAIVSCKKDRVIAAVEKAKEVMEPAEIIEKGLAAGMNQVGTLFERGKLFLPHVMMAADSMTAGVNILEAEMPAGTETKKLGVIVNGTVEGDVHDIGKSIVSTMLQSAGFEVHDIGRDVPIKNFVEKAKEVNADMIGLSALMTTTMQGQRDVIELLKEEGMRERVKVMVGGAPATQAWADKIGADCYAENASEAVAKAKELLL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01625", "category": "Mutation", "instruction": "Describe the effect of mutation Q8W033-C310S on protein Aldehyde dehydrogenase family 3 member I1, chloroplastic.", "input": "### Wild-type:\nMTKLLEINHIQTLCFAKGFSPARLNVATSPFRISRRGGGGYCSNACIPYRLKFTCYATLSAVVKEQASDFSGKEAALLVDELRSNFNSGRTKSYEWRISQLQNIARMIDEKEKCITEALYQDLSKPELEAFLAEISNTKSSCMLAIKELKNWMAPETVKTSVTTFPSSAQIVSEPLGVVLVISAWNFPFLLSVEPVIGAIAAGNAVVLKPSEIAPAASSLLAKLFSEYLDNTTIRVIEGGVPETTALLDQKWDKIFFTGGARVARIIMAAAARNLTPVVLELGGKCPALVDSDVNLQVAA\n\n### Mutant:\nMTKLLEINHIQTLCFAKGFSPARLNVATSPFRISRRGGGGYCSNACIPYRLKFTCYATLSAVVKEQASDFSGKEAALLVDELRSNFNSGRTKSYEWRISQLQNIARMIDEKEKCITEALYQDLSKPELEAFLAEISNTKSSCMLAIKELKNWMAPETVKTSVTTFPSSAQIVSEPLGVVLVISAWNFPFLLSVEPVIGAIAAGNAVVLKPSEIAPAASSLLAKLFSEYLDNTTIRVIEGGVPETTALLDQKWDKIFFTGGARVARIIMAAAARNLTPVVLELGGKCPALVDSDVNLQVAA\n\n### Answer:", "output": "No effect on solubility, but no effect on activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01626", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "KIT, PROM1, CD34", "output": "The expression pattern is consistent with hematopoietic stem cell in a context associated with COVID-19, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "hematopoietic stem cell", "tissue": "colon", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["KIT", "PROM1", "CD34"]}}
{"id": "eval_01627", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTIKPVSVKLDADIKARVEHLAETRKRSSHWMMREAIREYVEREEKREAL\n\n### Answer:", "output": "<SEQ_3Di>DDDDDDDDDDPVVVVVLVVVCVVVVHDSVVVVVVVVVVVCVVVCVVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01628", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQGHMPSDSKKPTIIYPCLWDYRVIMTTKDTSTLKELLETYQRPFKLEFKNTSKNAKFYSFNVSMEVSNESERNEIFQKISQLDKVVQTLGS\n\n### Answer:", "output": "<SEQ_3Di>DDDDDCPPPDDDDDPFDKDKKKFKFQDPPDDLLVVLVCVVVFDKDKAWDDAPPVNRITIIMIIGTDRDVVCLVVSQVSSVVPVRTPDIDSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01629", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "INS, MAFA, NKX6-1, IAPP", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as IAPP, NKX6-1, MAFA support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "heart", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["INS", "MAFA", "NKX6-1", "IAPP"]}}
{"id": "eval_01630", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: O=C(NCCCCCCCCNC(=O)c1cnc2ccccc2n1)c1cnc2ccccc2n1\nDescriptors: MW=456.55, LogP=4.07, TPSA=109.76, HBD=2, HBA=6", "output": "Descriptors: MW=456.55, LogP=4.07, TPSA=109.76, HBD=2, HBA=6, QED=0.33, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "O=C(NCCCCCCCCNC(=O)c1cnc2ccccc2n1)c1cnc2ccccc2n1", "compound_name": null, "pubchem_cid": null, "MW": 456.55, "LogP": 4.07, "TPSA": 109.76, "QED": 0.33, "functional_groups": ["aromatic ring", "amide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01631", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nFNLPPGNYKKPKLLYCSNGGHFLRILPDGTVDGTRDRSDPHIQLQLIAESVGEVYIKSTETGQYLAMDTDGLLYGSQTPNEECLFLERLEENGYNTYISKKHAEKNWFVGLKKNGSCKRGPRTHYGQKAILFLPLPVSSD\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCEEEEETTTTEEEEECTTSCEEEECCTTCGGGCEEEEESSTTEEEEEETTTCCEEEECTTSCEEEESSCCGGGCEEEEECTTSCEEEEESTTGGGTCBCCBCTTSBBCCGGGCCTTCGGGCEEEEECCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01632", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nUVRB_STAAW\n\n### Answer:", "output": "The UvrABC repair system catalyzes the recognition and processing of DNA lesions.  A damage recognition complex composed of 2 UvrA and 2 UvrB subunits scans DNA for abnormalities.  Upon binding of the UvrA(2)B(2) complex to a putative damaged site, the DNA wraps around one UvrB monomer.  DNA wrap is dependent on ATP binding by UvrB and probably causes local melting of the DNA helix, facilitating insertion of UvrB beta-hairpin between the DNA strands.  Then UvrB probes one DNA strand for the presence of a lesion.  If a lesion is found the UvrA subunits dissociate and the UvrB-DNA preincision complex is formed.  This complex is subsequently bound by UvrC and the second UvrB is released.  If no lesion is found, the DNA wraps around the other UvrB subunit that will check the other stand for damage.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01633", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nDEOB_SHEFN\n\n### Answer:", "output": "Belongs to the phosphopentomutase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01634", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMIQTERAVQQVLEWGRSLTGFADEHAVEAVRGGQYILQRIHPSLRGTSARTGRDPQDETLIVTFYRELALLFWLCDCNDLGLISPEQLAAVEQALGQGVPCALPGFEGCAVLRASLATLAYDRRDYAQLLDDTRCYSAALRAGHAQAERWSYAEYLHNGIDSIAYANVFCCLSLLWGLDMATLRARPAFRQVLRLISAIGRLQNDLHGADNAVILLLQRYPAMPVVEFLNDELAGHTRMLHRVMAEERFPAPWGPLIEAMAAI\n\n### Answer:", "output": "<SEQ_3Di>DDDQVVLLVLLVVLQQLAQQQDSVLSNVLSVVLVLLLVLQVVQLVVCCVVQVFDSPDSLLSSLLSNVSSVLVRLLCCLVVVRDDPVQNVQLLCCQLVLDQRPDGRHRSSNVSLVSLCVQFSDNVLSNVLSVLSNLLNVLSNVLSVAVVHDDLVVNLVSQLSVLCPSNSVSSVCRRSVQPVVVVVVDPLVVLLSSLLSSLLLLCCLLPDSDGQQVRCCVVPVPDPSVVVSVVVNVVSVVVNVVSCVVDPGRPPVVSSSVSSNVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01635", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPUTP_STAA3\n\n### Answer:", "output": "Sodium/proline symporter, Proline permease.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01636", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVKIYAPASIGNVSVGFDVLGAAVSPVDGTLLGDCVSVTASDSFSLRNEGRFVSKLPDDPKENIVYQCWERFCQEMGKEIPVALVLEKNMPIGSGLGSSACSVVAGLMAMNEFCGKPLDKVTLLGMMGELEGRVSGSVHFDNVAPCYLGGMQLILEQPGYISQDVPGFDDWLWVMAYPGIKVSTAEARAILPAQYRRQDCIAHGRNLAGFIHACHTQQPSLAAKMMKDVIAEPYRTQLLPGFAAARQAAQDIGALACGISGSGPTLFAVCNDSETAQRMASWLQNHYLQNDEGFVHICRLDTAGARLLG\n\n### Sequence 2:\nMVKVYAPASIGNVSVGFDVLGAAVSPVDGSLLGDCVSVEAADLFSLRNEGRFVSKLPDNPKENIVYQCWELFCQEIGKTVPVAMTLEKNMPIGSGLGSSACSVVAGLMAMNEFCGKPLDDTRLLTLMGELEGRISGSVHYDNVAPCFLGGVQLMLEENGIISQPVPSFDDWLWVMAYPGIKVSTAEARAILPAQYRRQDCISHGRYLAGFIHACHTGQAALAAKLMKDVIAEPYRTKLLPGFAAARQAAEDIGALACGISGSGPTLFSVCNDMTSAQRLADWLRDNYLQNDEGFVHICRLDTTGARQLG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01637", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRNP1_METMA\n\n### Answer:", "output": "Belongs to the eukaryotic/archaeal RNase P protein component 1 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01638", "category": "Mutation", "instruction": "Describe the effect of mutation B5A7L9-R318A on protein Glucosyltransferase 3.", "input": "### Wild-type:\nMRVYITNINGQSIQSTAQLCQNTVTDVAVSLGYRELGIYCYQIHTDSESELSKRLDGIVAGLRHGDVVIFQTPTWNTTEFDEKLMNKLKLYDIKIVLFIHDVVPLMFSGNFYLMDRTIAYYNKADVVVAPSQKMIDKLRDFGMNVSKTVVQGMWDHPTQAPMFPAGLKREIHFPGNPERFSFVKEWKYDIPLKVYTWQNVELPQNVHKINYRPDEQLLMEMSQGGFGLVWMDDKDKEYQSLYCSYKLGSFLAAGIPVIVQEGIANQELIENNGLGWIVKDVEEAIMKVKNVNEDEYIELV\n\n### Mutant:\nMRVYITNINGQSIQSTAQLCQNTVTDVAVSLGYRELGIYCYQIHTDSESELSKRLDGIVAGLRHGDVVIFQTPTWNTTEFDEKLMNKLKLYDIKIVLFIHDVVPLMFSGNFYLMDRTIAYYNKADVVVAPSQKMIDKLRDFGMNVSKTVVQGMWDHPTQAPMFPAGLKREIHFPGNPERFSFVKEWKYDIPLKVYTWQNVELPQNVHKINYRPDEQLLMEMSQGGFGLVWMDDKDKEYQSLYCSYKLGSFLAAGIPVIVQEGIANQELIENNGLGWIVKDVEEAIMKVKNVNEDEYIELV\n\n### Answer:", "output": "25% glycosyltransferase activity, the protein dimerizes but does not tetramerize, partially restores Fap1 glycosylation in vivo. The critical amino acid residue R318 is required for UDP- or UDP-glucose binding and for oligomerization of Gtf3. It is crucial for Fap1 glycosylation in S. parasanguinis in vivo.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01639", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPIELLPETPSQTAGPYVHIGLALEAAGNPTRDQEIWNRLAKPDAPGEHILLLGQVYDGNGHLVRDSFLEVWQADANGEYQDAYNLENAFNSFGRTATTFDAGEWTLHTVKPGVVNNAAGVPMAPHINISLFARGINIHLHTRLYFDDEAQANAKCPVLNLIEQPQRRETLIAKRCEVDGKTAYRFDIRIQGEGETVFFDF\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCTTTHHHHCHHHHTCCCCSSCCCSBCBCTTCSSCEEEEEECEECTTSCBCTTCEEEEECCCTTSCCCCBCCTTSSBCSEEEEECCSSSCCEEEEEECCCCCBCTTSCBCCSCEEEEEECTTCSSCEEEEEEEGGGHHHHHHCTTGGGSSCHHHHGGGEEEEEESSSCEEEECCEECSSTTCCCCEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01640", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMAMSKPILSTASVLAFERKLDPSDALMSAGAWAQRDASQEWPAVTVREKSVRGTISNRLKTKDRDPAKLDASIQSPNLQTVDVANLPSDADTLKVRFTLRVLGGAGTPSACNDAAYRDKLLQTVATYVNDQGFAELARRYAHNLANARFLWRNRVGAEAVEVRINHIRQGEVARAWRFDALAIGLRDFKADAELDALAELIASGLSGSGHVLLEVVAFARIGDGQEVFPSQELILDKGDKKGQKSKTLYSVRDAAAIHSQKIGNALRTIDTWYPDEDGLGPIAVEPYGSVTSQGKAYRQPKQKLDFYTLLDNWVLRDEAPAVEQQHYVIANLIRGGVFGEAEEK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCSCEEECCSSCBCCEEEEEECTTCCTTTTSCCBCCCCCBCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCBCCCBCCSSCCEEEEEECBBCCCCCSSCSBCCCSHHHHHHHHHHHHHHHHCCSHHHHHHHHHHTTSSCSSSSTTTTCSSCCEEECCCSSSSCSCCEECCSSSSCSSSCCCCHHHHHHHHHHHHHHSSSSCCBCCEEEEEECCSSCBCCCCBCCCCCCCCCCCCCCCCBCCBTTBCEECSHHHHHHTTCCBCCCSSCTTSCCBCCSTTCCCTTTTCCCSCTTTTCSHHHHHHHHHTSCCCCCHHHHHHHHHHHHSCEEECCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01641", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMKFFLDTANVEKIKEFNALGLVDGVTTNPSLIKKEGRDFYEVIKEICAIVDGPVSAEVIALDAEGMVKEARELVKLAENVVVKIPMTKEGMKAVNILSKEGIKTNVTLIFSANQALLAAKAGASYVSPFVGRLDDVGQDGMFLISEVMQVFSAYGIETEVIVASVRHPIHVIESAKMGADIATIPFDVLDKLFNHPLTDNGIEKFLADWEAHMNR\n\n### Sequence 2:\nMKFFLDTANVDKIKEFNALGLVDGVTTNPSLIKKEGRDFYEVIKEICSIVEGPVSAEVIALDAEGMVNEAKELVKIAKNVVIKIPMTKEGMKAVNILSKEGIKTNVTLIFSANQALLAAKAGATYVSPFVGRLDDIGQDGLLLISEIMQIFGAYGIETEVIVASVRHPIHVTESAKMGADIATIPFDVLDKLFNHSLTDIGIEKFLADWDAHMKR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01642", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nHPR_BACCZ\n\n### Answer:", "output": "Negative regulator of protease production and sporulation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01643", "category": "Mutation", "instruction": "Describe the effect of mutation P0A8H3-E152A on protein Zinc transporter ZupT.", "input": "### Wild-type:\nMSVPLILTILAGAATFIGAFLGVLGQKPSNRLLAFSLGFAAGIMLLISLMEMLPAALAAEGMSPVLGYGMFIFGLLGYFGLDRMLPHAHPQDLMQKSVQPLPKSIKRTAILLTLGISLHNFPEGIATFVTASSNLELGFGIALAVALHNIPEGLAVAGPVYAATGSKRTAILWAGISGLAEILGGVLAWLILGSMISPVVMAAIMAAVAGIMVALSVDELMPLAKEIDPNNNPSYGVLCGMSVMGFSLVLLQTAGIG\n\n### Mutant:\nMSVPLILTILAGAATFIGAFLGVLGQKPSNRLLAFSLGFAAGIMLLISLMEMLPAALAAEGMSPVLGYGMFIFGLLGYFGLDRMLPHAHPQDLMQKSVQPLPKSIKRTAILLTLGISLHNFPEGIATFVTASSNLELGFGIALAVALHNIPAGLAVAGPVYAATGSKRTAILWAGISGLAEILGGVLAWLILGSMISPVVMAAIMAAVAGIMVALSVDELMPLAKEIDPNNNPSYGVLCGMSVMGFSLVLLQTAGIG\n\n### Answer:", "output": "Binds Zn(2+) and Cd(2+) but with a 10-fold lower affinity. Retains binding of Fe(2+) at the high affinity M2 site, but loses Fe(2+) at the M1 site. Strong decrease in Zn(2+) uptake, but slight increase in Fe(2+) uptake. Cannot transport Co(2+) and Mn(2+). The mutation in the ZupT protein causes a change in its function.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01644", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTPIS_CARHZ\n\n### Answer:", "output": "Belongs to the triosephosphate isomerase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01645", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "TCF4, CLEC4C, GZMB, LILRA4", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as LILRA4, CLEC4C, GZMB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "bone marrow", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TCF4", "CLEC4C", "GZMB", "LILRA4"]}}
{"id": "eval_01646", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CN(C)c1cc(NC(=O)CNC(C)(C)C)c(O)c2c1CC1CC3C(N(C)C)C(=O)C(C(N)=O)=C(O)C3(O)C(=O)C1=C2O", "output": "This molecule is relatively large and polar, with hydrophilic character. It contains aromatic ring, amide, amine, hydroxyl, ketone. Its descriptor profile: MW=585.66, LogP=0.36, TPSA=205.76, HBD=7, HBA=11, QED=0.18.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "SIDER", "canonical_smiles": "CN(C)c1cc(NC(=O)CNC(C)(C)C)c(O)c2c1CC1CC3C(N(C)C)C(=O)C(C(N)=O)=C(O)C3(O)C(=O)C1=C2O", "compound_name": null, "pubchem_cid": null, "MW": 585.66, "LogP": 0.36, "TPSA": 205.76, "QED": 0.18, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "ketone", "phenol"], "moleculenet_labels": {"Hepatobiliary disorders": 1, "Metabolism and nutrition disorders": 1, "Product issues": 1, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 1, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01647", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMVMMLRTWRLLPMVLLAAYCYCVFGTCSIGTTTAPVEWKSPDRQIPKNITCANYSGTINGNVTFRGLQNKTEDFLHWLLGWGHKSICSFFPKLQGNYNEQHYRYEVANLTYNCTYNRLTLLNLTTENSGKYYFKREDANFTFYYSCYNLTVS\n\n### Sequence 2:\nMTETANLFVLGAAGGVEWGTVIVQVLTFIVLLALLKKFAWGPLKDVMDKRERDINRDIDDAEQAKLNAQKLEEENKQKLKETQEEVQKILEDAKVQARQQQEQIIHEANVRANGMIETAQSEINSQKERAIADINNQVSELSVLIASKVLRKEISEQDQKALVDKYLKEAGDK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01648", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "COc1cc2ncc(C(N)=O)c(Nc3cccc(Cl)c3)c2cc1OC", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, ether, halide, pyridine. Its descriptor profile: MW=357.8, LogP=3.75, TPSA=86.47, HBD=2, HBA=5, QED=0.73.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "COc1cc2ncc(C(N)=O)c(Nc3cccc(Cl)c3)c2cc1OC", "compound_name": null, "pubchem_cid": null, "MW": 357.8, "LogP": 3.75, "TPSA": 86.47, "QED": 0.73, "functional_groups": ["aromatic ring", "amide", "ether", "halide", "pyridine"], "moleculenet_labels": {"label": 3.5}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01649", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: CCn1ccnc1CC1COc2ccccc2O1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CCn1ccnc1CC1COc2ccccc2O1", "compound_name": null, "pubchem_cid": null, "MW": 244.29, "LogP": 2.29, "TPSA": 36.28, "QED": 0.83, "functional_groups": ["aromatic ring", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01650", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCBPN_HUMAN\n\n### Answer:", "output": "Protects the body from potent vasoactive and inflammatory peptides containing C-terminal Arg or Lys (such as kinins or anaphylatoxins) which are released into the circulation.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01651", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "CD34, KIT, PROM1", "output": "This profile is most consistent with hematopoietic stem cell, which is commonly found in pancreas under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "hematopoietic stem cell", "tissue": "pancreas", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CD34", "KIT", "PROM1"]}}
{"id": "eval_01652", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSNTQAERSIIGMIDMFHKYTRRDDKIDKPSLLTMMKENFPNFLSACDKKGTNYLADVFEKKDKNEDKKIDFSEFLSLLGDIATDYHKQSHGAAPCSGGSQ\n\n### Answer:", "output": "<SEQ_3Di>DDDPVVVVVVVVVVLQQVQDDPVSWHALVSVVVSCCVPVVVLCVQCVVVVHRCSPCLQVVLPPVPPRIGHPVSVVVVVVVQVVQVVVVVVPDRGPPDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01653", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nLPXB_BURP6\n\n### Answer:", "output": "Belongs to the LpxB family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01654", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAKYEILYIIRPNIEEEAKNALVARFDSILTDNGATVVESKTWEKRRLAYEIQDFREGLYHIVNVEANDDAALKEFDRLSKINADILRHMIVKIDA\n\n### Sequence 2:\nMAVKIRLKRMGAKKTPFYRVVVADSRSPRDGRFIEEIGTYNPVAQPAEVKINEEAALKWLGNGAKPSDTVRNLFSNQGIMEKFHLSKQGK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01655", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGHMGHTLERSVFSERTEESSAVQYFQFYGYLSQQQNMMQDYVRTGTYQRAILQNHTDFKDKIVLDVGCGSGILSFFAAQAGARKIYAVEASTMAQHAEVLVKSNNLTDRIVVIPGKVEEVSLPEQVDIIISEPMGYMLFNERMLESYLHAKKYLKPSGNMFPTIGDVHLAPFTDEQLYMEQFTKANFWYQPSFHGVDLSALRGAAVDEYFRQPVVDTFDIRILMAKSVKYTVNFLEAKEGDLHRIEIPFKFHMLHSGLVHGLAFWFDVAFIGSIMTVWLSTAPTEPLTHWYQVRCLFQSPLFAKAGDTLSGTCLLIANKRQSYDISIVAQVDQTGSKSSNLLDLKNPFFRYTGTTPSPPPG\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCHHHHTSCHHHHHHHHHHTTCHHHHHHHHTCHHHHHHHHHHHHHTGGGTTTCEEEEETCTTSHHHHHHHHTTCSEEEEEECSTHHHHHHHHHHHTTCTTTEEEEESCTTTCCCSSCEEEEEECCCBTTBTTTSHHHHHHHGGGGEEEEEEEESCEEEEEEEEECCHHHHHHHHHHHGGGGCSCBTTBCCGGGHHHHHHHHHTSCEEECCCGGGBCSCCEEEEEETTTCCGGGGSEEEEEEEEECSSCEEEEEEEEEEEEEEECSSCEEEEECCTTSCCCTTCEEEEEEEEEEEECTTCEEEEEEEEEEETTTEEEEEEEEEETTTCCEEEEEEETTSCEECCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01656", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "CCNC(=O)C1CCCN1C(=O)C(CCCN=C(N)N)NC(=O)C(CC(C)C)NC(=O)C(COC(C)(C)C)NC(=O)C(Cc1ccc(O)cc1)NC(=O)C(CO)NC(=O)C(Cc1c[nH]c2ccccc12)NC(=O)C(Cc1cnc[nH]1)NC(=O)C1CCC(=O)N1", "output": "The molecule contains the following functional groups: aromatic ring, amide, amine, hydroxyl, ether, phenol, indole, imidazole.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "SIDER", "canonical_smiles": "CCNC(=O)C1CCCN1C(=O)C(CCCN=C(N)N)NC(=O)C(CC(C)C)NC(=O)C(COC(C)(C)C)NC(=O)C(Cc1ccc(O)cc1)NC(=O)C(CO)NC(=O)C(Cc1c[nH]c2ccccc12)NC(=O)C(Cc1cnc[nH]1)NC(=O)C1CCC(=O)N1", "compound_name": null, "pubchem_cid": null, "MW": 1239.45, "LogP": -1.67, "TPSA": 440.77, "QED": 0.01, "functional_groups": ["aromatic ring", "amide", "amine", "hydroxyl", "ether", "phenol", "indole", "imidazole"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 1, "Product issues": 0, "Eye disorders": 1, "Investigations": 1, "Musculoskeletal and connective tissue disorders": 1, "Gastrointestinal disorders": 1, "Social circumstances": 1, "Immune system disorders": 1, "Reproductive system and breast disorders": 1, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 1, "General disorders and administration site conditions": 1, "Endocrine disorders": 1, "Surgical and medical procedures": 0, "Vascular disorders": 1, "Blood and lymphatic system disorders": 1, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 1, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 1, "Renal and urinary disorders": 1, "Pregnancy, puerperium and perinatal conditions": 1, "Ear and labyrinth disorders": 1, "Cardiac disorders": 1, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "amide", "amine", "hydroxyl", "ether", "phenol", "indole", "imidazole"]}}
{"id": "eval_01657", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "TTR, HNF4A, ALB, APOB", "output": "This cell is most consistent with hepatocyte, a liver hepatocyte. Key markers such as APOB, HNF4A, TTR support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "heart", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["TTR", "HNF4A", "ALB", "APOB"]}}
{"id": "eval_01658", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CN1C(=O)[C@@](c2ccc(OC(F)(F)F)cc2)(c2ccc(F)c(-c3cncnc3)c2)N=C1N", "output": "This molecule is medium-sized and polar, with lipophilic character. It contains aromatic ring, amide, amine, ether, halide. Its descriptor profile: MW=445.38, LogP=3.21, TPSA=93.7, HBD=1, HBA=6, QED=0.62.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BACE", "canonical_smiles": "CN1C(=O)[C@@](c2ccc(OC(F)(F)F)cc2)(c2ccc(F)c(-c3cncnc3)c2)N=C1N", "compound_name": null, "pubchem_cid": null, "MW": 445.38, "LogP": 3.21, "TPSA": 93.7, "QED": 0.62, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "halide"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01659", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nIRX2:medium TTR:medium GCG:medium\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as IRX2, GCG, TTR support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "bone marrow", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["IRX2", "TTR", "GCG"]}}
{"id": "eval_01660", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRL36_LISW6\n\n### Answer:", "output": "Belongs to the bacterial ribosomal protein bL36 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01661", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMNMA_MARMS\n\n### Answer:", "output": "tRNA-specific 2-thiouridylase MnmA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01662", "category": "Mutation", "instruction": "Describe the effect of mutation P97532-C248S on protein 3-mercaptopyruvate sulfurtransferase.", "input": "### Wild-type:\nMAAPQLFRALVSAQWVAEALKSPRASQPLKLLDASWYLPKLGRDARREFEERHIPGAAFFDIDRCSDHTSPYDHMLPSATHFADYAGSLGVSAATHVVIYDGSDQGLYSAPRVWWMFRAFGHHSVSLLDGGFRYWLSQNLPISSGKSPSEPAEFCAQLDPSFIKTHEDILENLDARRFQVVDARAAGRFQGTQPEPRDGIEPGHIPGSVNIPFTEFLTSEGLEKSPEEIQRLFQEKKVDLSKPLVATCGSGVTACHVVLGAFLCGKPDVPVYDGSWVEWYMRAQPEHVISQGRGKTL\n\n### Mutant:\nMAAPQLFRALVSAQWVAEALKSPRASQPLKLLDASWYLPKLGRDARREFEERHIPGAAFFDIDRCSDHTSPYDHMLPSATHFADYAGSLGVSAATHVVIYDGSDQGLYSAPRVWWMFRAFGHHSVSLLDGGFRYWLSQNLPISSGKSPSEPAEFCAQLDPSFIKTHEDILENLDARRFQVVDARAAGRFQGTQPEPRDGIEPGHIPGSVNIPFTEFLTSEGLEKSPEEIQRLFQEKKVDLSKPLVATSGSGVTACHVVLGAFLCGKPDVPVYDGSWVEWYMRAQPEHVISQGRGKTL\n\n### Answer:", "output": "Loss of both enzyme activities. Greatly reduced redox potential. The oxidation with a stoichiometric concentration of hydrogen peroxide protected the enzyme against reaction by iodoacetate, which modifies a catalytic C(247), suggesting that C(247) is a target of the oxidants. There is a strong possibility that C248 is a catalytic site of MST.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01663", "category": "Mutation", "instruction": "Describe the effect of mutation Q9SXS3-Q106E on protein 2-hydroxyisoflavanone synthase.", "input": "### Wild-type:\nMLVELAITLLVIALFIHLRPTLSAKSKSLRHLPNPPSPKPRLPFVGHLHLLDKPLLHYSLIDLSKRYGPLYSLYFGSMPTVVASTPELFKLFLQTHEASSFNTRFQTSAIRRLTYDNSVAMVPFGPYWKFIRKLIMNDLLNATTVNKLRPLRSQEIRKVLRVMAQSAESQVPLNVTEELLKWTNSTISRMMLGEAEEIRDIARDVLKIFGEYSLTDFIWPLKKLKVGQYEKRIDDIFNRFDPVIERVIKKRQEIRKKRKERNGEIEEGEQSVVFLDTLLDFAEDETMEIKITKEQIKGLV\n\n### Mutant:\nMLVELAITLLVIALFIHLRPTLSAKSKSLRHLPNPPSPKPRLPFVGHLHLLDKPLLHYSLIDLSKRYGPLYSLYFGSMPTVVASTPELFKLFLQTHEASSFNTRFETSAIRRLTYDNSVAMVPFGPYWKFIRKLIMNDLLNATTVNKLRPLRSQEIRKVLRVMAQSAESQVPLNVTEELLKWTNSTISRMMLGEAEEIRDIARDVLKIFGEYSLTDFIWPLKKLKVGQYEKRIDDIFNRFDPVIERVIKKRQEIRKKRKERNGEIEEGEQSVVFLDTLLDFAEDETMEIKITKEQIKGLV\n\n### Answer:", "output": "No effect on catalytic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01664", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSKRPAFPGMGGMNMQQMMKQAKKLQEQMAQEQENITTQEFTGKAADDMVVATFTGDRTLKSLFIKPEAIDPDDPDMLEDLVIDAVNKGLKQIDQATQQSLGKYTKGLM\n\n### Sequence 2:\nMLIFILSILPYNICSLIGTELYDCIFLYTKSNGCSRLQTCIKVSCSMLEFMYLAEDIPIQFDGIGNCLLSMVIMEYHLLFANI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01665", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "MRC1, APOE, SPP1, CD163", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as CD163, MRC1, APOE support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "liver", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["MRC1", "APOE", "SPP1", "CD163"]}}
{"id": "eval_01666", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "SLC17A7, RBFOX3, SATB2", "output": "The expression pattern is consistent with excitatory neuron in a context associated with lung adenocarcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "excitatory neuron", "tissue": "liver", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_disease_state", "top_genes": ["SLC17A7", "RBFOX3", "SATB2"]}}
{"id": "eval_01667", "category": "Mol", "instruction": "Provide the compound name and basic info for this molecule.", "input": "Cc1cccc2oc3c(c(=O)c12)CC(O)C(C)(C)O3", "output": "The SMILES represents an unnamed molecule with MW=260.29, LogP=2.18, TPSA=59.67. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "HIV", "canonical_smiles": "Cc1cccc2oc3c(c(=O)c12)CC(O)C(C)(C)O3", "compound_name": null, "pubchem_cid": null, "MW": 260.29, "LogP": 2.18, "TPSA": 59.67, "QED": 0.79, "functional_groups": ["aromatic ring", "hydroxyl", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01668", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nMIAB_BORPA\n\n### Answer:", "output": "Belongs to the methylthiotransferase family. MiaB subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01669", "category": "Mutation", "instruction": "Describe the effect of mutation P04391-K87Q on protein Ornithine carbamoyltransferase subunit I.", "input": "### Wild-type:\nMSGFYHKHFLKLLDFTPAELNSLLQLAAKLKADKKSGKEEAKLTGKNIALIFEKDSTRTRCSFEVAAYDQGARVTYLGPSGSQIGHKESIKDTARVLGRMYDGIQYRGYGQEIVETLAEYASVPVWNGLTNEFHPTQLLADLLTMQEHLPGKAFNEMTLVYAGDARNNMGNSMLEAAALTGLDLRLVAPQACWPEAALVTECRALAQQNGGNITLTEDVAKGVEGADFIYTDVWVSMGEAKEKWAERIALLREYQVNSKMMQLTGNPEVKFLHCLPAFHDDQTTLGKKMAEEFGLHGGME\n\n### Mutant:\nMSGFYHKHFLKLLDFTPAELNSLLQLAAKLKADKKSGKEEAKLTGKNIALIFEKDSTRTRCSFEVAAYDQGARVTYLGPSGSQIGHQESIKDTARVLGRMYDGIQYRGYGQEIVETLAEYASVPVWNGLTNEFHPTQLLADLLTMQEHLPGKAFNEMTLVYAGDARNNMGNSMLEAAALTGLDLRLVAPQACWPEAALVTECRALAQQNGGNITLTEDVAKGVEGADFIYTDVWVSMGEAKEKWAERIALLREYQVNSKMMQLTGNPEVKFLHCLPAFHDDQTTLGKKMAEEFGLHGGME\n\n### Answer:", "output": "Much less active than the wild-type. Substitutions at either of two conserved residues, S 55 and K 86, in the proposed active site of OTCase had deleterious effects parallel to those caused by the analogous ATCase substitutions.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01670", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CC(=O)NC(Cc1cc(F)cc(F)c1)C(O)C[NH2+]C1(c2cccc(C3CCCCO3)c2)CC1\nDescriptors: MW=459.56, LogP=2.87, TPSA=75.17, HBD=3, HBA=3", "output": "Descriptors: MW=459.56, LogP=2.87, TPSA=75.17, HBD=3, HBA=3, QED=0.54, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "BACE", "canonical_smiles": "CC(=O)NC(Cc1cc(F)cc(F)c1)C(O)C[NH2+]C1(c2cccc(C3CCCCO3)c2)CC1", "compound_name": null, "pubchem_cid": null, "MW": 459.56, "LogP": 2.87, "TPSA": 75.17, "QED": 0.54, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01671", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nTRMN6_BACFN\n\n### Answer:", "output": "tRNA1(Val) (adenine(37)-N6)-methyltransferase, tRNA m6A37 methyltransferase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01672", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSDLEVQVPTAFDPFADANAEDSGAGTKEYVHIRVQQRNGRKSLTTVQGLKKEYSYSKILKDLKKEFCCNGTVVQDSELGQVIQLQGDQRKNVSTFLVQAGLVKKDNIKIHGF\n\n### Sequence 2:\nMSHRVSILSSHFSPASAVMASEKEAALAATPSDSPTIFDKIISKEIPSTVVFEDDKVLAFRDITPQGPVHILLIPKVRDGLTGLSKAEERHIDILGRLLYTAKLVAKQEGLAEGFRIVINDGPQGCQSVYHIHVHLIGGRQMNWPPG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01673", "category": "Cell", "instruction": "Which biological pathway is most consistent with these markers?", "input": "CD3E, CD8A, SELL, CD3D, CCR7", "output": "The marker profile suggests functional enrichment consistent with the typical biology of naive CD8+ T cell.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "pathway_explain", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "lung", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_pathway", "top_genes": ["CD3E", "CD8A", "SELL", "CD3D", "CCR7"]}}
{"id": "eval_01674", "category": "Mutation", "instruction": "Describe the effect of mutation P71590-A116T on protein FHA domain-containing protein FhaA.", "input": "### Wild-type:\nMGSQKRLVQRVERKLEQTVGDAFARIFGGSIVPQEVEALLRREAADGIQSLQGNRLLAPNEYIITLGVHDFEKLGADPELKSTGFARDLADYIQEQGWQTYGDVVVRFEQSSNLHAGQFRARGTVNPDVETHPPVIDCARPQSNHAFGAEPGVAPMSDNSSYRGGQGQGRPDEYYDDRYARPQEDPRGGPDPQGGSDPRGGYPPETGGYPPQPGYPRPRHPDQGDYPEQIGYPDQGGYPEQRGYPEQRGYPDQRGYQDQGRGYPDQGQGGYPPPYEQRPPVSPGPAAGYGAPGYDQGYRQ\n\n### Mutant:\nMGSQKRLVQRVERKLEQTVGDAFARIFGGSIVPQEVEALLRREAADGIQSLQGNRLLAPNEYIITLGVHDFEKLGADPELKSTGFARDLADYIQEQGWQTYGDVVVRFEQSSNLHTGQFRARGTVNPDVETHPPVIDCARPQSNHAFGAEPGVAPMSDNSSYRGGQGQGRPDEYYDDRYARPQEDPRGGPDPQGGSDPRGGYPPETGGYPPQPGYPRPRHPDQGDYPEQIGYPDQGGYPEQRGYPEQRGYPDQRGYQDQGRGYPDQGQGGYPPPYEQRPPVSPGPAAGYGAPGYDQGYRQ\n\n### Answer:", "output": "Enhanced phosphorylation. The degree and the pattern of phosphorylation of the juxtamembrane domain enhances the affinity of the substrate (Rv0020c) toward its kinase (PknB).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01675", "category": "Mutation", "instruction": "Describe the effect of mutation Q96P20-A5S on protein NACHT, LRR and PYD domains-containing protein 3.", "input": "### Wild-type:\nMKMAATRCKLARYLEDLEDVDLKKFKMHLEDYPPQKGCIPLPRGQTEKADHVDLATLMIDFNGEEKAWAMAVWIFAAINRRDLYEKAKRDEPKWGSDNARVSNPTVICQEDSIEEEWMGLLEYLSRISICKMKKDYRKKYRKYVRSRFQCIEDRNARLGESVSLNKRYTRLRLIKEHRSQQEREQELLAIGKTKTCESPVSPIKMELLFDPDDEHSEPVHTVVFQGAAGIGKTILARKMMLDWASGTLYQDRFDYLFYIHCREVSLVTQRSLGDLIMSCCPDPNPPIHKIVRKPSRILFL\n\n### Mutant:\nMKMASTRCKLARYLEDLEDVDLKKFKMHLEDYPPQKGCIPLPRGQTEKADHVDLATLMIDFNGEEKAWAMAVWIFAAINRRDLYEKAKRDEPKWGSDNARVSNPTVICQEDSIEEEWMGLLEYLSRISICKMKKDYRKKYRKYVRSRFQCIEDRNARLGESVSLNKRYTRLRLIKEHRSQQEREQELLAIGKTKTCESPVSPIKMELLFDPDDEHSEPVHTVVFQGAAGIGKTILARKMMLDWASGTLYQDRFDYLFYIHCREVSLVTQRSLGDLIMSCCPDPNPPIHKIVRKPSRILFL\n\n### Answer:", "output": "Increased phosphorylation; decreased activation of the NLRP3 inflammasome. Phosphorylation of residue S5 in NLRP3 PYD disrupts charge-charge interaction between two PYDs that are essential for NLRP3 activation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01676", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "NC(=O)c1ccc(O[C@H]2C[C@H]3CC[C@@H](C2)N3Cc2ccccc2)nc1", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains aromatic ring, amide, amine, ether, pyridine. Its descriptor profile: MW=337.42, LogP=2.75, TPSA=68.45, HBD=1, HBA=4, QED=0.91.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "NC(=O)c1ccc(O[C@H]2C[C@H]3CC[C@@H](C2)N3Cc2ccccc2)nc1", "compound_name": null, "pubchem_cid": null, "MW": 337.42, "LogP": 2.75, "TPSA": 68.45, "QED": 0.91, "functional_groups": ["aromatic ring", "amide", "amine", "ether", "pyridine"], "moleculenet_labels": {"label": 1.25}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01677", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGPGYQDPNSAKLLQILNVRVVGSGERVVVLSHGFGTDQSAWSRVLPYLTRDHRVVLYDLVCAGSVNPDHFDFRRYDNLDAYVDDLLAILDALRIPRCAFVGHSVSAMIGILASIRRPDLFAKLVLIGASPRFLNDSDYHGGFELEEIQQVFDAMGANYSAWATGYAPLAVGADVPAAVQEFSRTLFNMRPDISLHVCQTVFKTDLRGVLGMVRAPCVVVQTTRDVSVPASVAAYLKAHLGGRTTVEFLQTEGHLPHLSAPSLLAQVLRRALARY\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCTTTTTCEEEECCSSEEEEECCTTCCGGGGTTTGGGTTTTCEEEEECCTTSTTSCGGGCCTTGGGSHHHHHHHHHHHHHHTTCCCEEEEEETHHHHHHHHHHHHCTTTEEEEEEESCCSCSBCBTTBCCSBCHHHHHHHHHHHHHCHHHHHHHHHHHHHCSCCHHHHHHHHHHHHTSCHHHHHHHHHHHHTCCCGGGGGGCCSCEEEEECSCCTTSCHHHHHHHHHHCSSCEEEEECSSCCSCHHHHCHHHHHHHHHHHTSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01678", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nVAFGADAAKTTQEKFDALKEAGVFSGYPGTTDAKLGQDMTRAEFAKVLVKLFGLKEIHGQYSYKDKNYDAKNWAAPFIEAVTAEGLMQGKDLTKKIFDFNGKITVEEASKTLVTALKLEPVKDAQNKATDWAKGYFEAAVNAGLFSKDANPKANATRAQLVEAAFAADEMSKGSGSHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCHHHHHHHHHHTTSSCCCTTCCCCCTTSBCBHHHHHHHHHHHHTCCCCCSCCCCCCTTCSTTCTTHHHHHHHHHHTSCCCSBTTTTBCCSSSBCBHHHHHHHHHHHTTCCCCTTCCSSSCGGGHHHHHHHHHTTSSCTTSCTTSBCBHHHHHHHHHHHHHHHCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01679", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMAALTRDPQFQKLQQWYREHRSELNLRRLFDANKDRFNHFSLTLNTNHGHILVDYSKNLVTEDVMRMLVDLAKSRGVEAARERMFNGEKINYTEGRAVLHVALRNRSNTPILVDGKDVMPEVNKVLDKMKSFCQRVRSGDWKGYTGKTITDVINIGIGGSDLGPLMVTEALKPYSSGGPRVWYVSNIDGTHIAKTLAQLNPESSLFIIASKTFTTQETITNAETAKEWFLQAAKDPSAVAKHFVALSTNTTKVKEFGIDPQNMFEFWDWVGGRYSLWSAIGLSIALHVGFDNFEQLLSGAHWMDQHFRTTPLEKNAPVLLALLGIWYINCFGCETHAMLPYDQYLHRFAAYFQQGDMESNGKYITKSGTRVDHQTGPIVWGEPGTNGQHAFYQLIHQGTKMIPCDFLIPVQTQHPIRKGLHHKILLANFLAQTEALMRGKSTEEARKELQAAGKSPEDLERLLPHKVFEGNRPTNSIVFTKLTPFMLGALVAMYEHKIFVQGIIWDINSFDQWGVELGKQLAKKIEPELDGSAQVTSHDASTNGLINFIKQQREARVQ\n\n### Answer:", "output": "<SEQ_2D>CCTTTTCHHHHHHHHHHHHHGGGCCHHHHHHHCTTHHHHTEEEEECSSCEEEEECTTSSCCHHHHHHHHHHHHHTTHHHHHHHHHHTCCCBTTTTBCCCHHHHTCTTCCCCEETTEESHHHHHHHHHHHHHHHHHHHHTCSBCTTSCBCCEEEEECCGGGTHHHHHHHHHTGGGCTTSCEEEEECCSSHHHHHHHHTTCCTTSEEEEEECSSSCCHHHHHHHHHHHHHHHTSSCCTTGGGGTEEEESSCHHHHHHHTCCGGGEECCCTTSCGGGCTTSGGGHHHHHHHCHHHHHHHHHHHHHHHHHHHHSCGGGCHHHHHHHHHHHHHHTSCCCEEEEEESCGGGTTHHHHHHHHTHHHHCCCBCTTSCBCSSCCCCEEECCCTTTGGGTTHHHHHHSSCCCCEEEEEEEECSCCGGGGHHHHHHHHHHHHHHHHHHHCBCHHHHHHHHHHHTCCHHHHHHHGGGGCBCCCCCEEEEEEEECCHHHHHHHHHHHHHHHHHHHHHHTCCTTCCGGGHHHHHHHHHHHHHTSSSCCCCSSCHHHHHHHHHHHHHHTCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01680", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRUVA_CERS4\n\n### Answer:", "output": "Holliday junction branch migration complex subunit RuvA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01681", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPPAC_BACSU\n\n### Answer:", "output": "Manganese-dependent inorganic pyrophosphatase, Pyrophosphate phospho-hydrolase, PPase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01682", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGACDNDTEPGGTAVEKMAGDWWVTVNAFIDGKEVEDPFGAGHLQMSTYNTASNSETEMWLDDLGNFWEYKLKVNVNYAARTFSTTGFVDNVTYESKVKITDGKVLEKAATTPSGMPADSIVYMVQFDDDEDGLTYKVSGFRRTGFPADDF\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCGGGCEEEEEEEEEEETTEEESSTTCCCSEEEEEEECTTCCSSEEEEECTTSSSCCEEEEEEETTTTEEECSSCEECSSSSCEEEEEEEEEEEEEEECTTSCEEEEEEEEEEETTCTTCCCEEEEEEECCSCGGGCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01683", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMLTVNEYFEGKVKSIGFSPAGLPATIGVMAIGEYTFGTDCREIMTVVSGELTVKLPEQADWHTYIAGQTFEVEANQSFDLKVAVETAYLCQYDR\n\n### Sequence 2:\nMFKVNEYFDGTVKSIAFGTAEGPATIGVMAPGEYEFGTAQREIMHVVSGALTVRLPDSADWETFPAGSQFNVPANSKFQLKVAVDTAYLCEY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01684", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMLVTTTERISGQEYEIIGEVFGLTTRSKNMFKDLGAGLKSVVGGEIKGYTDMQREARDQAIERLKAEASKLGADAVVMMRFDSGTIGTDMQSVVAYGTAVKYI\n\n### Sequence 2:\nILVTTTENIPGRKYEIIGEVFGVTTQSKNAIRDFGAGLKSIVGGEIKAYTSMLTESRDQSIARLRQNASEMGANAVVMMRFDSGSIAGDMQSVVAYGTAVKFI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01685", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMGSSHHHHHHSSGLVPRGSSAEETIQYAAAEISASRQYTVAIETLHENLRESIADVLYMEPYEVDIDEAFIDIGMDSITGLEWIKAVNKQYGTSFTVTRVYDYPTIRDFAEMLKSELGTHLDRKIEHTDSFEAAQQKPAASSHPKPAERPLQPVQHPIKKEHEKKTVPVLQDRPEDAIAIVGMSGRYPGARNVREYWDNLVHARNAIRDIPTSRWDVDKYYDPVLNKKGKVYCKSMGMLDDIEHFDPLFFNIPPSEAELMDPQHRIFLQEGYKAFEDAGYNARTLNEKKCGVYLGIMSNEYGVMLNRQSRANATGNSFAIAAARIPYFLNLKGPAIPIDTACSSSLVGTHLARQALINKEIDMALVGGVSLYLTPESYMSMCEAGMLSPDGQCKAFDNGANGFVPGEGAGALVLKRLKDAEADRDHIYGIIIGSGINQDGKTNGITAPSAKSQMDLERDIYETYGIHPESISYVEMHGTGTKQGDPIELEALSTVFQEKTDKKQFCAIGSVKSNIGHTSAAAGVAGVQKVLLCMNHKTLVPTLNFTTPNEHFEFEHSPLYVNTELKPWETADGKPRRACVSSFGYSGTNAHIVIEEYQPEKRNDRLTKQHRSALFVLSAKKEKQLKAYAEAMKDFVTSNEDIDLEDMAYTLQTGREAMDYRMAFLADSREMLIKALDDYLAEMPNGSIFAAHVKTKKSEIKLFETDHDAKALLQTWIEKKRLEKVAELWVKGLQIDWNKLYGEYTPRRISLPAYPFAEEYYWLP\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCEEEEEEEEEBTTBSSHHHHHHHHHTTCCCCEECCTTTCCGGGTBCCCCCCCCCBSCCEEBCCSCTTEECTTTTTCCHHHHTTSCHHHHHHHHHHHHHHHHHTCCHHHHTTCCEEEEEECCCCHHHHHHCCCCCCCCCCCSHHHHHHHHHHHHTCCSCEEEECSGGGHHHHHHHHHHHHHHTTSCSEEEEEEEECCCSHHHHHCCCCCCCBCTTCCCCTTBTTCCCBCBBCEEEEEEEEEHHHHHHTTCCCCEEEEEEEEEECCSCSSTTSCCHHHHHHHHHHHHHHHTCCGGGCCEEECCCCCCTTHHHHHHHHHHHHHHTTCCCCSCEEEECTHHHHCBCGGGHHHHHHHHHHHHHHTTEECCCCSCSSBCTTSCCTTSSEEECCSCEECCCCTTCCCEEEEEEECTTSEEEEEEEECCCCCCCCCCCCCCCCCEEEEEEESSHHHHHHHHHHHHHHHHHCTTCCHHHHHHHHHHSSCCCSEEEEEEESSHHHHHHHHHHHHHTCCCTTEEEEEGGGCHHHHHHHHHCHHHHHHHHHHHHHTCHHHHHHHHHHTCCCCGGGGCSSCCCCCCCCCCCCCCCEECCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01686", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nDNTIQWDKDADGIVTLTMDDPSGSTNVMNEAYIESMGKAVDRLVAEKDSITGVVVASAKKTFFAGGDVKTMIQARPEDAGDVFNTVETIKRQLRTLETLGKPVVAAINGAALGGGLEIALACHHRIAADVKGSQLGLPEVTLGLLPGGGGVTRTVRMFGIQNAFVSVLAQGTRFKPAKAKEIGLVDELVATVEELVPAAKAWIKEELKANPDGAGVQPWDKKGYKMPGGTPSLPSFPSNLRKQLKGAPMPAPRAILAAAVEGAQVDFDTASRIESRYFASLVTGQVAKNMMQAFFFDLQAINAGGSRPEGIGKTPIKRIGVLGAGMMGAGIAYVSAKAGYEVVLKDVSLEAAAKGKGYSEKLEAKALERGRTTQERSDALLARITPTADAADFKGVDFVIEAVFENQELKHKVFGEIEDIVEPNAILGSNTSTLPITGLATGVKRQEDFIGIHFFSPVDKMPLVEIIKGEKTSDEALARVFDYTLAIGKTPIVVNDSRGFFTSRVIGTFVNEALAMLGEGVEPASIEQAGSQAGYPAPPLQLSDELNLELMHKIAVATRKGVEDAGGTYQPHPAEAVVEKMIELGRSGRLKGAGFYEYADGKRSGLWPGLRETFKSGSSQPPLQDMIDRMLFAEALETQKCLDEGVLTSTADANIGSIMGIGFPPWTGGSAQFIVGYSGPAGTGKAAFVARARELAAAYGDRFLPPESLLS\n\n### Answer:", "output": "<SEQ_3Di>DAQWDWDADQQQAIETEGHDPPDQAAAADPNVLVNLVVVLVVCVVPVVSHAHYEYAYPHQFGHAAHDLVVLLPQALVCLVVLLVSLVSLQVSLQSQLQSLHAYEYLGQAAQEQVRLLSRLSGPAYEYEQDPPGKYAYQCVLLLFFRLNCLLLSLCQVANQLLSQPPGTQHTDIDHPVVCCVSPSHPHYDHHSVCRVVVVVVVRVVCCPVPVRCHSPHQLNDDPDAGHPGAVCVVCQPSDDVSNDDPQLRPSNNLSSVLNRVLRVDGRVVSSSSSSSSRSVRSNDLLSSLSCVFVPVLQVCQVVQVPADPDADADDFAAEEEEDLPLARLLVLQLLLLLPHQYEYAYQAQVRQVVSLVVNLVVVVVCPVVVVDHPVVSVSSSVSYHYGHDLLVQAAGAEYEYDDDPDLVVLLVVCLSRVVRHDPNHAAEYADQFDWQQSSLVSDPQSQRYWYWHADPNRVPQQEIETEDEPNHDSNNVNNVSVSSVSSPHGYFYFYTDGRLQQRVLLCLQLLLLLLCLLVAAANVLQQVLQVSLAFPGGSQLVQQVVWLVVVLVSQVVVQCVQVVDPHHDDDRSSNVSSVVQVVVVAGHQVSQHHQFHHDPRDTDGGDPCSCVVNVHPPDDDDSVLSNCSSLLSSLLVLVVCVVVVRHDDVSSQQCCCVSPNRRHCSCSGSPSVQQAPDDPVDGGNVVVQVVLVVSCVVPHDSSHHDCVSPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01687", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nCD3D IL2RA CTLA4 FOXP3 IKZF2\n</cell>", "output": "This cell is most consistent with regulatory T cell, a FOXP3+ regulatory T cell. Key markers such as IKZF2, CD3D, FOXP3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "regulatory T cell", "tissue": "colon", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD3D", "IL2RA", "CTLA4", "FOXP3", "IKZF2"]}}
{"id": "eval_01688", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nHIS2_MYCGI\n\n### Answer:", "output": "Belongs to the PRA-PH family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01689", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nRHMKVLLLGFEFLPVKVGGLAEALTAISEALASLGHEVLVFTPSHGRFQGEEIGKIRVFGEEVQVKVSYEERGNLRIYRIGGGLLDSEDVYGPGWDGLIRKAVTFGRASVLLLNDLLREEPLPDVVHFHDWHTVFAGALIKKYFKIPAVFTIHRLNKSKLPAFYFHEAGLSELAPYPDIDPEHTGGYIADIVTTVSRGYLIDEWGFFRNFEGKITYVFNGIDCSFWNESYLTGSRDERKKSLLSKFGMDEGVTFMFIGRFDRGQKGVDVLLKAIEILSSKKEFQEMRFIIIGKGDPELEGWARSLEEKHGNVKVITEMLSREFVRELYGSVDFVIIPSYFEPFGLVALEAMCLGAIPIASAVGGLRDIITNETGILVKAGDPGELANAILKALELSRSDLSKFRENCKKRAMSFSWEKSAERYVKAYTGSIDRAFDFIL\n\n### Answer:", "output": "<SEQ_2D>CCCEEEEECSCCTTSCSSSHHHHHHHHHHHHHHTTCEEEEEEECTTCSCCEEEEEEEETTEEEEEEEEEEEETTEEEEEEESGGGGCSSTTCSHHHHHHHHHHHHHHHHHHHHHHHTTTSCCCSEEEEESGGGHHHHHHHHHHHCCCEEEEESCCCCCCEEHHHHHHTTCGGGCCSSEECHHHHHHHHCSEEEESCHHHHHHTHHHHGGGTTSEEECCCCCCTTTSCGGGSCSCHHHHHHHHHHHHTCCSCEEEEEECCSSCTTBCHHHHHHHHHHHHTSGGGGGEEEEEECCCCHHHHHHHHHHHHHCTTEEEECSCCCHHHHHHHHTTCSEEEECBSCCSSCHHHHHHHHTTCEEEEESSTHHHHHCCTTTCEEECTTCHHHHHHHHHHHHHHTTTTTHHHHHHHHHHHHTSCHHHHHHHHHHHHHTCSCCBCSSBC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01690", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMETLTAISRWLAKQHVVTWCVQQEGELWCANAFYLFDAQKVAFYILTEEKTRHAQMSGPQAAVAGTVNGQPKTVALIRGVQFKGEIRRLEGEESERARQAYNRRFPVARMLSAPVWEIRLDEIKFTDNTLGFGKKMIWLRNSGTEQA\n\n### Sequence 2:\nMETLTAISRWLAKQHVVTWCVQQEGELWCANAFYLFDAQKVAFYILTEEKTRHAQMSGPQAAVAGTVNGQPKTVALIRGVQFKGEIRRLEGEESDLARKAYNRRFPVARMLSAPVWEIRLDEIKFTDNTLGFGKKMIWLRDSGTEQA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01691", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMTRGERVIAFIERFCIVPIGQPMRLDPFQKDFILAVYDNPAGTDMAILSIARKNGKTGLIAGILLAHLVGPEAVQNTQIVSGALSREQAAIVFNLAVKMVNLNPKLQEIVHITPSGKKLIGLPCNVEYKALSAEGKTTHGLSPILAILDETGQVRGPQDDFIDAITTAQGAHENPLLIVISTQAANDADLLSIWIDDAVKSKDPHIVCHVYEAPKDADISKRESWLAANPALGTFRSEKDMARQAEKAGRMPSFENTFRNLNLNQRVSTVSPFISRSVWELCGEMPINTPRKWYAGLDLSARNDLTALVIAGEADDGVWDVFPFFWTPQKTLEERTKTDRAPYDVWVREGLLRTTPGASVDYSFVVADIAEIIGDFDLTSMAFDRWRIDQFRKDADAIGLSLPLVEFGQGFKDMGPAVDTLESLMLNGRVRHGMHPVLTMCAVNAVVVKDAAGNRKLDKSKATGRIDGMVAMTMSVGAA\n\n### Answer:", "output": "<SEQ_3Di>DFPLVQLQVQQQPQFAPVPDHGDDDDPVLSVQSCLQRVLPVHAQEEEDAFFPQQCVLSSVLSVVVCCQANPVDAAPFEEEEEEADPVVLVSSLVSNVVSQVVGVVSVVQWDADPDARWIHRHPRRYIYHRDHNPDDDDDDAAHLEYEYEAQLPAAALDDPVVVVNVVNPVVDDRGYYYYYHQHRQFCRGNVVLVQVVCVVVVPSNYHHHHQAQDPPDDQQDVVLQCSRRVCDPNRDDPVVSNVLSVVVVVDVVCVSVCRRRRRSDHLKDKDALDDLVLVVLQQAAFDQDDFAKFKEWAQPCFQAKTKMWIWGQDPPRAIEIDMAMEHEDVCQVVVCVVQVPPSVVCCVVVLYDYDYDRTGDLLVVLVVVVVVCVSPDYPEYEDAQVCVVVSQVSNVVVVHHGRYDYAHPDQVRLQVLSVVVSVCSVVSRYHHNPSPVQSQLNRQWMWMAHPVGGIDTHQSVGRHDNNRVVRVSRRSSVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01692", "category": "Mutation", "instruction": "Describe the effect of mutation P05186-T292P on protein Alkaline phosphatase, tissue-nonspecific isozyme.", "input": "### Wild-type:\nMISPFLVLAIGTCLTNSLVPEKEKDPKYWRDQAQETLKYALELQKLNTNVAKNVIMFLGDGMGVSTVTAARILKGQLHHNPGEETRLEMDKFPFVALSKTYNTNAQVPDSAGTATAYLCGVKANEGTVGVSAATERSRCNTTQGNEVTSILRWAKDAGKSVGIVTTTRVNHATPSAAYAHSADRDWYSDNEMPPEALSQGCKDIAYQLMHNIRDIDVIMGGGRKYMYPKNKTDVEYESDEKARGTRLDGLDLVDTWKSFKPRYKHSHFIWNRTELLTLDPHNVDYLLGLFETGDMQYELN\n\n### Mutant:\nMISPFLVLAIGTCLTNSLVPEKEKDPKYWRDQAQETLKYALELQKLNTNVAKNVIMFLGDGMGVSTVTAARILKGQLHHNPGEETRLEMDKFPFVALSKTYNTNAQVPDSAGTATAYLCGVKANEGTVGVSAATERSRCNTTQGNEVTSILRWAKDAGKSVGIVTTTRVNHATPSAAYAHSADRDWYSDNEMPPEALSQGCKDIAYQLMHNIRDIDVIMGGGRKYMYPKNKTDVEYESDEKARGTRLDGLDLVDTWKSFKPRYKHSHFIWNRTELLTLDPHNVDYLLGLFEPGDMQYELN\n\n### Answer:", "output": "In HOPS; 4% of wild-type inactivity. The missense mutation resulted in gain of most enzymatic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01693", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nSLC32A1 GAD1 GAD2\n</cell>", "output": "This cell is most consistent with inhibitory neuron, an inhibitory (GABAergic) neuron. Key markers such as SLC32A1, GAD1, GAD2 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "inhibitory neuron", "tissue": "tumor", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SLC32A1", "GAD1", "GAD2"]}}
{"id": "eval_01694", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMRFTLKTTAIVSAAALLAGFGPPPRAAELPPGRLATTEDYFAQQAKQAVTPDVMAQLAYMNYIDFISPFYSRGCSFEAWELKHTPQRVIKYSIAFYAYGLASVALIDPKLRALAGHDLDIAVSKMKCKRVWGDWEEDGFGTDPIEKENIMYKGHLNLMYGLYQLVTGSRRYEAEHAHLTRIIHDEIAANPFAGIVCEPDNYFVQCNSVAYLSLWVYDRLHGTDYRAATRAWLDFIQKDLIDPERGAFYLSYHPESGAVKPWISAYTTAWTLAMVHGMDPAFSERYYPRFKQTFVEVYDEGRKARVRETAGTDDADGGVGLASAFTLLLAREMGDQQLFDQLLNHLEPPAKPSIVSASLRYEHPGSLLFDELLFLAKVHAGFGALLRMPPPAAKLAGK\n\n### Sequence 2:\nMSDLSPERFKLAVTSPSSIPESSSALQLHHSYSRKQKSLGLLCTNFLALYNREGIEMVGLDDAASKLGVERRRIYDIVNVLESVGVLTRRAKNQYTWKGFSAIPGALKELQEEGVKDTFHRFYVNENVKGSDDEDDDEESSQPHSSSQTDSSKPGSLPQSSDPSKIDNRREKSLGLLTQNFIKLFICSEAIRIISLDDAAKLLLGDAHNTSIMRTKVRRLYDIANVLSSMNLIEKTHTLDSRKPAFKWLGYNGEPTFTLSSDLLQLESRKRAFGTDITNVNVKRSKSSSSSQENATERRLKMKKHSTPESSYNKSFDVHESRHGSRGGYHFGPFAPGTGTYPTAGLEDNSRRAFDVENLDSDYRPSYQNQVLKDLFSHYMDAWKTWFSEVTQENPLPNTSQHR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01695", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: VEGFR2\nLigand SMILES: CNC1=Nc2ncccc2C(c2cc[nH]c2)=NC1c1cccs1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "CNC1=Nc2ncccc2C(c2cc[nH]c2)=NC1c1cccs1", "compound_name": null, "pubchem_cid": null, "MW": 321.41, "LogP": 3.31, "TPSA": 65.43, "QED": 0.76, "functional_groups": ["amine", "pyridine"], "moleculenet_labels": {"label": 1.73}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_01696", "category": "Mutation", "instruction": "Describe the effect of mutation Q8IXW5-A100C on protein Putative RNA polymerase II subunit B1 CTD phosphatase RPAP2.", "input": "### Wild-type:\nMADFAGPSSAGRKAGAPRCSRKAAGTKQTSTLKQEDASKRKAELEAAVRKKIEFERKALHIVEQLLEENITEEFLMECGRFITPAHYSDVVDERSIVKLAGYPLCQKKLGIVPKQKYKISTKTNKVYDITERKSFCSNFCYQASKFFEAQIPKTPVWVREEERHPDFQLLKEEQSGHSGEEVQLCSKAIKTSDIDNPSHFEKQYESSSSSTHSDSSSDNEQDFVSSILPGNRPNSTNIRPQLHQKSIMKKKAGHKANSKHKDKEQTVVDVTEQLGDCKLDSQEKDATCELPLQKVNTQSS\n\n### Mutant:\nMADFAGPSSAGRKAGAPRCSRKAAGTKQTSTLKQEDASKRKAELEAAVRKKIEFERKALHIVEQLLEENITEEFLMECGRFITPAHYSDVVDERSIVKLCGYPLCQKKLGIVPKQKYKISTKTNKVYDITERKSFCSNFCYQASKFFEAQIPKTPVWVREEERHPDFQLLKEEQSGHSGEEVQLCSKAIKTSDIDNPSHFEKQYESSSSSTHSDSSSDNEQDFVSSILPGNRPNSTNIRPQLHQKSIMKKKAGHKANSKHKDKEQTVVDVTEQLGDCKLDSQEKDATCELPLQKVNTQSS\n\n### Answer:", "output": "Enables interaction with RNA polymerase II complex subunits; when associated with A-105.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01697", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nCUC3A_HAECO\n\n### Answer:", "output": "Nematode cuticles are composed largely of collagen-like proteins.  The cuticle functions both as an exoskeleton and as a barrier to protect the worm from its environment.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01698", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "NKG7, CD8B, CD3D, CD8A, PRF1, GZMB, GZMA", "output": "This profile is most consistent with cytotoxic T cell, which is commonly found in tumor under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "cytotoxic T cell", "tissue": "tumor", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["NKG7", "CD8B", "CD3D", "CD8A", "PRF1", "GZMB", "GZMA"]}}
{"id": "eval_01699", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMEKHMKVVAFERQQQGTGASRRLRNAGKTTGIVYGGEAAPQMIELDHNALWHALKKEAFHSSILDLEVAGKSQRVLLRDVQYHPFRQLVLHVDFQRIDPKKKLHTKAPLHFLNAETSPAVKLSSAVVSHVVTEIEIECLPADLPEFLEVDLSKIEAGQSLHAKDIALPNGVALTAHVDAENPVIASATIPAGAVSDEAAAGEGETPAA\n\n### Sequence 2:\nMSTAKLVKSKATNLLYTRNDVSDSEKKATVELLNRQVIQFIDLSLITKQAHWNMRGANFIAVHEMLDGFRTALIDHLDTMAERAVQLGGVALGTTQVINSKTPLKSYPLDIHNVQDHLKELADRYAIVANDVRKAIGEAKDDDTADILTAASRDLDKFLWFIESNIE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01700", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGAGE1_HUMAN\n\n### Answer:", "output": "G antigen 1, GAGE-1, Antigen MZ2-F, Cancer/testis antigen 4.1, CT4.1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01701", "category": "Mutation", "instruction": "Describe the effect of mutation P30793-S249R on protein GTP cyclohydrolase 1.", "input": "### Wild-type:\nMEKGPVRAPAEKPRGARCSNGFPERDPPRPGPSRPAEKPPRPEAKSAQPADGWKGERPRSEEDNELNLPNLAAAYSSILSSLGENPQRQGLLKTPWRAASAMQFFTKGYQETISDVLNDAIFDEDHDEMVIVKDIDMFSMCEHHLVPFVGKVHIGYLPNKQVLGLSKLARIVEIYSRRLQVQERLTKQIAVAITEALRPAGVGVVVEATHMCMVMRGVQKMNSKTVTSTMLGVFREDPKTREEFLTLISS\n\n### Mutant:\nMEKGPVRAPAEKPRGARCSNGFPERDPPRPGPSRPAEKPPRPEAKSAQPADGWKGERPRSEEDNELNLPNLAAAYSSILSSLGENPQRQGLLKTPWRAASAMQFFTKGYQETISDVLNDAIFDEDHDEMVIVKDIDMFSMCEHHLVPFVGKVHIGYLPNKQVLGLSKLARIVEIYSRRLQVQERLTKQIAVAITEALRPAGVGVVVEATHMCMVMRGVQKMNSKTVTSTMLGVFREDPKTREEFLTLIRS\n\n### Answer:", "output": "The mutation in GTPCH is associated with the dominant dopa-responsive dystonia (DRD) disease. The mutant protein expression level is higher than the wild-type protein, suggesting a stabilizing effect. This mutation enhances the enzyme activity and kinetics in transfected eukaryotic cells.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01702", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nQLSSTFYDTSCPNALSTIRSAVNAAVAQENRMGASLLRLHFHDCFVQGCDASVLLNDTNGGEQNALPNAGSLRGFGVIDNIKAQVEALCPQTVSCADILAVAARDSVVALGGPSWTVPLGRRDSTNSSAALANSDLPPPQFNLSQLITAFGNKNLDPTDLVALSGAHTIGQAQCLNFRAHITEPNINPTFAASLRANCPATGGDTNLAPLDVTTPNTFDNAYYTNLLNQRGLLHSDQELFNNASTDSTVRNFASNAAAFTTAFTTAMIKMGNLQPLTGTQGQIRRNCWRVN\n\n### Answer:", "output": "<SEQ_3Di>DFALCQCVPQFVCLLVLLLVLLLVVCVVPLQLLLLLLVQLLLCQQFLHLALQQQADDPQHDLCVAPVRDPFHDDSVSLQVSLLVSCVRHNFDAASSQSSLSSNQSSQVSQPADHDRATGFFHHAQHGDNPSSVVQDDALPDAPVVLQVSCVVLVHHPLLSLLQVLSLLAHKDWLLRQVVVLPDDQADVVLSVVSCVQRPPDDSRGGIDRQQNPRHSHSWLSLLVCLVVRNGRTNHSVSCCDPDPNNVVSVCCVVPGVVSRVSNVVSSNSSNPRNGQGDPGHGRDSGSHHGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01703", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "KRT14, KRT5, TP63", "output": "The expression pattern is consistent with a healthy basal epithelial cell without obvious disease features.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "basal epithelial cell", "tissue": "bone marrow", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["KRT14", "KRT5", "TP63"]}}
{"id": "eval_01704", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL10_SYNFM\n\n### Answer:", "output": "Forms part of the ribosomal stalk, playing a central role in the interaction of the ribosome with GTP-bound translation factors.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01705", "category": "Mutation", "instruction": "Describe the effect of mutation P06858-T98A on protein Lipoprotein lipase.", "input": "### Wild-type:\nMESKALLVLTLAVWLQSLTASRGGVAAADQRRDFIDIESKFALRTPEDTAEDTCHLIPGVAESVATCHFNHSSKTFMVIHGWTVTGMYESWVPKLVATLYKREPDSNVIVVDWLSRAQEHYPVSAGYTKLVGQDVARFINWMEEEFNYPLDNVHLLGYSLGAHAAGIAGSLTNKKVNRITGLDPAGPNFEYAEAPSRLSPDDADFVDVLHTFTRGSPGRSIGIQKPVGHVDIYPNGGTFQPGCNIGEAIRVIAERGLGDVDQLVKCSHERSIHLFIDSLLNEENPSKAYRCSSKEAFEKG\n\n### Mutant:\nMESKALLVLTLAVWLQSLTASRGGVAAADQRRDFIDIESKFALRTPEDTAEDTCHLIPGVAESVATCHFNHSSKTFMVIHGWTVTGMYESWVPKLVAALYKREPDSNVIVVDWLSRAQEHYPVSAGYTKLVGQDVARFINWMEEEFNYPLDNVHLLGYSLGAHAAGIAGSLTNKKVNRITGLDPAGPNFEYAEAPSRLSPDDADFVDVLHTFTRGSPGRSIGIQKPVGHVDIYPNGGTFQPGCNIGEAIRVIAERGLGDVDQLVKCSHERSIHLFIDSLLNEENPSKAYRCSSKEAFEKG\n\n### Answer:", "output": "Increases the specific activity of the enzyme; enhances the secretion of the mutant protein significantly; the total LPL mass is increased compared to that of the wild-type construct. Increases the secretion of the mutant proteins significantly.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01706", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nGPMA_BORDL\n\n### Answer:", "output": "Catalyzes the interconversion of 2-phosphoglycerate and 3-phosphoglycerate.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01707", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "KRT5, TP63, KRT14", "output": "This cell is most consistent with basal epithelial cell, a basal epithelial cell. Key markers such as KRT14, KRT5, TP63 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "basal epithelial cell", "tissue": "brain", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["KRT5", "TP63", "KRT14"]}}
{"id": "eval_01708", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTHSTDLQWLLVRQNSKFLQKRNGIRLSSDPFNNNANWTKRHAGFLNTKAAVVKTKGDRILVTTKDGKAGNKPKSMYKKAVMDAGVEASVVSKAVAAVRPDLASIASRRARKMASTLEHMKKVRAARKERSSKITFQRKAVRPKRH\n\n### Sequence 2:\nMPGVLLWFFKGVIALLLAFAVGLIVYYYLEIRPIAQTALQSASFWLSESPQTRFLRRAAAKIHPKSYTARLLYTQAGVDGHFRIAIWVFWLDSLYRDDELYAMMLAQAYYGRDSQGNAVYGTKNAALTLFHVPVTEMACQQQVQLIYMFKAPSLYRPGSARLVESSKHYMTLCQEQIQQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01709", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSEKHFILPSSMLMIVSAVFGGIGIITTIVFVILTVLHSKSAVCKPAGKEDMKKLNGIEGMQTIKEECGGSTETSSSKPKKKAKKEV\n\n### Sequence 2:\nMIPGEYQIQDGEIELNAGRRTLTLNVANSGDRPIQVGSHYHFFETNDALVFDRALARGMRLNIPAGTAVRFEPGQSREVELVELAGLRRVYGFAGRVMGEL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01710", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSMAISDPPNSSSVPLEGNGTILVKGNVTIIVEGNADITVKGDATTLVEGNQTNTVNGNLSWKVAGTVDWDVGGDWTEKMASMSSISSGQYTIDGSRIDIGSVEGYIPEAPRDGQAYVRKDGEWVFLSTFLVEHHHHHH\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCBCCEEEESCEEEEESSCEEEEESSCEEEEESSCEEEEESSCEEEEESSCEEEEESSEEEEEESEEEEEESSEEEEECSEEEECCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01711", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nTRUB_FRAT1\n\n### Answer:", "output": "Responsible for synthesis of pseudouridine from uracil-55 in the psi GC loop of transfer RNAs.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01712", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMLPKDKLNPPTPSIYLENKRDAFFPPLHQFCTNPKNPVTVIRGLAGALKLDLGLFSTKTLVEANNEHMVEVRTQLLQPADENWDPTGTKKIWRCESNRSHTTIAKYAQYQASSFQESLREENEKRTQHKDHSDNESTSSENSGRRRKGPFKTIKFGTNIDLSDNKKWKLQLHELTKLPAFARVVSAGNLLTHVGHTILGMNTVQLYMKVPGSRTPGHQENNNFCSVNINIGPGDCEWFVVPEDYWGVLNDFCEKNNLNFLMSSWWPNLEDLYEANVPVYRFIQRPGDLVWINAGTVHWVQAVGWCNNIAWNVGPLTACQYKLAVERYEWNKLKSVKSPVPMVHLSWNMARNIKVSDPKLFEMIKYCLLKILKQYQTLREALVAAGKEVIWHGRTNDEPAHYCSICEVEVFNLLFVTNESNTQKTYIVHCHDCARKTSKSLENFVVLEQYKMEDLIQVYDQFTLALSLSSSSAENLYFQ\n\n### Answer:", "output": "<SEQ_2D>CCCCGGGSCCCCEEECCSHHHHHCHHHHHHHHCTTCSEEEEETHHHHHTCCGGGGSHHHHHHHCTTCEEEEEEEECCCTTCCBCTTSSCBCSBCCCEEEEEEHHHHHHHHHHHHHHHHHHHTTCCCCCCCCCCCCCCCCCCCCCCCCCCCCEEEEEEEEECCCTTTSHHHHHHGGGSCGGGSSSCTTCTTTTTCSCCTTTSSCEEEEECTTCEEEEECCGGGCEEEEEEEEESCEEEEEECGGGHHHHHHHHHTTTCCTTTSCBCCCHHHHHHTTCCCEEEEECTTCEEEECTTCEEEEEESSSEEEEEEEECCSSHHHHHHHHHHHHHHHHTTCCCSSCHHHHHHHHHHHCCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHHTTCCEEECCCCTTCCCEECTTTCCEECSEEEEEHHHHHTTCCCCBCHHHHHHHSTTSTTEEEEECSCHHHHHHHHHTCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01713", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BQ69-N174A on protein ADP-ribose glycohydrolase MACROD1.", "input": "### Wild-type:\nMSLQSRLSGRLAQLRAAGQLLVPPRPRPGHLAGATRTRSSTCGPPAFLGVFGRRARTSAGVGAWGAAAVGRTAGVRTWAPLAMAAKVDLSTSTDWKEAKSFLKGLSDKQREEHYFCKDFVRLKKIPTWKEMAKGVAVKVEEPRYKKDKQLNEKISLLRSDITKLEVDAIVNAANSSLLGGGGVDGCIHRAAGPLLTDECRTLQSCKTGKAKITGGYRLPAKYVIHTVGPIAYGEPSASQAAELRSCYLSSLDLLLEHRLRSVAFPCISTGVFGYPCEAAAEIVLATLREWLEQHKDKVDR\n\n### Mutant:\nMSLQSRLSGRLAQLRAAGQLLVPPRPRPGHLAGATRTRSSTCGPPAFLGVFGRRARTSAGVGAWGAAAVGRTAGVRTWAPLAMAAKVDLSTSTDWKEAKSFLKGLSDKQREEHYFCKDFVRLKKIPTWKEMAKGVAVKVEEPRYKKDKQLNEKISLLRSDITKLEVDAIVNAAASSLLGGGGVDGCIHRAAGPLLTDECRTLQSCKTGKAKITGGYRLPAKYVIHTVGPIAYGEPSASQAAELRSCYLSSLDLLLEHRLRSVAFPCISTGVFGYPCEAAAEIVLATLREWLEQHKDKVDR\n\n### Answer:", "output": "Slightly reduced ADP-ribosyl hydrolase activity; when associated with A-184. Reduces O-acetyl-ADP-ribose deacetylase activity by 93%; when associated with A-184. No significant effect on affinity for substrate.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01714", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAKKSKQLRAALEKIDSTKAYSVEEAVALAKETNFAKFDATVEVAYNLNIDVKKADQQIRGAMVLPNGTGKTQRVLVFARGAKAEEAKEAGADFVGEDDLVQKINGGWLDFDVVIATPDMMAIVGRLGRVLGPRNLMPNPKTGTVTMDVAKAVEESKGGKITYRADKAGNVQTIIGKVSFDADKLVENFKAFNDVIAKAKPATAKGTYITNLVLTTTQGPGIKVDANSF\n\n### Sequence 2:\nMAKKSKQLRAALEKIDSTKAYSVEEAVALAKETNFAKFDATVEVAYNLNIDVKKADQQIRGAMVLPNGTGKTQRVLVFARGAKAEEAKEAGADFVGEDDLVQKINGGWLDFDVVIATPDMMAIVGRLGRVLGPRNLMPNPKTGTVTMDVAKAVEESKGGKITYRADKAGNVQTIIGKVSFDADKLVENFKAFNDVIAKAKPATAKGTYITNLVLTTTQGPGIKVDANSF\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01715", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nPTSPFETLRAAAAPRYFGAALGVPHLLNFTHDPLFDVTAVLQFNGATPENEMKWAYIEPERNQFNFTGGDIVAAFSAANDYVLRGHNLVWYQELAPWVETLTGEDLWNATVNHITTVMTHYKESFNIYAWDVVNEAFNDNGTYRENVWYTQLGPDYIPNAYAVARSVNTPSKLYINDYNTEGINNKSDALLAVVQSMKAHNLVDGVGFQCHFFVGELPPDLEQNFARFVAAGVEIAVTELDIRMNLPPSQADIEQQARDYATVVNACKAQGAACVGITTWGITDLYSWIPSTYPGEGYALLFDDNYVPHPAFNATIQALLA\n\n### Answer:", "output": "<SEQ_2D>CCCTTCSHHHHHTTSEEEEEECHHHHTTTTTCHHHHHHHHHHCSEEEESSTTSHHHHCSBTTBCCCHHHHHHHHHHHHTTCEEEEEEEECSSSCCTTGGGCCHHHHHHHHHHHHHHHHHHHHHHSCEEEEEEEECCBCTTSSBCCCHHHHHHCTTHHHHHHHHHHHTCCSCEEEEEESSCSSSSHHHHHHHHHHHHHHHTTCCCEEEECCEEETTCCCTTHHHHHHHHHHTTCEEEEEEEEEEEESSCCHHHHHHHHHHHHHHHHHHHTTGGGEEEEEESCSBGGGCSHHHHSTTEECCSSBCTTSCBCHHHHHHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01716", "category": "Mutation", "instruction": "Describe the effect of mutation O14862-E147A on protein Interferon-inducible protein AIM2.", "input": "### Wild-type:\nMESKYKEILLLTGLDNITDEELDRFKFFLSDEFNIATGKLHTANRIQVATLMIQNAGAVSAVMKTIRIFQKLNYMLLAKRLQEEKEKVDKQYKSVTKPKPLSQAEMSPAASAAIRNDVAKQRAAPKVSPHVKPEQKQMVAQQESIREGFQKRCLPVMVLKAKKPFTFETQEGKQEMFHATVATEKEFFFVKVFNTLLKDKFIPKRIIIIARYYRHSGFLEVNSASRVLDAESDQKVNVPLNIIRKAGETPKINTLQTQPLGTIVNGLFVVQKVTEKKKNILFDLSDNTGKMEVLGVRNED\n\n### Mutant:\nMESKYKEILLLTGLDNITDEELDRFKFFLSDEFNIATGKLHTANRIQVATLMIQNAGAVSAVMKTIRIFQKLNYMLLAKRLQEEKEKVDKQYKSVTKPKPLSQAEMSPAASAAIRNDVAKQRAAPKVSPHVKPEQKQMVAQQESIRAGFQKRCLPVMVLKAKKPFTFETQEGKQEMFHATVATEKEFFFVKVFNTLLKDKFIPKRIIIIARYYRHSGFLEVNSASRVLDAESDQKVNVPLNIIRKAGETPKINTLQTQPLGTIVNGLFVVQKVTEKKKNILFDLSDNTGKMEVLGVRNED\n\n### Answer:", "output": "Strongly reduced ability to homooligomerize upon double-stranded DNA (dsDNA)-binding.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01717", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYCEI_SALTY\n\n### Answer:", "output": "Protein YceI.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01718", "category": "Mutation", "instruction": "Describe the effect of mutation P52597-R116A on protein Heterogeneous nuclear ribonucleoprotein F.", "input": "### Wild-type:\nMMLGPEGGEGFVVKLRGLPWSCSVEDVQNFLSDCTIHDGAAGVHFIYTREGRQSGEAFVELGSEDDVKMALKKDRESMGHRYIEVFKSHRTEMDWVLKHSGPNSADSANDGFVRLRGLPFGCTKEEIVQFFSGLEIVPNGITLPVDPEGKITGEAFVQFASQELAEKALGKHKERIGHRYIEVFKSSQEEVRSYSDPPLKFMSVQRPGPYDRPGTARRYIGIVKQAGLERMRPGAYSTGYGGYEEYSGLSDGYGFTTDLFGRDLSYCLSGMYDHRYGDSEFTVQSTTGHCVHMRGLPYKA\n\n### Mutant:\nMMLGPEGGEGFVVKLRGLPWSCSVEDVQNFLSDCTIHDGAAGVHFIYTREGRQSGEAFVELGSEDDVKMALKKDRESMGHRYIEVFKSHRTEMDWVLKHSGPNSADSANDGFVRLAGLPFGCTKEEIVQFFSGLEIVPNGITLPVDPEGKITGEAFVQFASQELAEKALGKHKERIGHRYIEVFKSSQEEVRSYSDPPLKFMSVQRPGPYDRPGTARRYIGIVKQAGLERMRPGAYSTGYGGYEEYSGLSDGYGFTTDLFGRDLSYCLSGMYDHRYGDSEFTVQSTTGHCVHMRGLPYKA\n\n### Answer:", "output": "Decreases affinity for RNA oligonucleotide 100-fold.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01719", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTKTVFVLNGPNLNLLGKREPGIYGVATLDDIEASCKREAGQLELQIDFRQSNHEGDLVSWIQEAGEKNAYVLINPAAYSHTSVAIHDAIRSARVTVVEVHLSNIHAREAFRHHSHVSAVAKGVICGFGAEGYLLGLRALAAIAKEEENNGQSIKGA\n\n### Sequence 2:\nMPSTIFVLNGPNLNALGKREPGIYGGKTLADIEAMCKEEAKLLGFDVDFRQSNHEGTLVDWLHEAGEKAVGIAINPAAYGHTSIAMHDAIRAITVPVVELHLSNIHAREEFRHKSMIAPAVKGVICGFGAQSYILALHALKNLTEKSK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01720", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nSPP1 APOE CD206 MRC1 CD163\n</cell>", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as SPP1, APOE, MRC1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "colon", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["SPP1", "APOE", "CD206", "MRC1", "CD163"]}}
{"id": "eval_01721", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAKIGVEPSLTDVQEALKERGYDVVPLRGEQDARGCDCCVITGLDANIAGIHNIVTSGPVIEASGLTAEEICQKVEEKLR\n\n### Sequence 2:\nMGMRMMFTVFLLVVLATTVVSFPSDRASDGRDDEAKDERSDMHESDRKEICCNPACGPKYSCGR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01722", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNMTVEWSPQRTRATEQGQPLLRLAGVSRRFMAGDREFLALKDINLQIRAGELVAIIGASGSGKSTLMNILGCLDNASSGSYQVNGQETRDLDDDALAALRRDHFGFIFQRYHLLPHLDALRNVEIPAVYAGVPQAGRHQRARELLTRLGLAGHLANRPSQLSGGQQQRVSICRALMNGGEVILADEPTGALDTASGKEVMNILLELHAAGHTVILVTHDPKVAAHAERIIEVSDGQIVDDRRTVREVAKPVEEVQPAAERAPRRLVASLGLFREAFKMAWIALISHRMRTLLTMLGIIIGITSVVSISAIGEGAKGYVLKDIQAIGSNTIDIYSGKSFGDSRAKAIETLSPADVTALNELYYVDSATPVIGQGSLVRYRNIDADVQLNGVSEQYFQVRNIPLAEGIVFSADDARRQAQVVVIDHNTRKRLFAPGVEALGQVILVGSLPCTVIGVTAENKNLFVAGNSLNVWMPYETAAGRVLGQRHLDSISVRIKDGLPSKRVEEEVNKLMLQRHGTKDFFTNNLDSIMQTVQKTSRSLTLLLSLIAVISLVVGGIGVMNIMLVSVTERTREIGIRMAVGARQSDIRQQFLVEAVMVCLIGGVIGIGLSYGIGYLFALFVKQWEMVFSLGSIVTAFVCSTLIGIVFGFVPARNAARLDPIEALARD\n\n### Sequence 2:\nMIAHLLCILSYNFNTSVILNVYSKLTMFCTTNTLPMDLLLKQGSLKQEVESFCYQIVSESNDQKVGILQSENEQLQPSVSKKSEGELSRVKFMSSSNKITTFSKKPKRRKYDESYLSFGFTYFGNRDAPHAQCVLCKKILSNSSLAPSKLRRHLETKHAAYKDKDISFFKQHLDSPENNKPPTPKIVNTDNESATEASYNVSYHIALSGEAHTIGELLIKPCAKDVVMRMFDEQYSKKIDAVQLSNSTVARRIKDLAADIEEELVCRLKICDGFSLQLDESADVSGLAVLLVFVRYRFNKSIEEDLLLCESLQSNATGEEIFNCINSFMQKHEIEWEKCVDVCSDASRAMDGKIAEAVTLIKYVAPESTSSHCLLYRHALAVKIMPTSLKNVLDQAVQIINYIKARPHQSRLLKILCEEMGAQHTALLLNTEVRWLSRGKVLVRLFELRRELLVFMDSAFRLSDCLTNSSWLLRLAYLADIFTKLNEVNLSMQGKNVTVFTVFDKMSSLLRKLEFWASSVEEENFDCFPTLSDFLTEINSTVDKDICSAIVQHLRGLRSTLLKYFPVTNDNNTWVRNPFTVTVKPASLVARDYESLIDLTSDSQVKQNFSELSLNDFWSSLIQEYPSIARRAVRVLLPFATMHLCETGFSYYAATKTKYRKRLDAAPHMRIRLSNITPNIKRICDKKTQKHCSH\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01723", "category": "Mutation", "instruction": "Describe the effect of mutation Q86XR7-C117H on protein TIR domain-containing adapter molecule 2.", "input": "### Wild-type:\nMGIGKSKINSCPLSLSWGKRHSVDTSPGYHESDSKKSEDLSLCNVAEHSNTTEGPTGKQEGAQSVEEMFEEEAEEEVFLKFVILHAEDDTDEALRVQNLLQDDFGIKPGIIFAEMPCGRQHLQNLDDAVNGSAWTILLLTENFLRDTWCNFQFYTSLMNSVNRQHKYNSVIPMRPLNNPLPRERTPFALQTINALEEESRGFPTQVERIFQESVYKTQQTIWKETRNMVQRQFIA\n\n### Mutant:\nMGIGKSKINSCPLSLSWGKRHSVDTSPGYHESDSKKSEDLSLCNVAEHSNTTEGPTGKQEGAQSVEEMFEEEAEEEVFLKFVILHAEDDTDEALRVQNLLQDDFGIKPGIIFAEMPHGRQHLQNLDDAVNGSAWTILLLTENFLRDTWCNFQFYTSLMNSVNRQHKYNSVIPMRPLNNPLPRERTPFALQTINALEEESRGFPTQVERIFQESVYKTQQTIWKETRNMVQRQFIA\n\n### Answer:", "output": "Loss of ability to dimerize. Loss of RANTES-inducing activity and ability to induce NF-kappa-B activation. Inhibition of TLR4-dependent activation of IRF3 and IRF7. Loss of interaction with TLR4.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01724", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nELP1_MOUSE\n\n### Answer:", "output": "Component of the elongator complex which is required for multiple tRNA modifications, including mcm5U (5-methoxycarbonylmethyl uridine), mcm5s2U (5-methoxycarbonylmethyl-2-thiouridine), and ncm5U (5-carbamoylmethyl uridine) (By similarity).  The elongator complex catalyzes the formation of carboxymethyluridine in the wobble base at position 34 in tRNAs (PubMed:23717213).  Regulates the migration and branching of projection neurons in the developing cerebral cortex, through a process depending on alpha-tubulin acetylation (PubMed:22854966).  ELP1 binds to tRNA, mediating interaction of the elongator complex with tRNA (By similarity).  May act as a scaffold protein that assembles active IKK-MAP3K14 complexes (IKKA, IKKB and MAP3K14/NIK) (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01725", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMTIENDYAKLKELMEFPAKMTFKVAGINREGLAQDLIQVVQKYIKGDYIPKEKRSSKGTYNSVSIDIIAENFDQVETLYKELAKVEGVKMVI\n\n### Sequence 2:\nMKIKDAVNMIRWKYREKIDDYVVIIIDRLTENGLKEISFSEIDDVDNNYLYLKSEENTVIPLHRVLMIKRKSDNALIWKR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01726", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSRVAKAPVAIPAGVEVTLNEQTITVKGTKGSLTRVINADVSVVVEDNEIKCSSVEGVKTAAQAGTARALINNMVVGVTAGFEKKLQLIGVGYRAKIAGKGVDLTLGFSHPLVHELPDGVTAECPSQTEIVLKGTDKQLIGQVAAEIRGYRPPEPYKGKGVRYADEQVRRKEAKKK\n\n### Sequence 2:\nMFRGATLVNLDSKGRLSVPTRYREQLLENAAGQMVCTIDIHHPCLLLYPLPEWEIIEQKLSRLSSMNPVERRVQRLLLGHASECQMDGAGRLLIAPVLRQHAGLTKEVMLVGQFNKFELWDETTWHQQVKEDIDAEQLATGDLSERLQDLSL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01727", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRLUB_XYLFT\n\n### Answer:", "output": "Ribosomal large subunit pseudouridine synthase B, 23S rRNA pseudouridine(2605) synthase, rRNA pseudouridylate synthase B, rRNA-uridine isomerase B.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01728", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPMATTIEVADYTNIMPITVSITVYLGVSIGIDTSTTGFTCFSRY\n\n### Sequence 2:\nMPMATTIDGTDYTNIMPITVFTTVYLGVFIGIDTSTTGFTCFSRY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01729", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRECO_TOLAT\n\n### Answer:", "output": "DNA repair protein RecO, Recombination protein O.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01730", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTSIGTGYDLSNSVFSPDGRNFQVEYAVKAVENGTTSIGIKCNDGVVFAVEKLITSKLLVPQKNVKIQVVDRHIGCVYSGLIPDGRHLVNRGREEAASFKKLYKTPIPIPAFADRLGQYVQAHTLYNSVRPFGVSTIFGGVDKNGAHLYMLEPSGSYWGYKGAATGKGRQSAKAELEKLVDHHPEGLSAREAVKQAAKIIYLAHEDNKEKDFELEISWCSLSETNGLHKFVKGDLLQEAIDFAQKEINGDDDEDEDDSDNVMSSDDENAPVATNANATTDQEGDIHLE\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCTTSSTTCCCTTSCCHHHHHHHHHHHTSCCEEEEEETTEEEEEEEEECCCTTBCTTCSCCCEEETTTEEEEEEECHHHHHHHHHHHHHHHHHHHHHHSSCCCHHHHHHHHHHHHHHTTSCTTBCCCCEEEEEEEEETTEEEEEEECTTSCEEEBSEEEESTTHHHHHHHHHHHHHHCTTCCCHHHHHHHHHHHHHHHGGGGCSSCEEEEEEEEETTTTTTCCEEECHHHHHHHHHHHHHHHCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01731", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "MYH6, TNNT2, MYH7, TNNI3", "output": "The expression pattern is consistent with cardiomyocyte in a context associated with rheumatoid arthritis, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "cardiomyocyte", "tissue": "skin", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["MYH6", "TNNT2", "MYH7", "TNNI3"]}}
{"id": "eval_01732", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nIRF7:low TCF4:medium CLEC4C:high\n</cell>", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as CLEC4C, IRF7, TCF4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "brain", "disease": "normal", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["IRF7", "TCF4", "CLEC4C"]}}
{"id": "eval_01733", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "CD34, PROM1, KIT", "output": "This profile is most consistent with hematopoietic stem cell, which is commonly found in kidney under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "hematopoietic stem cell", "tissue": "kidney", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CD34", "PROM1", "KIT"]}}
{"id": "eval_01734", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMIQHAIAKLLEGEDLSRAEAAQVMTEIADGGATPAQSGAFLAALRMKGETVEEIAGAADVMRQRADRVRVSRDVFIDTCGTGGDGRHTFNISTTAAFVAAGAGVCVAKHGNRAVSSRSGSADVLAALGVNVDADKETVERCIEEVGIGFLFAVRLHPAFKAIAGVRRELGVRTIFNLLGPLANPAGARHQVLGVYEARWVPVLGGVLAALGAAHAFVVHGEGLDEIAVTGMTHVCEVKDGAVERYTIRPEDLGLPRRDAAELAGGDAAANARIVTHVLEGQQGGPRDAVLANAAAALVCAGAATDLRDGVARAARSIDSGAAREKLRQLVAATTVPA\n\n### Sequence 2:\nMFSQTLSKIINRKDLSREEMDRIFSDIFSGNLTDAQIGAFMAALATKGETFEELAGAAEAMRRKATRIQAASPVVVDTCGTGGDGAHTFNISTTSAFVVAGAGICVAKHGNRSVSSKCGSADVLEALGVKLDTQPEVAEEAVNEIGIGFLFAPLFHGAMKYAIVPRKELGVRTIFNMLGPLTNPAAANCQVLGVFAPQLTEMFADALNLLGARRAFVVHGHDGLDEISVCASTRVSELNDGRVQTYDISPEHFFEDRARPEDMAGGTPSENAQITRHILSGKEKGPRRNVVVVNAGAALVAAGKAEDLKAGVALAGQIIDSGKAHEKLEQLIEFT\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01735", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "CN1CC[C@@]2(C)c3cc(OC(=O)N4CCc5ccccc5C4)ccc3N(C)[C@@H]12", "output": "The molecule contains the following functional groups: aromatic ring, ester, amide, amine, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "BBBP", "canonical_smiles": "CN1CC[C@@]2(C)c3cc(OC(=O)N4CCc5ccccc5C4)ccc3N(C)[C@@H]12", "compound_name": null, "pubchem_cid": null, "MW": 377.49, "LogP": 3.61, "TPSA": 36.02, "QED": 0.76, "functional_groups": ["aromatic ring", "ester", "amide", "amine", "ether"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "ester", "amide", "amine", "ether"]}}
{"id": "eval_01736", "category": "Mutation", "instruction": "Describe the effect of mutation P04637-P152S on protein Cellular tumor antigen p53.", "input": "### Wild-type:\nMEEPQSDPSVEPPLSQETFSDLWKLLPENNVLSPLPSQAMDDLMLSPDDIEQWFTEDPGPDEAPRMPEAAPPVAPAPAAPTPAAPAPAPSWPLSSSVPSQKTYQGSYGFRLGFLHSGTAKSVTCTYSPALNKMFCQLAKTCPVQLWVDSTPPPGTRVRAMAIYKQSQHMTEVVRRCPHHERCSDSDGLAPPQHLIRVEGNLRVEYLDDRNTFRHSVVVPYEPPEVGSDCTTIHYNYMCNSSCMGGMNRRPILTIITLEDSSGNLLGRNSFEVRVCACPGRDRRTEEENLRKKGEPHHELP\n\n### Mutant:\nMEEPQSDPSVEPPLSQETFSDLWKLLPENNVLSPLPSQAMDDLMLSPDDIEQWFTEDPGPDEAPRMPEAAPPVAPAPAAPTPAAPAPAPSWPLSSSVPSQKTYQGSYGFRLGFLHSGTAKSVTCTYSPALNKMFCQLAKTCPVQLWVDSTPSPGTRVRAMAIYKQSQHMTEVVRRCPHHERCSDSDGLAPPQHLIRVEGNLRVEYLDDRNTFRHSVVVPYEPPEVGSDCTTIHYNYMCNSSCMGGMNRRPILTIITLEDSSGNLLGRNSFEVRVCACPGRDRRTEEENLRKKGEPHHELP\n\n### Answer:", "output": "In sporadic cancers; somatic mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01737", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMASLKLWVCLVLLLVLELTSVHECRPLVAEERFSGSSRLKKIRRELFERLKEMKGRSEGEETILGNTLDSKRLSPGGPDPRHH\n\n### Sequence 2:\nMASFKLWVCLIL--LLLEFSVHQCRPLVAEESPSDSGNIRKIMRELLKRSEELKVRSKDGQTVLG-TLDSKRLSPGGPDPRHH\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01738", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSMSDVTIVKEGWVQKRGEYIKNWRPRYFLLKTDGSFIGYKEKPQDVDLPYPLNNFSVAKCQLMKTERPKPNTFIIRCLQWTTVIERTFHVDTPEEREEWTEAIQAVADRLQRQEEERMN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCEEEEEEEEEECSSSCCEEEEEEEEETTSEEEEESSSCCGGGTTCCSEEEECTTCEEEEECSSSTTEEEEEEEETTEEEEEEEECSSHHHHHHHHHHHHHHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01739", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMLAVEVSEAEGGIREWEEEEKEEEAAERTPTGEKSPNSPRTLLSLRGKARTGGPMEVKLELHPLQNRWALWFFKNDRSRAWQDNLHLVTKVDTVEDFWALYSHIQLASKLSSGCDYALFKDGIQPMWEDSRNKRGGRWLVSLAKQQRHIELDRLWLETLLCLIGESFEEHSREVCGAVVNIRTKGDKIAVWTREAENQAGVLHVGRVYKERLGLSPKTIIGYQAHADTATKSNSLAKNKFVV\n\n### Sequence 2:\nMSVLDLNALNALPKVERILALAETNAKLEKLSAEERVAWALENLPGEYALSSSFGIQAAVSLHLVNQLRPDIPVILTDTGYLFPETYQFIDELTDKLKLNLKVYRAQESAAWQEARYGKLWEQGVEGIEKYNEINKVEPMNRALKELNTQTWFAGLRREQSGSRATLPVLAIQRGVFKVLPIIDWDNRTVYQYLQKHGLKYHPLWDQGYLSVGDTHTTRKWEPGMAEEETRFFGLKRECGLHEG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01740", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "SFTPA1, ABCA3, SFTPC, SFTPB", "output": "This cell is most consistent with alveolar type II cell, a pulmonary alveolar type II cell. Key markers such as SFTPC, SFTPA1, ABCA3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "colon", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["SFTPA1", "ABCA3", "SFTPC", "SFTPB"]}}
{"id": "eval_01741", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMPEKDTGGQHRATRRTEEHDETIDEATATSDVQERREKLDADVDAILDEIDDVLEENAEEFVRSYIQKGGE\n\n### Sequence 2:\nMPEKDTGGQHRATRRTEEHDETIDEATATSDVQERREKLDADVDAILDEIDDVLEENAEEFVRSYIQKGGE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01742", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCCCCN(CCCOC)C(=O)C(CCC(=O)O)NC(=O)c1ccc(Cl)c(Cl)c1", "output": "This molecule is medium-sized and polar, with lipophilic character. It contains aromatic ring, amide, hydroxyl, ether, halide. Its descriptor profile: MW=461.39, LogP=4.01, TPSA=95.94, HBD=2, HBA=4, QED=0.41.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CCCCCN(CCCOC)C(=O)C(CCC(=O)O)NC(=O)c1ccc(Cl)c(Cl)c1", "compound_name": null, "pubchem_cid": null, "MW": 461.39, "LogP": 4.01, "TPSA": 95.94, "QED": 0.41, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "ether", "halide"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": null, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": 0.0, "SR-ATAD5": null, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": null}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01743", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMMAAGTALGLALWLLLPPVGVGGAGPPPIQDGEFTFLLPAGRKQCFYQSAPANASLETEYQVIGGAGLDVDFSLESPQGVLLVSESRKADGVHTVEPTEAGDYKLCFDNSFSTISEKLVFFELIFDSLQDEEEVEGWAEAVEPEEILEVKMEDIKESIETMRIRLERSIQVLTLLRAFEARDRNLQEGNLERVNFWSAVNVAVLLLVAVLQVCTLKRFFQDKRPVPM\n\n### Sequence 2:\nMYHVIAATTNPAKINAITLAFDDVYGPGQYRIEGVNVDSGVPLQPIGSTETRIGARQRVKNARQVRPEADFWVGIEAGIEDNMTFAWMVIEHLQARGESRSASLMLPDIILQGIRQGRELGDEMAVLSGISNVKQQGGAIGIFTQGKLTRTSVYHQALLLALVPFHNEIYQRPSPSKPAI\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01744", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: VEGFR2\nLigand SMILES: O=CNC(N1C=CN(C(NC=O)C(Cl)(Cl)Cl)C=C1)C(Cl)(Cl)Cl\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "ESOL", "canonical_smiles": "O=CNC(N1C=CN(C(NC=O)C(Cl)(Cl)Cl)C=C1)C(Cl)(Cl)Cl", "compound_name": null, "pubchem_cid": null, "MW": 430.93, "LogP": 2.43, "TPSA": 64.68, "QED": 0.5, "functional_groups": ["amide", "amine", "halide"], "moleculenet_labels": {"label": -4.19}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_01745", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nYIDC_MYCGE\n\n### Answer:", "output": "Required for the insertion and/or proper folding and/or complex formation of integral membrane proteins into the membrane.  Involved in integration of membrane proteins that insert both dependently and independently of the Sec translocase complex, as well as at least some lipoproteins.  Aids folding of multispanning membrane proteins (By similarity).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01746", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "MAFA, INS, IAPP", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as IAPP, MAFA, INS support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "skin", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["MAFA", "INS", "IAPP"]}}
{"id": "eval_01747", "category": "Mol", "instruction": "Identify the compound name from the given SMILES.", "input": "CN1C(=O)C2(CC(C)(C)Oc3ccc(-c4cncnc4)cc32)N=C1N", "output": "The SMILES represents an unnamed molecule with MW=337.38, LogP=1.69, TPSA=93.7. PubChem lookup would be required to retrieve a canonical compound name or synonyms.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "compound_lookup", "source": "BACE", "canonical_smiles": "CN1C(=O)C2(CC(C)(C)Oc3ccc(-c4cncnc4)cc32)N=C1N", "compound_name": null, "pubchem_cid": null, "MW": 337.38, "LogP": 1.69, "TPSA": 93.7, "QED": 0.85, "functional_groups": ["aromatic ring", "amide", "amine", "ether"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_compound_lookup", "name": null, "cid": null}}
{"id": "eval_01748", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTVILGIDPGSRITGYGIVRQEGCKLTYLGSGCIRTKVDDLPARLKLIYAGVSEIITQFQPDCLAIEQVFMAKNADSALKLGQARGVAIVVAMNLDLPVFEYAARQVKQTVVGSGAAEKSQVQHMVRTLLKLPANPQADAADALAIAITHCHVSQNALRMGSGGLNLARGRLR\n\n### Sequence 2:\nMSRLQEQYKNKISLILQKELGMSNPMQTPKIEKITINMGLGNALGDKKILQSGLEEMSLISGQKPLTCNARKSVASFKLREGNPIGCKVTLRKQKMYEFLDRLVNITIPRIRDFRGLKTTAFDGRGNYNMGITEQITFPEIDFEKVTKIRGMDIAITTTAKSDEDAKKLLAMFKFPFKG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01749", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nPCAY_PSEP1\n\n### Answer:", "output": "Methyl-accepting chemotaxis protein PcaY, Aromatic acid chemoreceptor PcaY.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01750", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "DCN, PDGFRA, COL1A1, COL3A1", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as COL3A1, PDGFRA, COL1A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "lymph node", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["DCN", "PDGFRA", "COL1A1", "COL3A1"]}}
{"id": "eval_01751", "category": "Mutation", "instruction": "Describe the effect of mutation O95999-E53R on protein B-cell lymphoma/leukemia 10.", "input": "### Wild-type:\nMEPTAPSLTEEDLTEVKKDALENLRVYLCEKIIAERHFDHLRAKKILSREDTEEISCRTSSRKRAGKLLDYLQENPKGLDTLVESIRREKTQNFLIQKITDEVLKLRNIKLEHLKGLKCSSCEPFPDGATNNLSRSNSDESNFSEKLRASTVMYHPEGESSTTPFFSTNSSLNLPVLEVGRTENTIFSSTTLPRPGDPGAPPLPPDLQLEEEGTCANSSEMFLPLRSRTVSRQ\n\n### Mutant:\nMEPTAPSLTEEDLTEVKKDALENLRVYLCEKIIAERHFDHLRAKKILSREDTREISCRTSSRKRAGKLLDYLQENPKGLDTLVESIRREKTQNFLIQKITDEVLKLRNIKLEHLKGLKCSSCEPFPDGATNNLSRSNSDESNFSEKLRASTVMYHPEGESSTTPFFSTNSSLNLPVLEVGRTENTIFSSTTLPRPGDPGAPPLPPDLQLEEEGTCANSSEMFLPLRSRTVSRQ\n\n### Answer:", "output": "Abolished homomultimerization and formation of a CBM complex, abolished ability to activate NF-kappa-B.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01752", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CCCCCCCCCCCCCCOS(=O)(=O)O", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains hydroxyl. Its descriptor profile: MW=294.46, LogP=4.51, TPSA=63.6, HBD=1, HBA=3, QED=0.38.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "SIDER", "canonical_smiles": "CCCCCCCCCCCCCCOS(=O)(=O)O", "compound_name": null, "pubchem_cid": null, "MW": 294.46, "LogP": 4.51, "TPSA": 63.6, "QED": 0.38, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"Hepatobiliary disorders": 0, "Metabolism and nutrition disorders": 0, "Product issues": 0, "Eye disorders": 0, "Investigations": 0, "Musculoskeletal and connective tissue disorders": 0, "Gastrointestinal disorders": 1, "Social circumstances": 0, "Immune system disorders": 1, "Reproductive system and breast disorders": 0, "Neoplasms benign, malignant and unspecified (incl cysts and polyps)": 0, "General disorders and administration site conditions": 1, "Endocrine disorders": 0, "Surgical and medical procedures": 0, "Vascular disorders": 0, "Blood and lymphatic system disorders": 0, "Skin and subcutaneous tissue disorders": 1, "Congenital, familial and genetic disorders": 0, "Infections and infestations": 0, "Respiratory, thoracic and mediastinal disorders": 1, "Psychiatric disorders": 0, "Renal and urinary disorders": 0, "Pregnancy, puerperium and perinatal conditions": 0, "Ear and labyrinth disorders": 0, "Cardiac disorders": 0, "Nervous system disorders": 1, "Injury, poisoning and procedural complications": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01753", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSKKGGYKRTGVDVRYGDESVARFINAVMLDGKKDVATKIVYDAFAIIGEKMTEETPLEVYHRAMSNIAPVVEVRSKRVGGATYQIPMEVKPSRRGALAFRWLKQYATKRGGRSMAEKLAAELMDAAGEQGASVKKRDEVHRMADANKAFAHFRF\n\n### Sequence 2:\nMKKLIAKTKEELIEKIKECENEEEIYINLELDRDVAIALLESCEPKKIYLPKSKYKRSSKKIIKALEGVDVEVIPIKAKTGRPTNVDKIIKKYLDKKPKEIAEITGINIKTVEYHYYKLKKKEKQ\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01754", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nVGFKAGVKDYKLTYYTPSYETKDTDILAAFRVTPQPGVPPEEAGAAVAAESSTGTWTTVWTDGLTSLDRYKGRCYHIEPVAGEETQFIAYVAYPLDLFEEGSVTNMFTSIVGNVFGFKALRALRLEDLRIPTAYVKTFQGPPHGIQVERDKLNKYGRPLLGCTIKPKLGLSAKNYGRAVYECLPGGLDFTKDDENVNSQPFMRWRDRFLFCAEAIFKSQAETGEIKGHYLNATAGTCEEMMKRAVFARELGVPIVMHDYLTGGFTANTTLAHYCRDNGLLLHIHRAMHAVIDRQKKHGMHFRVLAKALRMSGGDHIHAGTVVGKLEGEREITLGFVDLLRDDFIEKDRSRGLFFTQDWVSLPGVLPVASGGIHVWHMPALTEIFGDDSVLQFGGGTLGHPWGNAPGAVANRVALEACVQARNEGRDLAREGNDIIREASKWSPELAAACEVWKEIKFEFEAMDTL\n\n### Sequence 2:\nMVNTSEENNSNAVMPRKFQGGMNQVGHGGGRIVGQNRRALGGINQNFVHGRPYPCVVHKRVLSEKHEICEKKQADLGHRPITRRFAAKIAGSQQSYAEKTKNSNPLNLNEFGNSIAIDDELKSPEDQPEPMTLEHTEPMHSDPLEMEEVEMEDIEGEMILDIDSCDANNSLAVVEYIEDLHAYYRKIEYLGCVSPTYMDEQLDLNERMRAILVDWLIEVHDKFDLMQETLFLTVNLIDRFLAKQNVVRKKLQLVGLVAMLLACKYEEVSVPVVSDLIHIADRAYTRKDILEMEKLMLNTLQYNMSLPTAYVFMRRFLKAAQADKKLELVAFFLVDLSLVEYEMLKFPPSLVAAAAVYTAQCTVSGFKHWNKTCEWHTNYSEDQLLECSMLMVGFHQKAGAGKLTGVHRKYGSAKFSFTAKCEPACFLLENKNQP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01755", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMPYVGVGAQTVSTSLTGAPMVKAYIAIAASLIFVFCIAALGVHHSERKFNKFNKVSIDDIHKSDAGVIQDNIKTENIKKYLRIFTKDPHVAGTEANKKVAYEIANAWSEAGLEDVHTLPYEVLLSYPDFENPNSVIIKSSAGKEVFKSKGVSPVIIPDEQSGKYAGHQWLAYAGNGSASADVVYINHGTANDFKNLKLMGVDIKGKIALMRYGHGFRGDKIHKAQQAGAIGAILFSDTQDVAQDGVESENVYPKKIWMPNEGVQRGSLMHGDGDALSPYYPSKKELFKGRTIEEAKEDGVLPSIPVLPVSYTTGYEILKRLSGRPAPSDWQGFVGGNLTYKLGPGFVNGEKLSINVHSELRTKRIRNVIGYIRGSEEPDSYIMLGNHFDAWVYGSIDPNSGTAVLAEVARAMMQTINETSWKPARTIVFNAWDAEEFGLIGSTEFVEEFVNILQKRAVVYINMDCIQGNISLHVDTVPILEHAVIEASKQVENPSKRERSRGRKTLYDTWMKVFPDKKAGVPKIRVPGGGSDHAPFLNFAGVPVINFTFKNYTTWDTYPLYHTMYETPFSNIHLLDTDNLSVHKAIGQYWAELAKTFADDVILPMNTTHFASVMLKTYLPQLKTTISGINVSRSDFEDIRTQYALLSKSAQDLLTMSKKFQETIHFTQHSFSQNPYDPKHVNAVNERLKSTERCFINPRGVSMHNPSARHVLFSVSDSDSYSSSLMAGVQNAINSYDLNPTKKGLREIINQISIVQYSVICVVNTLRDVI\n\n### Sequence 2:\nMSSVEELTQLFSQVGFEDKKVKEIVKNKKVSDSLYKLIKETPSDYQWNKSTRALVHNLASFVKGTDLPKSELIVNGIINGDLKTSLQVDAAFKYVKANGEASTKMGMNENSGVGIEITEDQVRNYVMQYIQENKERILTERYKLVPGIFADVKNLKELKWADPRSFKPIIDQEVLKLLGPKDERDLIKKKTKNNEKKKTNSAKKSSDNSASSGPKRTMFNEGFLGDLHKVGENPQAYPELMKEHLEVTGGKVRTRFPPEPNGYLHIGHSKAIMVNFGYAKYHNGTCYLRFDDTNPEKEAPEYFESIKRMVSWLGFKPWKITYSSDYFDELYRLAEVLIKNGKAYVCHCTAEEIKRGRGIKEDGTPGGERYACKHRDQSIEQNLQEFRDMRDGKYKPGEAILRMKQDLNSPSPQMWDLIAYRVLNAPHPRTGTKWRIYPTYDFTHCLVDSMENITHSLCTTEFYLSRESYEWLCDQVHVFRPAQREYGRLNITGTVLSKRKIAQLVDEKFVRGWDDPRLFTLEAIRRRGVPPGAILSFINTLGVTTSTTNIQVVRFESAVRKYLEDTTPRLMFVLDPVEVVVDNLSDDYEELATIPYRPGTPEFGERTVPFTNKFYIERSDFSENVDDKEFFRLTPNQPVGLIKVSHTVSFKSLEKDEAGKIIRIHVNYDNKVEEGSKPKKPKTYIQWVPISSKYNSPLRVTETRVYNQLFKSENPSSHPEGFLKDINPESEVVYKESVMEHNFGDVVKNSPWVVDSVKNSEFYVEEDKDSKEVCRFQAMRVGYFTLDKESTTSKVILNRIVSLKDATSK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01756", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMIDLNALCSECGVKFYADELISENSRQIYRVYITKKNGVNLDDCEALSRLLSPILDVEPPTSGAFTLEVSSPGLERKLTTPSNFENSLGELVKITLKNGEKISGEILSFKDEILKLKTAENNEISVNFNEIKKAKTYIEW\n\n### Sequence 2:\nMNLEALCKEAELSFYDDELVSENGKKIYRIYVQKEGGVNLDDCARLSEILSPIFDVEPPVNGEYFLEVSSPGLERKLSKIEHFAKSIGELVKITTNEKEKIEAKIIAVDDENITLENLENKEKTTINFNDIKKARTFMEW\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01757", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAFKHQLILSTAILLAVLAAASASFREQCVPGREITYESLNARREYAVRQTCGYYLSAERQKRRCCDELSKVPELCWCEVLRILMDRRVTKEGVVKDSLLQDMSRCKKLTREFIAGIVGRE\n\n### Sequence 2:\nMAFKHQLILSTAILLAVLAAASASFREQCVPGREITYESLNARREYAVRQTCGYYLSAERQKRRCCDELSKVPELCWCEVLRILMDRRVTKEGVVKDSLLQDMSRCKKLTREFIAGIVGRE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01758", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMKKLLTVMTMAVLTAGTLLLPAQSVTPAAHAVQISNSERELPFKAKHAYSTISQLSEAIGPRIAGTAAEKKSALLIASSMRKLKLDVKVQRFNIPDRLEGTLSSAGRDILLQAASGSAPTEEQGLTAPLYNAGLGYQKDFTADAKGKIALISRGDLTYYEKAKNAEAAGAKAVIIYNNKESLVPMTPNLSGNKVGIPVVGIKKEDGEALTQQKEATLKLKAFTNQTSQNIIGIKKPKNIKHPDIVYVTAHYDSVPFSPGANDNGSGTSVMLEMARVLKSVPSDKEIRFIAFGAEELGLLGSSHYVDHLSEKELKRSEVNFNLDMVGTSWEKASELYVNTLDGQSNYVWESSRTAAEKIGFDSLSLTQGGSSDHVPFHEAGIDSANFIWGDPETEEVEPWYHTPEDSIEHISKERLQQAGDLVTAAVYEAVKKEKKPKTIKKQMKAKASDIFEDIK\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCHHHHHHHHHHHHTTTCCCCTTSHHHHHHHHHHHHHHHHTTCEEEEEEEECCCEEEEEEEETTEEECEEECTTSCCCCTTCEEEEEEECTTCCGGGCCGGGTTSEEEEECSSSCHHHHHHHHHHTTCSEEEEECSSCCSSCCCCCCTTCCCSSCEEEECHHHHHHHTTCSEEEEEEEEECSCEEEEEEEEECCTTCSSCCEEEEEEECCCCTTCCCTTTTHHHHHHHHHHHHHHTTCCCSSEEEEEEESCGGGTSHHHHHHHHTSCHHHHHHEEEEEEECSCCBCCTTCCEEEEEETTSCCCHHHHHHHHHHHHHTCCCEEEEECCSSTHHHHHHTTCEEEEEEEECTTTCCCCTTTTSTTCSGGGBCHHHHHHHHHHHHHHHHHHHCCCCCCCCCCCCCCCCGGGTCSCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01759", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSGPAKECEKDQFQCRNERCIPSVWRCDEDDDCLDHSDEDDCPKKTCADSDFTCDNGHCIHERWKCDGEEECPDGSDESEATCTKQVCPAEKLSCGPTSHKCVPASWRCDGEKDCEGGADEAGCATSLGTCRGDEFQCGDGTCVLAIKHCNQEQDCPDGSDEAGCLQGLNECLHNNGGCSHICTDLKIGFECTCPAGFQLLDQKTCGDIDECKDPDACSQICVNYKGYFKCECYPGYEMDLLTKNCKAAAGKSPSLIFTNRHEVRRIDLVKRNYSRLIPMLKNVVALDVEVATNRIYWCDLSYRKIYSAYMDKASDPKEQEVLIDEQLHSPEGLAVDWVHKHIYWTDSGNKTISVATVDGGRRRTLFSRNLSEPRAIAVDPLRGFMYWSDWGDQAKIEKSGLNGVDRQTLVSDNIEWPNGITLDLLSQRLYWVDSKLHQLSSIDFSGGNRKTLISSTDFLSHPFGIAVFEDKVFWTDLENEAIFSANRLNGLEISILAENLNNPHDIVIFHELKQPRAPDACELSVQPNGGCEYLCLPAPQISSHSPKYTCACPDTMWLGPDMKRCYRASS\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCTTEEECTTSCEEEGGGTTSSCCCSSSSTTTTTCCSCSCCTTCEECSSCCEECGGGTTSSSCCSTTCGGGSSSSCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCSSSGGGGCSSEEECCSSSCEEECCSSCCEESSSCEECCCTTSSTTSSSSEEEECSSCEEEECCTTEECCTTTTCCEECSSCCCEEEEECBSCEEEEETTTTEEEEEECCCSSEEEEEEETTTTEEEEEETTTTEEEEEETTTTTSGGGCEEEECSSCSCCCEEEEETTTTEEEEEETTTTEEEEEETTSSCEEEEECSSCSSEEEEEEETTTTEEEEEECSSSCEEEEEETTSCCCEEEECSSCSCEEEEEEETTTTEEEEEETTTTEEEEEETTSCCCEEEECCTTTCSSEEEEEEETTEEEEEETTTTEEEEEETTTSCCCEEEECCCSSCCCEEEECGGGSCCCCCHHHHSSSGGGGCSSEEEECCCSSSSSCSEEEECCTTCEECTTSSCEECCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01760", "category": "Mutation", "instruction": "Describe the effect of mutation B2GUB3-K24A on protein Tubulin tyrosine ligase 3.", "input": "### Wild-type:\nMAHHTAVNPDRLKHAKALVEKAIKQKKIFAIHGPYPVIRSCLRSRGWVEKKFPKSGKAKQKKEKASDEDMEDDDGDGSSNDDDDGENSDEEENGDPDGTCDLMSRLLRNEDPNFFWTTKRDAVDCRFLKKDQMLNHYAKAGSFTTKVGLCLNLRNLHWFDDADPDSFFPRCYRLGAEDEKQSFKEDFWHTAARSILKRVANRRDICSPAATGGAKASHREPGANNGAQLLAKRGSRKRAESVPVQIILTALEACERYLNSLEHNDIDMETEATPAMTDTQWEEFLHGYYQVIHDGATIEH\n\n### Mutant:\nMAHHTAVNPDRLKHAKALVEKAIAQKKIFAIHGPYPVIRSCLRSRGWVEKKFPKSGKAKQKKEKASDEDMEDDDGDGSSNDDDDGENSDEEENGDPDGTCDLMSRLLRNEDPNFFWTTKRDAVDCRFLKKDQMLNHYAKAGSFTTKVGLCLNLRNLHWFDDADPDSFFPRCYRLGAEDEKQSFKEDFWHTAARSILKRVANRRDICSPAATGGAKASHREPGANNGAQLLAKRGSRKRAESVPVQIILTALEACERYLNSLEHNDIDMETEATPAMTDTQWEEFLHGYYQVIHDGATIEH\n\n### Answer:", "output": "Decreased monoglycylation activity; when associated with A-26. The site specificity of TTLL enzymes and their biochemical interplay remain largely unknown.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01761", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "C[C@@H]1CC[C@H]2C(C)(C)[C@H]3C[C@@]12CC[C@@]3(C)O", "output": "The molecule contains the following functional groups: hydroxyl.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "C[C@@H]1CC[C@H]2C(C)(C)[C@H]3C[C@@]12CC[C@@]3(C)O", "compound_name": null, "pubchem_cid": null, "MW": 222.37, "LogP": 3.61, "TPSA": 20.23, "QED": 0.66, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 1.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["hydroxyl"]}}
{"id": "eval_01762", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKKIILTIIDGLGLRKERQGNAYLQAKHPCFDYLFSMCPNSVLQASGQYVGLPEGQIGNSEVGHLNIGAGRVVYTGLSLINKAIENNTFKDNEILNDVIDKTIKNNTTLHVMGLLSNGGVHSLDLHLFEILKLAHSKGLKNVSVHVFGDGRDVKPQSIKNSLETLKDLCQKFGYKISSISGRFYAMDRDSIFSRNQEAYDAILGQSKNVIENIDDYIESQYKKGIFDEFFEPAQLKDGVFVKNGDSIIFFNFRPDRARQLSHMFIGSNLYTYKPKNQVQIDNFVSLMKYEGINSKIAFKEMEVVNPLGKVLESNDIKQLRLAETQKYAHVTFFFDGGVDIEYKNENRILVDSLKVDSFADYPHMSAKEITDSLLNNIEKNDFIIMNYANPDMVGHTGNLNATIEAIEFLDSQFQRILEYVSLNHENVTWFITADHGNAEITEDENNKPATKHTTSPVMFICTDKNVNLGNGSLCDVAPTILDYLKINKPKEMTGKSLLK\n\n### Sequence 2:\nMSVIKEYRTASEVVGPLMIVEQVNNVSYNELVEIQLHNGEIRRGQVLEIHEDKAMVQLFEGSSGINLEKSKIRFAGHALELAVSEDMVGRIFNGMGKPIDGGPDLIPEKYLDIDGQAINPVSRDYPDEFIQTGISSIDHLNTLVRGQKLPVFSGSGLPHNELAAQIARQATVLNSDENFAVVFAAMGITFEEAEFFMEELRKTGAIDRSVLFMNLANDPAIERIATPRIALTAAEYLAFEKDMHVLVIMTDMTNYCEALREVSAARREVPGRRGYPGYLYTNLSTLYERAGRLVGKKGSVTQIPILTMPEDDITHPIPDLTGYITEGQIILSHELYNQGYRPPINVLPSLSRLKDKGSGEGKTRGDHAPTMNQLFAAYAQGKKVEELAVVLGESALSDVDKLYVRFTKRFEEEYINQGFYKNRNIEDTLNLGWELLSILPRTELKRIKDDLLDKYLPLVEV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01763", "category": "Mutation", "instruction": "Describe the effect of mutation Q99558-T852I on protein Mitogen-activated protein kinase kinase kinase 14.", "input": "### Wild-type:\nMAVMEMACPGAPGSAVGQQKELPKAKEKTPPLGKKQSSVYKLEAVEKSPVFCGKWEILNDVITKGTAKEGSEAGPAAISIIAQAECENSQEFSPTFSERIFIAGSKQYSQSESLDQIPNNVAHATEGKMARVCWKGKRRSKARKKRKKKSSKSLAHAGVALAKPLPRTPEQESCTIPVQEDESPLGAPYVRNTPQFTKPLKEPGLGQLCFKQLGEGLRPALPRSELHKLISPLQCLNHVWKLHHPQDGGPLPLPTHPFPYSRLPHPFPFHPLQPWKPHPLESFLGKLACVDSQKPLPDPH\n\n### Mutant:\nMAVMEMACPGAPGSAVGQQKELPKAKEKTPPLGKKQSSVYKLEAVEKSPVFCGKWEILNDVITKGTAKEGSEAGPAAISIIAQAECENSQEFSPTFSERIFIAGSKQYSQSESLDQIPNNVAHATEGKMARVCWKGKRRSKARKKRKKKSSKSLAHAGVALAKPLPRTPEQESCTIPVQEDESPLGAPYVRNTPQFTKPLKEPGLGQLCFKQLGEGLRPALPRSELHKLISPLQCLNHVWKLHHPQDGGPLPLPTHPFPYSRLPHPFPFHPLQPWKPHPLESFLGKLACVDSQKPLPDPH\n\n### Answer:", "output": "In an ovarian mucinous carcinoma sample; somatic mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01764", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAPSRKFFVGGNWKMNGRKKNLGELITTLNAAKVPADTEVVCAPPTAYIDFARQKLDPKIAVAAQNCYKVTNGAFTGEISPGMIKDCGATWVVLGHSERRHVFGESDELIGQKVAHALSEGLGVIACIGEKLDEREAGITEKVVFEQTKVIADNVKDWSKVVLAYEPVWAIGTGKTATPQQAQEVHEKLRGWLKSNVSDAVAQSTRIIYGGSVTGATCKELASQPDVDGFLVGGASLKPEFVDIINAKQ\n\n### Answer:", "output": "<SEQ_2D>CCCCCEEEEEECCBCCCHHHHHHHHHHHHHSCCCTTEEEEEECCGGGHHHHHHHSCTTSEEEESCCCSSSSBSCTTCCCHHHHHHTTCCEEEESCHHHHHTSCCCHHHHHHHHHHHHHTTCEEEEEECCCHHHHHTTCHHHHHHHHHHHHHTTCSCGGGEEEEECCGGGSSSSCCCCHHHHHHHHHHHHHHHHHHTCHHHHHHSEEEECSCCCTTTHHHHHTSTTCCEEEESGGGGSTHHHHHHTTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01765", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSMLPEAEVRLGQAGFMQRQFGAMLQPGVNKFSLRMFGSQKAVEREQERVKSAGFWIIHPYSDFRFYWDLTMLLLMVGNLIIIPVGITFFKDENTTPWIVFNVVSDTFFLIDLVLNFRTGIVVEDNTEIILDPQRIKMKYLKSWFMVDFISSIPVDYIFLIVETRIDSEVYKTARALRIVRFTKILSLLRLLRLSRLIRYIHQWEEIFHMTYDLASAVVRIVNLIGMMLLLCHWDGCLQFLVPMLQDFPDDCWVSINNMVNNSWGKQYSYALFKAMSHMLCIGYGRQAPVGMSDVWLTMLSMIVGATCYAMFIGHATALIQSLDSSRRQYQEKYKQVEQYMSFHKLPPDTRQRIHDYYEHRYQGKMFDEESILGELSEPLREEIINFNCRKLVASMPLFANADPNFVTSMLTKLRFEVFQPGDYIIREGTIGKKMYFIQHGVVSVLTKGNKETKLADGSYFGEICLLTRGRRTASVRADTYCRLYSLSVDNFNEVLEEYPMMRRAFETVALDRLDRIGKKNS\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCTTTTTSTTSCCSSTTHHHHSSSHHHHHHHHHHHHTTSTTCBCTTSSHHHHHHHHHHHHHHHHHHHHHHHHHTCSSCCSHHHHHHHHHHHHHHHHHHHHTTSBCEECTTCSSEECCHHHHHHHHHHSSSHHHHHHHSCHHHHHHHHHHHHCGGGTTSSHHHHHHHHHHHHGGGGGGGHHHHHHHHHHHHHHSHHHHSSCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTTCCTTSHHHHTTCSSSCHHHHHHHHHHHHHHHHTTCCCSSSCCCSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTSSHHHHHHHHHHHHHHHHHHHHTCCHHHHHHHHHHHHHHTTTCCCCHHHHHHHSCHHHHHHHHHHHTHHHHHHCTTGGGSCHHHHHHHHTTCEEEEECTTCEEECTTSCCCEEEEEEESCEEEECTTSCCEEECTTCEECCHHHHHCCCCCSEEEECSCEEEEEEEHHHHHHHHHHCTTSHHHHHHHHHHHHHTCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01766", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSYSGERDNFAPHMALVPMVIEQTSRGERSFDIYSRLLKERVIFLTGQVEDHMANLIVAQMLFLEAENPEKDIYLYINSPGGVITAGMSIYDTMQFIKPDVSTICMGQAASMGAFLLTAGAKGKRFCLPNSRVMIHQPLGGYQGQATDIEIHAREILKVKGRMNELMALHTGQSLEQIERDTERDRFLSAPEAVEYGLVDSILTHRN\n\n### Sequence 2:\nMSYSGERDNLAPHMALVPMVIEQTSRGERSFDIYSRLLKERVIFLTGQVEDHMANLIVAQMLFLEAENPEKDIYLYINSPGGVITAGMSIYDTMQFIKPDVSTICMGQAASMGAFLLTAGAKGKRFCLPNSRVMIHQPLGGYQGQATDIEIHAREILKVKGRMNELMAHHTGQSLEQIERDTERDRFLSAPEAVEYGLVDSILTHRN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01767", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CN1c2c(O)nc(N)nc2NCC1CNc1ccc(C(=O)NC(CCC(=O)[O-])C(=O)[O-])cc1\nDescriptors: MW=457.45, LogP=-2.51, TPSA=208.69, HBD=5, HBA=12", "output": "Descriptors: MW=457.45, LogP=-2.51, TPSA=208.69, HBD=5, HBA=12, QED=0.26, Lipinski violations=1. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "CN1c2c(O)nc(N)nc2NCC1CNc1ccc(C(=O)NC(CCC(=O)[O-])C(=O)[O-])cc1", "compound_name": null, "pubchem_cid": null, "MW": 457.45, "LogP": -2.51, "TPSA": 208.69, "QED": 0.26, "functional_groups": ["aromatic ring", "amide", "hydroxyl", "phenol"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01768", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nMRC1 CD206 SPP1 APOE CD163\n</cell>", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as APOE, MRC1, CD206 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "blood", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MRC1", "CD206", "SPP1", "APOE", "CD163"]}}
{"id": "eval_01769", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nTCDDYYYGFGCNKFCRPRDDFFGHYACDQNGNKTCMEGWMGPECNRAICRQGCSPKHGSCKLPGDCRCQYGWQGLYCDKCIPHPGCVHGICNEPWQCLCETNWGGQLCDKDLNYCGTHQPCLNGGTCSNTGPDKYQCSCPEGYSGPNCEI\n\n### Answer:", "output": "<SEQ_2D>CCCTTEESTTSCEECCCEEETTEEEEECSSSCEEECTTEESTTSCEECCCTTCCTTTEECSSTTCCEECTTEESTTSCEECCSTTCSSEECSSSSCCEECTTEETTTTCEETTHHHHHCCCCTTCEEEEEETTEEEEECCTTCCSTTSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01770", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSEYQYNKVTPEMIEKFKEIAPKRVLVGDEINEDFTHDEMAIYGKARPEVLVEATSTEEVAAVVKLCNENKIPVTPSGARTGLVGGAVSIGGGVMISLTKMNKILGYDKENFVVKIQSGVLLNDLAQDAEKQGLLYPPDPGEKFATVGGNVATNAGGMRAVKYGCTRDYVRAMTVVLPTGEIVKLGATVSKTSSGYSLLNLMIGSEGTLGIITELTLKVIPAPKSVISLIIPYENLEDCIATVPQFFMHHLAPQALEFMEKEVVMDTEKFLGKQVYPKELEGTEIGAYLLATFDGNSEEQLEDIIEQASEVVLEAGAIDVLVADTPALKKDAWAVRGALLEAIEADTVLLDECDVVVPTNKIAEFLTYTKSLEAEADFRVKSFGHAGDGNLHIYTCSNDMEEGEFKKQVAVFMDKVYAKATEFGGMISGEHGIGHGKMDYLAESLGPVQMRIMEGVKEVFDPNMILNPGKICYKL\n\n### Sequence 2:\nMSEYQYNKVTPEMIEKFKEIAPKRVLVGDEINEDFTHDEMAIYGKARPEVLVEATSTEEVAAVVKLCNENKIPVTPSGARTGLVGGAVSIGGGVMISLTKMNKILGYDKENFVVKIQSGVLLNDLAQDAEKQGLLYPPDPGEKFATVGGNVATNAGGMRAVKYGCTRDYVRAMTVVLPTGEIVKLGATVSKTSSGYSLLNLMIGSEGTLGIITELTLKVIPAPKSVISLIIPYENLEDCIATVPQFFMHHLAPQALEFMEKEVVMDTEKFLGKQVYPKELEGTEIGAYLLATFDGNSEEQLEDIIEQASEVVLEAGAIDVLVADTPALKKDAWAVRGALLEAIEADTVLLDECDVVVPTNKIAEFLTYTKSLEAEADFRVKSFGHAGDGNLHIYTCSNDMEEGEFKKQVAVFMDKVYAKATEFGGMISGEHGIGHGKMDYLAESLGPVQMRIMEGVKEVFDPNMILNPGKICYKL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01771", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSTKTNSEILEQLKQASDGLLFMSESEYPFEVFLWEGSAPPVTHEIVLQQTGHGQDAPFKVVDIDSFFSRATTPQDWYEDEENAVVAKFQKLLEVIKSNLKNPQVYRLGEVELDVYVIGETPAGNLAGISTKVVET\n\n### Answer:", "output": "<SEQ_3Di>DPDDPVVLLVVLCVLQPPAWDDAVHTKGKDWDWDPDDAQPDDVVVVCVVVVDDPVWDKDKDAPCVVCVVQQDDDPPDDVVSVVSNVSSVVSVVSCPVFFAPWIWMWIHPQKTKIWTWGAGPVRTIIGMIIIHGDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01772", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "Cc1sc2ncnc(SCC(=O)N3CCCC3)c2c1C", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains amide. Its descriptor profile: MW=307.44, LogP=3.02, TPSA=46.09, HBD=0, HBA=5, QED=0.65.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "Cc1sc2ncnc(SCC(=O)N3CCCC3)c2c1C", "compound_name": null, "pubchem_cid": null, "MW": 307.44, "LogP": 3.02, "TPSA": 46.09, "QED": 0.65, "functional_groups": ["amide"], "moleculenet_labels": {"label": 2.89}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01773", "category": "Mol", "instruction": "Evaluate whether this molecule is compatible with oral-drug-like descriptors.", "input": "SMILES: CC(=O)OCCOC(C)=O\nDescriptors: MW=146.14, LogP=0.11, TPSA=52.6, HBD=0, HBA=4", "output": "Descriptors: MW=146.14, LogP=0.11, TPSA=52.6, HBD=0, HBA=4, QED=0.42, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "Tox21", "canonical_smiles": "CC(=O)OCCOC(C)=O", "compound_name": null, "pubchem_cid": null, "MW": 146.14, "LogP": 0.11, "TPSA": 52.6, "QED": 0.42, "functional_groups": ["ester", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01774", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nMS4A7 FCGR3A CDKN1C LST1\n</cell>", "output": "This cell is most consistent with CD16+ monocyte, a non-classical CD16+ monocyte. Key markers such as CDKN1C, FCGR3A, LST1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "prostate", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["MS4A7", "FCGR3A", "CDKN1C", "LST1"]}}
{"id": "eval_01775", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPKRVLQGVVVSDKTDKTIVVKVERRFTHPVMKKTVRRSKNYHAHDEANAAKIGQTVFIEESRPYSKTKTWKLVEDQAAAAEAAGTAA\n\n### Sequence 2:\nMPKRVLQGVVVSDKTDKTIVVKVERRFTHPVMKKTVRRSKNYHAHDEANTAKVGQTVFIEESRPYSKTKTWKLVEDQAAAAEAAGTAA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01776", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "FCGR3B, S100A9, CXCR2, S100A8, CSF3R", "output": "This cell is most consistent with neutrophil, a mature neutrophil. Key markers such as S100A8, CSF3R, S100A9 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "intestine", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["FCGR3B", "S100A9", "CXCR2", "S100A8", "CSF3R"]}}
{"id": "eval_01777", "category": "Mutation", "instruction": "Describe the effect of mutation P0ABQ0-D210N on protein Coenzyme A biosynthesis bifunctional protein CoaBC.", "input": "### Wild-type:\nMSLAGKKIVLGVSGGIAAYKTPELVRRLRDRGADVRVAMTEAAKAFITPLSLQAVSGYPVSDSLLDPAAEAAMGHIELGKWADLVILAPATADLIARVAAGMANDLVSTICLATPAPVAVLPAMNQQMYRAAATQHNLEVLASRGLLIWGPDSGSQACGDIGPGRMLDPLTIVDMAVAHFSPVNDLKHLNIMITAGPTREPLDPVRYISDHSSGKMGFAIAAAAARRGANVTLVSGPVSLPTPPFVKRVDVMTALEMEAAVNASVQQQNIFIGCAAVADYRAATVAPEKIKKQATQGDEL\n\n### Mutant:\nMSLAGKKIVLGVSGGIAAYKTPELVRRLRDRGADVRVAMTEAAKAFITPLSLQAVSGYPVSDSLLDPAAEAAMGHIELGKWADLVILAPATADLIARVAAGMANDLVSTICLATPAPVAVLPAMNQQMYRAAATQHNLEVLASRGLLIWGPDSGSQACGDIGPGRMLDPLTIVDMAVAHFSPVNDLKHLNIMITAGPTREPLDPVRYISNHSSGKMGFAIAAAAARRGANVTLVSGPVSLPTPPFVKRVDVMTALEMEAAVNASVQQQNIFIGCAAVADYRAATVAPEKIKKQATQGDEL\n\n### Answer:", "output": "Gain of activity, reaction intermediate undetectable. Using the mutant protein H-CoaB the putative acyl-cytidylate intermediate of 4'-phosphopantothenate was undetectable. Two more inactive-site residues, R206 and A276, were identified and shown to be uninvolved in the second half reaction of the (R)-4'-phospho-N-pantothenoylcysteine synthetase.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01778", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMVKSFYGYVRDAWKNPDETYVNELRWERLQVWRKQGSVTRIERPTRIDRARSLGYKAKQGIVVVRVNVRRGGLGHVRPNRGRRTQKMGKNKVSGGMSIQRIAEVRADRRYPNLEVLNSYWVGEDGKHKWFEVILVDPHHPVIKSDKNLNWVCDPSSRGRATRGKTSAGRKGRGMATRGKGTEKTRPSIRAYKSRGK\n\n### Sequence 2:\nMSKSFYGYIRDAWKNPDETYVRDLRWERLQVWRKEGSVTRVERPTRIDRARSLGYKAKQGIVVARVKVRRGSMRKSRYIRGRRTQHTGKNKITVGKSIQRISEERAARKYPNMEVLNSYWVGEDGKQKWYEVILVDPSHPVIKSDKNLNWICGKAHSGRVFRGKTSAGRKGRGMQTRGTGTEKTRPSVRSNLNRSK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01779", "category": "Mutation", "instruction": "Describe the effect of mutation Q99755-L223I on protein Phosphatidylinositol 4-phosphate 5-kinase type-1 alpha.", "input": "### Wild-type:\nMASASSGPSSSVGFSSFDPAVPSCTLSSAASGIKRPMASEVLEARQDSYISLVPYASGMPIKKIGHRSVDSSGETTYKKTTSSALKGAIQLGITHTVGSLSTKPERDVLMQDFYVVESIFFPSEGSNLTPAHHYNDFRFKTYAPVAFRYFRELFGIRPDDYLYSLCSEPLIELCSSGASGSLFYVSSDDEFIIKTVQHKEAEFLQKLLPGYYMNLNQNPRTLLPKFYGLYCVQAGGKNIRIVVMNNLLPRSVKMHIKYDLKGSTYKRRASQKEREKPLPTFKDLDFLQDIPDGLFLDADM\n\n### Mutant:\nMASASSGPSSSVGFSSFDPAVPSCTLSSAASGIKRPMASEVLEARQDSYISLVPYASGMPIKKIGHRSVDSSGETTYKKTTSSALKGAIQLGITHTVGSLSTKPERDVLMQDFYVVESIFFPSEGSNLTPAHHYNDFRFKTYAPVAFRYFRELFGIRPDDYLYSLCSEPLIELCSSGASGSLFYVSSDDEFIIKTVQHKEAEFLQKLLPGYYMNLNQNPRTLIPKFYGLYCVQAGGKNIRIVVMNNLLPRSVKMHIKYDLKGSTYKRRASQKEREKPLPTFKDLDFLQDIPDGLFLDADM\n\n### Answer:", "output": "Decreased 1-phosphatidylinositol-4-phosphate 5-kinase activity with 1-octadecanoyl-2-(5Z,8Z,11Z,14Z)-eicosatetraenoyl-sn-glycero-3-phospho-1D-myo-inositol 4-phosphate (arachidonate-PtdIns4P) as substrate. Increased enzyme activation by diarachidonoyl phosphatidic acid (DAPA). No change in enzyme activation by 1-stearoyl-2-oleoyl phosphatidic acid (SOPA) or 1-stearoyl-2-arachidonoyl phosphatidic acid (SAPA).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01780", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFABP4_CAEEL\n\n### Answer:", "output": "Fatty acid-binding protein homolog 4.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01781", "category": "Mutation", "instruction": "Describe the effect of mutation Q99KG5-A308S on protein Lipolysis-stimulated lipoprotein receptor.", "input": "### Wild-type:\nMAPAASACAGAPGSHPATTIFVCLFLIIYCPDRASAIQVTVPDPYHVVILFQPVTLHCTYQMSNTLTAPIVIWKYKSFCRDRVADAFSPASVDNQLNAQLAAGNPGYNPYVECQDSVRTVRVVATKQGNAVTLGDYYQGRRITITGNADLTFEQTAWGDSGVYYCSVVSAQDLDGNNEAYAELIVLGRTSEAPELLPGFRAGPLEDWLFVVVVCLASLLFFLLLGICWCQCCPHTCCCYVRCPCCPDKCCCPEALYAAGKAATSGVPSIYAPSIYTHLSPAKTPPPPPAMIPMRPPYGYP\n\n### Mutant:\nMAPAASACAGAPGSHPATTIFVCLFLIIYCPDRASAIQVTVPDPYHVVILFQPVTLHCTYQMSNTLTAPIVIWKYKSFCRDRVADAFSPASVDNQLNAQLAAGNPGYNPYVECQDSVRTVRVVATKQGNAVTLGDYYQGRRITITGNADLTFEQTAWGDSGVYYCSVVSAQDLDGNNEAYAELIVLGRTSEAPELLPGFRAGPLEDWLFVVVVCLASLLFFLLLGICWCQCCPHTCCCYVRCPCCPDKCCCPEALYAAGKAATSGVPSIYAPSIYTHLSPAKTPPPPPAMIPMRPPYGYP\n\n### Answer:", "output": "Increased phosphorylation levels.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01782", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMPKAENKKSRAAEIVTREYTINLHKRLHGVGFKKRAPRAVKEIKKFASKIMGTTDVRVDPRLNKFVWNQGIRSVPYRVRVRLARKRNDDEDAKEKLYTLVTYVPVTSFKGLQTQNVDAE\n\n### Sequence 2:\nKGEKKKGRSAINEVVTREYTINVHKRIHGVGFKKRAPRAIKEIRKFAVKEMGTPDVRIDTRLNKAVWSKGVRNVPYRIRVRLSRKRNEDEDSPNKLYTLVTYVPVTTCKGLQTVNVD\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01783", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nHHHHHHMNTDVLNQQKIEELSAEIGSDNVPVLLDIFLGEMDSYIGTLTELQGSEQLLYLKEISHALKSSAASFGADRLCERAIAIDKKAKANQLQEQGMETSEMLALLHITRDAYRSWTN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCTTTTTTHHHHHTHHHHHHHHHHHHHHHHHHHHHHHHHTTTTSSHHHHHHHHHHHHHHHHHHHHHTTTHHHHHHHHHHHHHHHHHHCTTTSTTTTTTHHHHHHHHHHHHHCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01784", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCD27 CD79A IGHA1 MS4A1\n</cell>", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as CD27, CD79A, MS4A1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "lymph node", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD27", "CD79A", "IGHA1", "MS4A1"]}}
{"id": "eval_01785", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSKTHFGFETVEETDKAKKVAGVFHSVANNYDLMNDLMSAGMHRAWKAFTIAQANVRPGYKVLDIAAGTGDLTKAFAKAAGPTGEVWHTDINESMLRVGRDRLLDKGIVTPSLLCDAEKIPFPDNYFDVVTVAFGLRNMTHKDVALAEMRRVAKPGGRVMVLEFSKVWEPLKKAYDLYSFKVLPWLGDKFAKDADSYRYLAESIRMHPDQDTLKTMMEQAGLDAVKYYNLSGGVVALHLGTKY\n\n### Sequence 2:\nMSKTHFGFETVEETDKAKKVAGVFHSVANNYDLMNDLMSAGMHRAWKAFTIAQANVRPGYKVLDIAAGTGDLTKAFAKAAGPTGEVWHTDINESMLRVGRDRLLDKGIVTPSLLCDAEKIPFPDNYFDVVTVAFGLRNMTHKDVALAEMRRVAKPGGRVMVLEFSKVWEPLKKAYDLYSFKVLPWLGDKFAKDADSYRYLAESIRMHPDQDTLKTMMEQAGLDAVKYYNLSGGVVALHLGTKY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01786", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "CCR7, CD3E, SELL, CD3D, CD8B", "output": "This profile is most consistent with naive CD8+ T cell, which is commonly found in kidney under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "kidney", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CCR7", "CD3E", "SELL", "CD3D", "CD8B"]}}
{"id": "eval_01787", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nMUTS2_BACLD\n\n### Answer:", "output": "Endonuclease MutS2.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01788", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSHO1_CANTT\n\n### Answer:", "output": "High osmolarity signaling protein SHO1, Osmosensor SHO1.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01789", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nPOA1_YARLI\n\n### Answer:", "output": "Belongs to the POA1 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01790", "category": "Cell", "instruction": "Identify the cell type from the given marker gene expression profile.", "input": "<cell>\nGCG:high TTR:medium IRX2:medium\n</cell>", "output": "This cell is most consistent with pancreatic alpha cell, a pancreatic alpha cell secreting glucagon. Key markers such as IRX2, GCG, TTR support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic alpha cell", "tissue": "colon", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["GCG", "TTR", "IRX2"]}}
{"id": "eval_01791", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTDNKYLTVTDLNYYITQKFKNDPYLHKVFLQGELSNFRYRRNSHQYFSLKDEKSKINVVMFRSYFDKVKFKPEEGMKVYVVGYVSVYGPQGSYQFYAENMEPAGLGALYEQLKQLQAKLAKEGLFNPEHKRKLPRFPDKIAVVTSASGAVIHDIMVTANRRFPHAEIDLFPAQVQGDTAADSLVKAMRQIAAQGDEYDVMIIGRGGGSLEDLWSFNEEEVVRQVYAMPMPVISSVGHETDTTLCDLVADARAATPTAAAEYATPNLADELAGIHQLQSRLIAGMQNNINQKRDRLNRIKNSVIMREPTRLYDQQIQQVDLLKQRLQNSIQNKVNNSLQNYRLLKQRLLSVSPDKQIAGMEQQEKFLAKQLNDNMKHYLKNKRSDFSKIVQQLDDYSPLKTLERGFVYTTNEEGETVSSVDQIKKDDHLNLHFKDGEVTATVTEVKK\n\n### Sequence 2:\nQYLTVTALTQYLKRKFEVDPYLGKVYLTGEVSNYRPRPNTHQYFSLKDDHAKISAIMFKSAFAKVKFQPEEGMKVLVVGRIGLYEPSGSYQIYVERMEPDGVGALYQAYEQLKKKLAAEGLFSAP-KKPLPRFPKRIAVVTSRSGAVIRDIITTTRRRFPIAQIVLFPSQVQGDAAAAEISRQIERANAQGD-FDTLIIGRGGGSIEDLWPFNEEVVARAIAQSQLPVISSVGHETDTTIADLVADVRAATPTAAAELAVPVYNDVLLQLKQDQTRVFNAFQNFVQRDRQRLNKLQTSYVFTQPNRLYEGYLQKLDFLNERLKQAGQNNFNLASQHYQRVFQQLRQQTPIHQVRQAQTQLLNLQQRLNRGTQLVVRQKRQQLTQTVQSLDLLSPLKIMTRGYAFVTADE-QVVHGVKQLQPEQTVAIHMADGEAQAQITKI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01792", "category": "Cell", "instruction": "Based on the top expressed genes, infer the cell type.", "input": "<cell>\nIGHG1 CD27 CD79A IGHA1\n</cell>", "output": "This cell is most consistent with memory B cell, a memory B cell. Key markers such as IGHG1, CD27, IGHA1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "memory B cell", "tissue": "bone marrow", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["IGHG1", "CD27", "CD79A", "IGHA1"]}}
{"id": "eval_01793", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMRMIRAIKQGQWAVLLWPYLLTASIPLDCRDEQGVLSHCPSISQEKLLDRVIQHAELIYRVSEESCSLFEEMFIPFPLQLQRNQAGYPCITKALPIPSSKSEIQQISDKWLLHSVLMLVQSWIEPLVYLQTTLNRYDGVPDMLLNKTKWVSEKLMSLEQGVVVLIKKMLDEEMMTTTYSEQGLFQDDGQPEMLEYVMRDYTLLSCFKKDAHKMEILLKLLKCRQNDMHSCR\n\n### Sequence 2:\nMIICASRVSHLALQKFRNQIPTMKLHKVRTVNCSADYPDVECNGYTAYDTAQKVIVISFRGTKGPNQNNQIVEGMTRDGLLPYFGNGSGKIFKVLYDSFMLLWNEWYSNKFPSSFHIIHRLDLIPRVPAIDPHTNTTVMFHPRTEVWYNNYMRLNDTYQICEEADGNNCSDAVTEGLNMNDHGFYFDINIANWGKDGCPKNTTGYSQP\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01794", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nUNG_XANC8\n\n### Answer:", "output": "Excises uracil residues from the DNA which can arise as a result of misincorporation of dUMP residues by DNA polymerase or due to deamination of cytosine.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01795", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nDEF_RICAH\n\n### Answer:", "output": "Removes the formyl group from the N-terminal Met of newly synthesized proteins.  Requires at least a dipeptide for an efficient rate of reaction.  N-terminal L-methionine is a prerequisite for activity but the enzyme has broad specificity at other positions.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01796", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nYBJQ_ECO7I\n\n### Answer:", "output": "UPF0145 protein YbjQ.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01797", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "VCAN, LYZ, S100A9, S100A8, CD14, FCN1", "output": "This cell is most consistent with CD14+ monocyte, a classical CD14+ monocyte. Key markers such as S100A9, CD14, VCAN support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "CD14+ monocyte", "tissue": "pancreas", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["VCAN", "LYZ", "S100A9", "S100A8", "CD14", "FCN1"]}}
{"id": "eval_01798", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "CD163, CD68, C1QA, MRC1, C1QB", "output": "This cell is most consistent with macrophage, a tissue macrophage. Key markers such as MRC1, C1QA, CD163 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "macrophage", "tissue": "pancreas", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD163", "CD68", "C1QA", "MRC1", "C1QB"]}}
{"id": "eval_01799", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSPPSIGIAVILVGTSDEVAIKDAHEKDDFHHLSVVPRVELVAMNETDPKSIITRICDLMSDRKIQGVVFADDTDQEAIAQILDFISAQTLTPILGIHGGSSMIMADKDESSMFFQFGPSIEQQASVMLNIMEEYDWYIFSIVTTYFPGYQDFVNKIRSTIENSFVGWELEEVLLLDMSLDDGDSKIQNQLKKLQSPIILLYCTKEEATYIFEVANSVGLTGYGYTWIVPSLVAGDTDTVPAEFPTGLISVSYDEWDYGLPARVRDGIAIITTAASDMLSEHSFIPEPKSSCYNTHEKRIYQSNMLNRYLINVTFEGRDLSFSEDGYQMHPKLVIILLNKERKWERVGKWKDKSLQMKYYVWPRM\n\n### Answer:", "output": "<SEQ_2D>CCCEEEEEEEEESSCCHHHHCCCCCCCCCCCCCSEEEEEEEEECCCSHHHHHHHHHHHHHHSCEEEEEEECSSCCTHHHHHHHHHHHHHTCCEEEEEGGGGSCCSSCCTTCCEEESSCCHHHHHHHHHHHHHHHTCCEEEEEECSSTTHHHHHHHHHHHHHTCSSCCEEEEEEECCTTSCCSSCHHHHHHTTCCCSEEEEESCHHHHHHHHHHHHHTTCSSTTCEEEEEHHHHCSTTCCCTTSCTTCEEEEECTTTSCHHHHHHHHHHHHHHHHHHHHHHHSCCCCCCSCSSSTTTSCTTCCSTTHHHHTSCEETTEECCBCTTSBBSSCCEEEEEECTTSCEEEEEEECSSCEEESCSSCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01800", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSGRGKGGKGLGKGGAKRHRKVLRDNIQGITKPAIRRLARRGGVKRISGLIYEETRGVLKIFLENVIRDSVTYTEHARRKTVTAMDVVYALKRQGRTLYGFGG\n\n### Sequence 2:\nMSGRGKGGKGLGKGGAKRHRRVLRDNIQGITKPAIRRLARRGGVKRISGLIYEETRGVLKVFLENVIRDAVTYTEHARRKTVTAMDVVYALKRQGRTLYGFGG\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01801", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCD1C CLEC10A FCER1A HLA-DQA1\n</cell>", "output": "This cell is most consistent with conventional dendritic cell, a conventional type-2 dendritic cell (cDC2). Key markers such as HLA-DQA1, CD1C, CLEC10A support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "conventional dendritic cell", "tissue": "blood", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CD1C", "CLEC10A", "FCER1A", "HLA-DQA1"]}}
{"id": "eval_01802", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPNP_BORT9\n\n### Answer:", "output": "Involved in mRNA degradation.  Catalyzes the phosphorolysis of single-stranded polyribonucleotides processively in the 3'- to 5'-direction.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01803", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "LILRA4, IRF7, GZMB", "output": "This cell is most consistent with plasmacytoid dendritic cell, a plasmacytoid dendritic cell (pDC). Key markers such as IRF7, LILRA4, GZMB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "plasmacytoid dendritic cell", "tissue": "liver", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["LILRA4", "IRF7", "GZMB"]}}
{"id": "eval_01804", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAGLNTAAKAKGLKYFGSATDNPELTDSAYVAQLSNTDDFGQITPGNSMKWDATEPSQNSFSFANGDAVVNLANKNGQLMRCHTLVWHSQLPNWVSSGSWTNATLLAAMKNHITNVVTHYKGKCYAWDVVNEALNEDGTFRNSVFYQIIGPAYIPIAFATAAAADPDVKLYYNDYNIEYSGAKATAAQNIVKMIKAYGAKIDGVGLQAHFIVGSTPSQSDLTTVLKGYTALGVEVAYTELDIRMQLPSTAAKLAQQSTDFQGVAAACVSTTGCVGVTIWDWTDKYSWVPSVFQGYGAPLPWDENYVKKPAYDGLMAGLGA\n\n### Answer:", "output": "<SEQ_2D>CCHHHHHHHTTCSEEEEEECGGGGGCHHHHHHHTCTTTCSEEEESSTTSHHHHCSBTTBCCCHHHHHHHHHHHHHTCEEEEEEEECSSSCCHHHHTCCCCHHHHHHHHHHHHHHHHHHTTTTCSEEEEEECCBCTTSSBCCCHHHHHHCTTHHHHHHHHHHHHCSSSEEEEEESSCSSSSHHHHHHHHHHHHHHHTTCCCCEEEECCEEETTCCCCHHHHHHHHHHHHTTTCEEEEEEEEEEEESSCCHHHHHHHHHHHHHHHHHHHHSTTEEEEEESCSBGGGCSSTTTSTTEECCSSBCTTSCBCHHHHHHHHHTTC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01805", "category": "Cell", "instruction": "Identify the cell type from the given marker genes.", "input": "IAPP, NKX6-1, MAFA", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as MAFA, NKX6-1, IAPP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "pancreas", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IAPP", "NKX6-1", "MAFA"]}}
{"id": "eval_01806", "category": "Mutation", "instruction": "Describe the effect of mutation P81446-I251F on protein Beta-galactoside-specific lectin 1.", "input": "### Wild-type:\nMNGHLASRRAWVWYFLMLGQVFGATVKAETKFSYERLRLRVTHQTTGEEYFRFITLLRDYVSSGSFSNEIPLLRQSTIPVSDAQRFVLVELTNEGGDSITAAIDVTNLYVVAYQAGDQSYFLRDAPRGAETHLFTGTTRSSLPFNGSYPDLERYAGHRDQIPLGIDQLIQSVTALRFPGGSTRTQARSILILIQMISEAARFNPILWRARQYINSGASFLPDVYMLELETSWGQQSTQVQQSTDGVFNNPIRLAIPPGNFVTLTNVRDVIASLAIMLFVCGERPSSSDVRYWPLVIRPVI\n\n### Mutant:\nMNGHLASRRAWVWYFLMLGQVFGATVKAETKFSYERLRLRVTHQTTGEEYFRFITLLRDYVSSGSFSNEIPLLRQSTIPVSDAQRFVLVELTNEGGDSITAAIDVTNLYVVAYQAGDQSYFLRDAPRGAETHLFTGTTRSSLPFNGSYPDLERYAGHRDQIPLGIDQLIQSVTALRFPGGSTRTQARSILILIQMISEAARFNPILWRARQYINSGASFLPDVYMLELETSWGQQSTQVQQSTDGVFNNPFRLAIPPGNFVTLTNVRDVIASLAIMLFVCGERPSSSDVRYWPLVIRPVI\n\n### Answer:", "output": "In MLA'. The functional changes caused by the mutation in mistletoe lectin I are not mentioned in the document.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01807", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "CC(C)(C)C1CCCCC1=O", "output": "The molecule contains the following functional groups: ketone.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "CC(C)(C)C1CCCCC1=O", "compound_name": null, "pubchem_cid": null, "MW": 154.25, "LogP": 2.79, "TPSA": 17.07, "QED": 0.52, "functional_groups": ["ketone"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["ketone"]}}
{"id": "eval_01808", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGHMAKTYDYLFKLLLIGDSGVGKTCVLFRFSEDAFNSTFISTIGIDFKIRTIELDGKRIKLQIWDTAGQERFRTITTAYYRGAMGIMLVYDITNEKSFDNIRNWIRNIEEHASADVEKMILGNKCDVNDKRQVSKERGEKLALDYGIKFMETSAKANINVENAFFTLARDIKAKMDKKLEGNSPQG\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCEEEEEEECCTTSCHHHHHHHHTTCSCCCCSTTTTSEEEEEEEEEETTEEEEEEEEEECCCGGGHHHHHHHHHTCSEEEEEEETTCHHHHHHHHHHHHHHHHHSCTTCEEEEEEETTTCGGGCCSCHHHHHHHHHHTTCEEEECBTTTTBSHHHHHHHHHHHHHHHHHHTCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01809", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMSQRKLQQDIDKLLKKVKEGIEDFDDIYEKFQSTDPSNSSHREKLESDLKREIKKLQKHRDQIKTWLSKEDVKDKQSVLMTNRRLIENGMERFKSVEKLMKTKQFSKEALTNPDIIKDPKELKKRDQVLFIHDCLDELQKQLEQYEAQENEEQTERHEFHIANLENILKKLQNNEMDPEPVEEFQDDIKYYVENNDDPDFIEYDTIYEDMGCEIQPSSSNNEAPKEGNNQTSLSSIRSSKKQERSPKKKAPQRDVSISDRATTPIAPGVESASQSISSTPTPVSTDTPLHTVKDDSIKFDNSTLGTPTTHVSMKKKESENDSEQQLNFPPDRTDEIRKTIQHDVETNAAFQNPLFNDELKYWLDSKRYLMQPLQEMSPKMVSQLESSLLNCPDSLDADSPCLYTKPLSLPHPTSIFFPNEPIRFVYPYDVPLNLTNNENDTDNKFGKDSKAKSKKDDDIYSRTSLARIFMKFDLDTLFFIFYHYQGSYEQFLAARELFKNRNWLFNKVDRCWYYKEIEKLPPGMGKSEEESWRYFDYKKSWLARRCGNDFVYNEEDFEKL\n\n### Sequence 2:\nMEPLQQQQQQQQQQQKQPHLAPLQMDAREKQGQQMREAQFLYAQKLVTQPTLLSATAGRPSGSTPLGPLARVPPTAAVAQVFERGNMNSEPEEEDGGLEDEDGDDEVAEVAEKETQAASKYFHVQKVARQDPRVAPMSNLLPAPGLPPHGQQAKEDHTKDASKASPSVSTAGQPNWNLDEQLKQNGGLAWSDDADGGRGREISRDFAKLYELDGDPERKEFLDDLFVFMQKRGTPINRIPIMAKQILDLYMLYKLVTEKGGLVEIINKKIWREITKGLNLPTSITSAAFTLRTQYMKYLYAYECEKKALSSPAELQAAIDGNRREGRRPSYSSSLFGYSPAAATAAAAAGAPALLSPPKIRFPILGLGSSSGTNTSSPRISPATTLRKGDGAPVTTVPVPNRLAVPVTLASQQAGTRTAALEQLRERLESGEPAEKKASRLSEEEQRLVQQAFQRNFFSMARQLPMKIRINGRAEDRAEASAAALNLTTSSIGSINMSVDIDGTTYAGVLFAQKPVVHLITGSAPQSLGSSASSSSSSHCSPSPTSSRGTPSAEPSTSWSL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01810", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nGP167_BPPH2\n\n### Answer:", "output": "DNA replication protein 16.7, Gene product 16.7, gp16.7, Protein p16.7.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01811", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMMVDVLIKDISQMVTMKGSNGPRRGKDMREVHLIEDGWIAIKGDKIVAVGSGLMEENLKIGKNTMVIDGKGKTVTPGLVDPHTHLVHGGSRENELALKLNGVPYLDILAQGGGILSTVKATRKATVEELMQQGKRSLDQMLSFGVTTVEVKSGYGLNTETELKQLEVIRQLNKEHPCDLVPTFMGAHAIPMEYKEDPDVFVDIVIDEMLPAVVERGLAEFCDVFCEKGVFTVAQSRRVLEAAREAGLKLRIHVDEIEALGGAELAAEMGAITAEHLMVTSEEDMKKMAKAGVIAALLPGTSFNLMVGKYAQARKMIDYGVPITLSTDYNPGSCPTENIQFIMTLGCLAMKMTPEEVFTAVTINGAAAVDRQGDIGSLEVGKKADVVIFNAPNINYIPYHFGINHVDKVLKNGKLVVDKGRVI\n\n### Sequence 2:\nMVADVLLTHFNQLFCLNDPGHPLTGQEMKKATIVEDGYIAIKDGLIVALGSGEPDAEL-VGPQTIMRSYKGKIATPGIIDCHTHLVYGGSREHEFAKKLAGVSYLDILAQGGGILSTVRATRSASFDNLYQKSKRLLDYMLLHGVTTVEAKSGYGLDWETEKRQLDVVAALEKDHPIDLVSTFMAAHAIPEEYKGNPKAYLDVIIKDMLPVVKEENLAEFCDIFCEKNVFTADESRYLLSKAKEMGFKLRIHADEIASIGGVDVAAELSAVSAEHLMMITNDGIAKLIGAGVIGNLLPATTFSLMEDTYAPARKMIDAGMAITLSTDSNPGSCPTANMQFVMQLGCFMLRLTPIEVLNAVTINAAYSVNRQERVGSLTVGKEADIAIFDAPNIDYLFYFFATNLIHQVYKKGQLTVDRGRIL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01812", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNSDLKSFDKIEQKIGGSRKISNYIIGGMLTIGGIGFLLASISSYTGRDLLPLGNPSTLLFIPQGIIMGAYGVIANLLNFYLWYLVYINFGSGSNYFDKSSKSIEIRRKGLFKDIEVKLNFDEIKSVKLDISEGFNPRRRIALVLKGRKKPLPLSGAGELKPLLQVEEEGARLAKFLDVNLEGLK\n\n### Sequence 2:\nMKKNVQIKGTKDGISIFLSDKASISELQQELTQLLADQKQNPYSGEKLEVQVQIGNRLFSEEEELEISTIIHENSQMEISAFYSNVMSKDEAKKWKENDQIFSMATIIRSGQVVQVPGDFLLIGDVNPGGQIRSNGNVFVLGNIKGIIHAGFEGNENAVVAGKFLYPSQVRIADKVYGFDSEDYKEVTETDLFSAFVNDAGEIVIDGIHKIRKIRPEISNFQGGR\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01813", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nCPNS1_MOUSE\n\n### Answer:", "output": "Cytoplasm. Cell membrane. Note=Translocates to the plasma membrane upon calcium binding.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01814", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nS100B:low GFAP:medium SLC1A3:medium\n</cell>", "output": "This cell is most consistent with astrocyte, an astrocyte. Key markers such as SLC1A3, S100B, GFAP support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "astrocyte", "tissue": "spleen", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["S100B", "GFAP", "SLC1A3"]}}
{"id": "eval_01815", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "CD3D, CCR7, CD3E, CD8B, CD8A, SELL", "output": "This cell is most consistent with naive CD8+ T cell, a naive cytotoxic T-cell subtype. Key markers such as CD3D, CD3E, CCR7 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "naive CD8+ T cell", "tissue": "bone marrow", "disease": "type 2 diabetes", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CD3D", "CCR7", "CD3E", "CD8B", "CD8A", "SELL"]}}
{"id": "eval_01816", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAPVPVTKLVCDGDTYKCTAYLDYGDGKWVAQWDTAVFHTT\n\n### Answer:", "output": "<SEQ_2D>CCEEEEEEEEBTTTTEEEEEECSTTSCEEEEEECEEEECC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01817", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMTKSDLIERLTSKHFQLSVKEVEDSVKETLTLMANSLAQGERIEVRGFGSFSLHYRAPRVGRNPKTGDRVELDGKYVPHFKPGKALRERVNAANA\n\n### Sequence 2:\nMAMRFLNIGYGNIVSAHRIIAIVSPESAPIKRTVQEAREHNALLDATYGRKTRAVIVMDDGHVVLSPIQPETIAHRLNNKEDLSEEG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01818", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKIRAIVTIKGLVQGVAFRHHTVQQAQRLGVSGWVKNLAGGDVQGCFEGEEEAVDALVAWCHHGPSRARVDRVILEREHYRGEFDDFDVRY\n\n### Sequence 2:\nMMKRVHVYVKGKVQGVFFRAHTRKAALRFHVNGWVRNLPNGRVEAVFEGRSPEVNALIDWCRQGPSHAIVEHLDVCEETYTGEFDDFRILY\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01819", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BTU6-R275E on protein Phosphatidylinositol 4-kinase type 2-alpha.", "input": "### Wild-type:\nMDETSPLVSPERAQPPDYTFPSGSGAHFPQVPGGAVRVAAAAGSGPSPPGSPGHDRERQPLLDRARGAAAQGQTQTVAAQAQALAAQAAAAAHAAQAHRERNEFPEDPEFEAVVRQAELAIERCIFPERIYQGSSGSYFVKDPQGRIIAVFKPKNEEPYGHLNPKWTKWLQKLCCPCCFGRDCLVLNQGYLSEAGASLVDQKLELNIVPRTKVVYLASETFNYSAIDRVKSRGKRLALEKVPKVGQRFNRIGLPPKVGSFQLFVEGYKDADYWLRRFEAEPLPENTNRQLLLQFERLVVL\n\n### Mutant:\nMDETSPLVSPERAQPPDYTFPSGSGAHFPQVPGGAVRVAAAAGSGPSPPGSPGHDRERQPLLDRARGAAAQGQTQTVAAQAQALAAQAAAAAHAAQAHRERNEFPEDPEFEAVVRQAELAIERCIFPERIYQGSSGSYFVKDPQGRIIAVFKPKNEEPYGHLNPKWTKWLQKLCCPCCFGRDCLVLNQGYLSEAGASLVDQKLELNIVPRTKVVYLASETFNYSAIDRVKSRGKRLALEKVPKVGQRFNRIGLPPKVGSFQLFVEGYKDADYWLERFEAEPLPENTNRQLLLQFERLVVL\n\n### Answer:", "output": "Reduces enzyme activity, probably due to impaired membrane-association; when associated with E-129 and E-276. The mutation in PI4KIIα affects its activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01820", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nGHRB_ECOLU\n\n### Answer:", "output": "Belongs to the D-isomer specific 2-hydroxyacid dehydrogenase family. GhrB subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01821", "category": "Mutation", "instruction": "Describe the effect of mutation Q9JHE5-C279S on protein Sodium-coupled neutral amino acid symporter 2.", "input": "### Wild-type:\nMKKTEMGRFNISPDEDSSSYSSNGDFNYSYPTKQAALKSHYVDVDPENQNFLLESNLGKKKYETDFHPGTTSFGMSVFNLSNAIVGSGILGLSYAMANTGIALFIILLTFVSIFSLYSVHLLLKTANEGGSLLYEQLGHKAYGLAGKLAASGSITMQNIGAMSSYLFIVKYELPLVIKALMNIEDTNGLWYLNGDYLVLLVSFVLILPLSLLRNLGYLGYTSGLSLLCMIFFLIVVICKKFQIPCPVEVALMANETVNGTFTQVALAALASNSTAADTCRPRYFIFNSQTVYAVPILTFS\n\n### Mutant:\nMKKTEMGRFNISPDEDSSSYSSNGDFNYSYPTKQAALKSHYVDVDPENQNFLLESNLGKKKYETDFHPGTTSFGMSVFNLSNAIVGSGILGLSYAMANTGIALFIILLTFVSIFSLYSVHLLLKTANEGGSLLYEQLGHKAYGLAGKLAASGSITMQNIGAMSSYLFIVKYELPLVIKALMNIEDTNGLWYLNGDYLVLLVSFVLILPLSLLRNLGYLGYTSGLSLLCMIFFLIVVICKKFQIPCPVEVALMANETVNGTFTQVALAALASNSTAADTSRPRYFIFNSQTVYAVPILTFS\n\n### Answer:", "output": "Does not affect amino acid transport activity. Significantly decreases alanine transport activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01822", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nWAPL_DROME\n\n### Answer:", "output": "Protein wings apart-like, Protein parallel sister chromatids.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01823", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nTMEM119:high CSF1R:high CX3CR1:medium\n</cell>", "output": "This cell is most consistent with microglia, a brain microglia cell. Key markers such as CSF1R, TMEM119, CX3CR1 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "microglia", "tissue": "breast", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["TMEM119", "CSF1R", "CX3CR1"]}}
{"id": "eval_01824", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMTAPTVPVALVTGAAKRLGRSIAEGLHAEGYAVCLHYHRSAAEANALSATLNARRPNSAITVQADLSNVATAPVSGADGSAPVTLFTRCAELVAACYTHWGRCDVLVNNASSFYPTPLLRNDEDGHEPCVGDREAMETATADLFGSNAIAPYFLIKAFAHRVAGTPAKHRGTNYSIINMVDAMTNQPLLGYTIYTMAKGALEGLTRSAALELAPLQIRVNGVGPGLSVLVDDMPPAVWEGHRSKVPLYQRDSSAAEVSDVVIFLCSSKAKYITGTCVKVDGGYSLTRA\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCEEEESSCSSHHHHHHHHHHHHTTCEEEEEESSCHHHHHHHHHHHHHHSTTCEEEEECCCSSCEECCCCCCCCCCCEEHHHHHHHHHHHHHHHHSCCCEEEECCCCCCCBCSCCCCCCCCCCCCCTTHHHHHHHHHHHHHHTHHHHHHHHHHHHHHHHSCGGGSCSCEEEEEECCGGGGSCCTTBHHHHHHHHHHHHHHHHHHHHTTTTTEEEEEEEESCBSCCCCCCCCCCCCSGGGCTTTTSCBCHHHHHHHHHHHHSGGGTTCCSCEEEESTTGGGCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01825", "category": "Mutation", "instruction": "Describe the effect of mutation Q99PE8-M93K on protein ATP-binding cassette sub-family G member 5.", "input": "### Wild-type:\nMGELPFLSPEGARGPHINRGSLSSLEQGSVTGTEARHSLGVLHVSYSVSNRVGPWWNIKSCQQKWDRQILKDVSLYIESGQIMCILGSSGSGMTTLLDAISGRLRRTGTLEGEVFVNGCELRRDQFQDCFSYVLQSDVFLSSLTVRETLRYTAMLALCRSSADFYNKKVEAVMTELSLSHVADQMIGSYNFGGISSGERRRVSIAAQLLQDPKVMMLDEPTTGLDCMTANQIVLLLAELARRDRIVIVTIHQPRSELFQHFDKIAILTYGELVFCGTPEEMLGFFNNCGYPCPEHSNPFD\n\n### Mutant:\nMGELPFLSPEGARGPHINRGSLSSLEQGSVTGTEARHSLGVLHVSYSVSNRVGPWWNIKSCQQKWDRQILKDVSLYIESGQIMCILGSSGSGKTTLLDAISGRLRRTGTLEGEVFVNGCELRRDQFQDCFSYVLQSDVFLSSLTVRETLRYTAMLALCRSSADFYNKKVEAVMTELSLSHVADQMIGSYNFGGISSGERRRVSIAAQLLQDPKVMMLDEPTTGLDCMTANQIVLLLAELARRDRIVIVTIHQPRSELFQHFDKIAILTYGELVFCGTPEEMLGFFNNCGYPCPEHSNPFD\n\n### Answer:", "output": "Disrupts sterol transport activity. Increases expression of both ABCG5 and ABCG8.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01826", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKIVIADDHAVVRTGFSMILNYQNDMEVVATAADGVEAYQKVMEYKPDVLLMDLSMPPGESGLIATSKIADSFPETKILILTMFDDEEYLFHVLRNGAKGYILKNAPDEQLLLAIRTVYKGETYVDMKLTTSLVNEFVSNSNQDTANTTDPFKILSKRELEILPLIAKGYGNKEIAEKLFVSVKTVEAHKTHIMTKLGLKSKPELVEYALKKKLLEF\n\n### Sequence 2:\nMENFFSILGGMGTMATESFVRLINHRTKATKDQEYLNYVLFNHATVPDRTAYILDRSEENPMPFLLDDIEKQNLLRPNFIVLTCNTAHYFFEELQAATDIPILHMPREAANELVRQHTTGRVAILGTEGSMKAGIYEREVKNLGFETVIPDTALQEKINYLIYHEIKESDYLNQELYYEILEEAVERLNCEKVILGCTELSLMHEFAEDNHYPVIDAQSILADRTIERALAERSEALDTASEK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01827", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMAYFMANATRENCQDECRLSPNRDAVEEDLKKIGVECTKVDFDDANVQEDLDQLIKKYNMNYHDEIHICRETMPNFDEKLKIFFEEHLHDDAELRVIKHGAGYFDVRAKDEQWIRIPVRRGDFVFLPPGIYHRFTTDRSEDVVALRLFRNNPKWTAFNRSENGDKQPIREQYLKDVSA\n\n### Sequence 2:\nMALLPDKEKLLRNFLRCANWEEKYLYIIELGQRLPELRDEDKSPQNSIQGCQSQVWIVMRQNAQGIIELQGDSDAAIVKGLIAVVFILYDQMTPQDIVNFDVRPWFEKMALTQHLTPSRSQGLEAMIRAIRAKAAALS\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01828", "category": "Mutation", "instruction": "Describe the effect of mutation P50747-G241W on protein Biotin--protein ligase.", "input": "### Wild-type:\nMEDRLHMDNGLVPQKIVSVHLQDSTLKEVKDQVSNKQAQILEPKPEPSLEIKPEQDGMEHVGRDDPKALGEEPKQRRGSASGSEPAGDSDRGGGPVEHYHLHLSSCHECLELENSTIESVKFASAENIPDLPYDYSSSLESVADETSPEREGRRVNLTGKAPNILLYVGSDSQEALGRFHEVRSVLADCVDIDSYILYHLLEDSALRDPWTDNCLLLVIATRESIPEDLYQKFMAYLSQGGKVLGLSSSFTFGGFQVTSKGALHKTVQNLVFSKADQSEVKLSVLSSGCRYQEGPVRLSP\n\n### Mutant:\nMEDRLHMDNGLVPQKIVSVHLQDSTLKEVKDQVSNKQAQILEPKPEPSLEIKPEQDGMEHVGRDDPKALGEEPKQRRGSASGSEPAGDSDRGGGPVEHYHLHLSSCHECLELENSTIESVKFASAENIPDLPYDYSSSLESVADETSPEREGRRVNLTGKAPNILLYVGSDSQEALGRFHEVRSVLADCVDIDSYILYHLLEDSALRDPWTDNCLLLVIATRESIPEDLYQKFMAYLSQGWKVLGLSSSFTFGGFQVTSKGALHKTVQNLVFSKADQSEVKLSVLSSGCRYQEGPVRLSP\n\n### Answer:", "output": "In HLCS deficiency. The sequencing of HCLS showed homozygosity for the novel missense variant (p. ). The second sibling had a similar presentation at birth: severe metabolic acidosis and respiratory distress. A urine organic acid profile was consistent with HS deficiency; he was treated with biotin powder (20 mg), and after 24 h, the lactate decreased significantly; by day 5 of life, the patient was tolerating 40 mg of biotin, feeding by mouth and off all other medications and support. This is the first report of the p. mutation. To our knowledge, this is also the first mutation described in West African patients with HS deficiency and the cases demonstrate that it is biotin responsive. Additionally, our experience suggests that the powdered form of biotin supplementation may be more digestible than tablets for the treatment of severe neonatal HS deficiency.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01829", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nBLUB_RHIME\n\n### Answer:", "output": "5,6-dimethylbenzimidazole synthase, DMB synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01830", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nFTSH7_ORYSJ\n\n### Answer:", "output": "Probable ATP-dependent zinc metallopeptidase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01831", "category": "Mutation", "instruction": "Describe the effect of mutation P32396-Y13F on protein Coproporphyrin III ferrochelatase.", "input": "### Wild-type:\nMSRKKMGLLVMAYGTPYKEEDIERYYTHIRRGRKPEPEMLQDLKDRYEAIGGISPLAQITEQQAHNLEQHLNEIQDEITFKAYIGLKHIEPFIEDAVAEMHKDGITEAVSIVLAPHFSTFSVQSYNKRAKEEAEKLGGLTITSVESWYDEPKFVTYWVDRVKETYASMPEDERENAMLIVSAHSLPEKIKEFGDPYPDQLHESAKLIAEGAGVSEYAVGWQSEGNTPDPWLGPDVQDLTRDLFEQKGYQAFVYVPVGFVADHLEVLYDNDYECKVVTDDIGASYYRPEMPNAKPEFIDAL\n\n### Mutant:\nMSRKKMGLLVMAFGTPYKEEDIERYYTHIRRGRKPEPEMLQDLKDRYEAIGGISPLAQITEQQAHNLEQHLNEIQDEITFKAYIGLKHIEPFIEDAVAEMHKDGITEAVSIVLAPHFSTFSVQSYNKRAKEEAEKLGGLTITSVESWYDEPKFVTYWVDRVKETYASMPEDERENAMLIVSAHSLPEKIKEFGDPYPDQLHESAKLIAEGAGVSEYAVGWQSEGNTPDPWLGPDVQDLTRDLFEQKGYQAFVYVPVGFVADHLEVLYDNDYECKVVTDDIGASYYRPEMPNAKPEFIDAL\n\n### Answer:", "output": "No change in activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01832", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL9_BACAA\n\n### Answer:", "output": "50S ribosomal protein L9.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01833", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nADD_ECO27\n\n### Answer:", "output": "Adenosine deaminase, Adenosine aminohydrolase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01834", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CCCCN(C)C(=O)Nc1ccc(Cl)c(Cl)c1\nDescriptors: MW=275.18, LogP=4.26, TPSA=32.34, HBD=1, HBA=1", "output": "Descriptors: MW=275.18, LogP=4.26, TPSA=32.34, HBD=1, HBA=1, QED=0.87, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a relatively high QED score suggests good drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "ESOL", "canonical_smiles": "CCCCN(C)C(=O)Nc1ccc(Cl)c(Cl)c1", "compound_name": null, "pubchem_cid": null, "MW": 275.18, "LogP": 4.26, "TPSA": 32.34, "QED": 0.87, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"label": -4.77}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01835", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSEQRVVVMGLGNLLWADEGFGVRVAERLYAHYHWPEYVEIVDGGTQGLNLLGYVESASHLLILDAIDYGLEPGTLRTYAGERIPAYLSAKKMSLHQNSFSEVLALADIRGHLPAHIALVGLQPAMLDDYGGSLSELAREQLPAAEQAALAQLAAWGIVPQPANESRCLNYDCLSMENYEGVRLRQYRMTQEEQG\n\n### Sequence 2:\nMTEAAPTRIYVDADACPVKDEIYRVAERHGLPVSVVAGSFIRVPTHPLIERIAAGSGMDAADDWIAERVQPGDIVVTADVPLASRCVKAGAEVLAPNGKPFSEESIGMTLAVRNLMTDLRSSGEITGGPRGFTPRDRSAFLAALDTTIRRIARRRPSPPAQPQT\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01836", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSHMARYDSLLQALGNTPLVGLQRLSPRWDDGRDGPHVRLWAKLEDRNPTGSIKDRPAVRMIEQAEADGLLRPGATILEPTSGNTGISLAMAARLKGYRLICVMPENTSVERRQLLELYGAQIIFSAAEGGSNTAVATAKELAATNPSWVMLYQYGNPANTDSHYCGTGPELLADLPEITHFVAGLGTTGTLMGTGRFLREHVANVKIVAAEPRYGEGVYALRNMDEGFVPELYDPEILTARYSVGAVDAVRRTRELVHTEGIFAGISTGAVLHAALGVGAGALAAGERADIALVVADAGWKYLSTGAYAGSLDDAETALEGQLWA\n\n### Answer:", "output": "<SEQ_2D>CCCCSSBSSGGGSCSCCCEEECTTTSSBCCCCCCCSCEEEEEEETTSSTTSBTTHHHHHHHHHHHHHTTCCCTTCEEEEECSSHHHHHHHHHHHHHTCEEEEEEESSSCHHHHHHHHHTTCEEEEECGGGHHHHHHHHHHHHHHHCTTSEECCTTTCHHHHHHHHHTHHHHHHHHCTTCCEEEEECSSSHHHHHHHHHHHHHSTTCEEEEEEECCCCCCCCCCCCCCCCCCTTCCTTTCSEEEEECHHHHHHHHHHHHHHHCCCBCHHHHHHHHHHHHHHHHHHHTTCCEEEEEEECBBGGGGGGGTCCCCCCCCCCCCCCCCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01837", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nLSPA_ANAPZ\n\n### Answer:", "output": "Lipoprotein signal peptidase, Prolipoprotein signal peptidase, Signal peptidase II, SPase II.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01838", "category": "Mutation", "instruction": "Describe the effect of mutation Q0WX57-C89S on protein Ubiquitin carboxyl-terminal hydrolase 17-like protein 24.", "input": "### Wild-type:\nMEDDSLYLRGEWQFNHFSKLTSSRPDAAFAEIQRTSLPEKSPLSCETRVDLCDDLAPVARQLAPREKLPLSSRRPAAVGAGLQNMGNTCYVNASLQCLTYTPPLANYMLSREHSQTCHRHKGCMLCTMQAHITRALHNPGHVIQPSQALAAGFHRGKQEDAHEFLMFTVDAMKKACLPGHKQVDHHSKDTTLIHQIFGGYWRSQIKCLHCHGISDTFDPYLDIALDIQAAQSVQQALEQLVKPEELNGENAYHCGVCLQRAPASKTLTLHTSAKVLILVLKRFSDVTGNKIAKNVQYPEC\n\n### Mutant:\nMEDDSLYLRGEWQFNHFSKLTSSRPDAAFAEIQRTSLPEKSPLSCETRVDLCDDLAPVARQLAPREKLPLSSRRPAAVGAGLQNMGNTSYVNASLQCLTYTPPLANYMLSREHSQTCHRHKGCMLCTMQAHITRALHNPGHVIQPSQALAAGFHRGKQEDAHEFLMFTVDAMKKACLPGHKQVDHHSKDTTLIHQIFGGYWRSQIKCLHCHGISDTFDPYLDIALDIQAAQSVQQALEQLVKPEELNGENAYHCGVCLQRAPASKTLTLHTSAKVLILVLKRFSDVTGNKIAKNVQYPEC\n\n### Answer:", "output": "Abolishes enzymatic activity. Loss of the pro-apoptotic function.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01839", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nATC9_SCHPO\n\n### Answer:", "output": "Belongs to the cation transport ATPase (P-type) (TC 3.A.3) family. Type V subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01840", "category": "Mutation", "instruction": "Describe the effect of mutation O75569-S287A on protein Interferon-inducible double-stranded RNA-dependent protein kinase activator A.", "input": "### Wild-type:\nMSQSRHRAEAPPLEREDSGTFSLGKMITAKPGKTPIQVLHEYGMKTKNIPVYECERSDVQIHVPTFTFRVTVGDITCTGEGTSKKLAKHRAAEAAINILKANASICFAVPDPLMPDPSKQPKNQLNPIGSLQELAIHHGWRLPEYTLSQEGGPAHKREYTTICRLESFMETGKGASKKQAKRNAAEKFLAKFSNISPENHISLTNVVGHSLGCTWHSLRNSPGEKINLLKRSLLSIPNTDYIQLLSEIAKEQGFNITYLDIDELSANGQYQCLAELSTSPITVCHGSGISCGNAQSDAAH\n\n### Mutant:\nMSQSRHRAEAPPLEREDSGTFSLGKMITAKPGKTPIQVLHEYGMKTKNIPVYECERSDVQIHVPTFTFRVTVGDITCTGEGTSKKLAKHRAAEAAINILKANASICFAVPDPLMPDPSKQPKNQLNPIGSLQELAIHHGWRLPEYTLSQEGGPAHKREYTTICRLESFMETGKGASKKQAKRNAAEKFLAKFSNISPENHISLTNVVGHSLGCTWHSLRNSPGEKINLLKRSLLSIPNTDYIQLLSEIAKEQGFNITYLDIDELSANGQYQCLAELSTSPITVCHGAGISCGNAQSDAAH\n\n### Answer:", "output": "Abrogates apoptosis induction under conditions of stress. Prevents activation of EIF2AK2 in stressed cells; when associated with A-246. Substitution of S287 with alanine was tolerated.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01841", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMEFPDLGKHCSEPTCKQLDFLPITCDACKQDFCKDHFSYVGHKCPFAFKKDVQVPVCPLCNAPIPVKRGEIPDVVVGEHMDRDCTFHPGRNRNKVFTHRCSKEGCRKKEMLQLACAQCHGNFCIQHRHPLDHNCQAGSSSASRGRTSTSRAAEQKPSGVSWLAQRLRRTVK\n\n### Sequence 2:\nMVMAMGFDTVVAAIMATAIIVAVAYTFLAGSTSIAELSVESYKDAVNSAVKKLRSDIEILSVSYDNSTSKIVAYFKNTGDERYPDFSEFDAIVYGRTSGGEMVSFYVNSTVFSITNELINPGIFDPQEVAKLEAVQPLQNGTYVLLICTPNAVCDSADFSV\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01842", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAQYEDGKQYTTLEKPVAGAPQVLEFFSFFCPHCYQFEEVLHISDNVKKKLPEGVKMTKYHVNFMGGDLGKDLTQAWAVAMALGVEDKVTVPLFEGVQKTQTIRSASDIRDVFINAGIKGEEYDAAWNSFVVKSLVAQQEKAAADVQLRGVPAMFVNGKYQLNPQGMDTSNMDVFVQQYADTVKYLSEKK\n\n### Answer:", "output": "<SEQ_2D>CCSCBTTTEEECSSCCTTCCSEEEEECTTCHHHHHHHHTSCHHHHHHHHSCTTCCEEEEECSSSSHHHHHHHHHHHHHHHHHTCHHHHHHHHHHHHHTSCCCCSHHHHHHHHHHTTCCHHHHHHHHHSHHHHHHHHHHHHHHHHTTCCSSSEEEETTTEEECGGGSCSSCHHHHHHHHHHHHHHHHTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01843", "category": "Mutation", "instruction": "Describe the effect of mutation Q9ULV1-S233A on protein Frizzled-4.", "input": "### Wild-type:\nMAWRGAGPSVPGAPGGVGLSLGLLLQLLLLLGPARGFGDEEERRCDPIRISMCQNLGYNVTKMPNLVGHELQTDAELQLTTFTPLIQYGCSSQLQFFLCSVYVPMCTEKINIPIGPCGGMCLSVKRRCEPVLKEFGFAWPESLNCSKFPPQNDHNHMCMEGPGDEEVPLPHKTPIQPGEECHSVGTNSDQYIWVKRSLNCVLKCGYDAGLYSRSAKEFTDIWMAVWASLCFISTAFTVLTFLIDSSRFSYPERPIIFLSMCYNIYSIAYIVRLTVGRERISCDFEEAAEPVLIQEGLKNT\n\n### Mutant:\nMAWRGAGPSVPGAPGGVGLSLGLLLQLLLLLGPARGFGDEEERRCDPIRISMCQNLGYNVTKMPNLVGHELQTDAELQLTTFTPLIQYGCSSQLQFFLCSVYVPMCTEKINIPIGPCGGMCLSVKRRCEPVLKEFGFAWPESLNCSKFPPQNDHNHMCMEGPGDEEVPLPHKTPIQPGEECHSVGTNSDQYIWVKRSLNCVLKCGYDAGLYSRSAKEFTDIWMAVWASLCFIATAFTVLTFLIDSSRFSYPERPIIFLSMCYNIYSIAYIVRLTVGRERISCDFEEAAEPVLIQEGLKNT\n\n### Answer:", "output": "Increased signaling activity in presence of NDP/norrin but not in presence of WNT3A.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01844", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMQLRITSRKKLTSLLCALGLISIVAIYPRQTVNFFYSTAVQITDYIHFYGYRPVKSFAIRIPASYTIHGIDVSRWQERIDWQRVAKMRDNGIRLQFAFIKATEGEKLVDPYFSRNWQLSRENGLLRGAYHYFSPSVSASVQARLFLQTVDFSQGDFPAVLDVEERGKLSAKELRKRVSQWLKMVEKSTGKKPIIYSGAVFYHTNLAGYFNEYPWWVAHYYQRRPDNDGMAWRFWQHSDRGQVDGINGPVDFNVFNGTVEELQAFVDGIKETP\n\n### Sequence 2:\nMQLRITSRKKFTVLLCALGLISIVAIYPRQTVNFFYSTAIQIKDYIHFYGYRPVKSFAIRIPASYTIHGIDVSRWQERIDWQRVAKMRDNGIRLQFAFIKATEGEKLVDPYFSRNWQLSRENGLLRGAYHYFSPSVSASVQARLFLQTVDFSQGDFPAVLDVEERGKLSAKELRKRVSQWLKMVEKRTGKKPIIYSGAVFYHTNLAGYFNEYPWWVAHYYQRRPDNDGMAWRFWQHSDRGQVDGINGPVDFNVFNGTVEELQGFVDGIKETP\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01845", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nS100A4 CD4 CD45RO CD3D\n</cell>", "output": "This cell is most consistent with memory CD4+ T cell, a memory CD4+ T-cell population. Key markers such as CD3D, CD45RO, CD4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "memory CD4+ T cell", "tissue": "lymph node", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["S100A4", "CD4", "CD45RO", "CD3D"]}}
{"id": "eval_01846", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMKKFDILDFFYIKSQDTQILRYYISKNPTNYGGLSLTNGAEGKTSTMAKSLNTTPTLDTMPMFTEDVCFAAYSIDMQKILHNMKLKTNLEYPGLLFDHEGPDCLPLIKEAYGFVDKSFWVLPGRGKELFKDVKKVR\n\n### Sequence 2:\nMSAQEERIPEDDDTTEAPLTMAASVVLTNLPRDASKALETAGSLNVQKITVRLHPIGSAPALTQRVFKLSTNQRFDTIVRFLRKRLGVKEHESVFCYVGSVFAPGLDEGVGGLWSGEELVVGYAMAPAFG\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01847", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMRHRHGLRKLNRTSSHRLAMLRNMSNSLIEHEVIKTTLPKAKELRKVVEPLITLGKKPSLANRRLAFNRLRDRDSVAKLFDVLGPRFANRPGGYLRVLKFGFRVGDNAPMALVELLDRPEVDETENVQEAE\n\n### Sequence 2:\nMRHRHGLRKLNRTSSHRLAMLRNMSNSLIEHEVIKTTLPKAKELRKVVEPLITLGKKPSLANRRLAFNRLRDRDSVAKLFDVLGPRFANRPGGYLRVLKFGFRVGDNAPMALVELLDRPEVDETENVQEAE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01848", "category": "Mutation", "instruction": "Describe the effect of mutation Q9V7W1-Q306A on protein GPI mannosyltransferase 2.", "input": "### Wild-type:\nMTEKVTKLALASRLIVLLVQLVANGALPEHKPDVFRMPVSSDQNASWIDKVIKRCLGGLRHWDGEYFLHIAENLYSYENTLAFYPLYPVVVRHVGQAVEAIGISLSQESILLVVAVALNFWLFCESANLLFQLTQVLFNDLNKSWNAALIYCFNPATIFFTAAYSETFFAYSSLHLMLECSKPTGSFRYLRLGTALAACLLCRSNGLITLGYPLYFFGRQLLLKNKEPNTCMQLTQMTLTILGAIGILHTYYFYIYRLYCLPNTRPNHPQHIVDYAVERKYLLSGQGSEGSPWCQYTLPF\n\n### Mutant:\nMTEKVTKLALASRLIVLLVQLVANGALPEHKPDVFRMPVSSDQNASWIDKVIKRCLGGLRHWDGEYFLHIAENLYSYENTLAFYPLYPVVVRHVGQAVEAIGISLSQESILLVVAVALNFWLFCESANLLFQLTQVLFNDLNKSWNAALIYCFNPATIFFTAAYSETFFAYSSLHLMLECSKPTGSFRYLRLGTALAACLLCRSNGLITLGYPLYFFGRQLLLKNKEPNTCMQLTQMTLTILGAIGILHTYYFYIYRLYCLPNTRPNHPQHIVDYAVERKYLLSGQGSEGSPWCQYTLPF\n\n### Answer:", "output": "Mild photoreceptor cell and microvillar disorganization; when associated with V-276. The mutation in GPI-MT2 leads to defects in chaoptin trafficking to the plasma membrane in Drosophila photoreceptor cells. Loss of sufficient chaoptin in the membrane leads to microvillar instability, photoreceptor cell pathology, and retinal degeneration.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01849", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSYL_SHIF8\n\n### Answer:", "output": "Belongs to the class-I aminoacyl-tRNA synthetase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01850", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEKIIYFAAYVITSVDEEMRHNELSTLEAEMAVERKAVEDQRDGELEARAQKLEADLAELEAEGAKADARRKVRDGGEREMRQIRDRAQRELDRLEDIWSTFTKLAPKQLIVDENLYRELVDRYGEYFTGAM\n\n### Answer:", "output": "<SEQ_3Di>DPDDFDDKKFFAAFQVVVCVVCVVVVVVVLVVVLVVLVVVLVVVLVVLVVVLVVVLVVCVVVVHDPVVNVVSVVVSVVVNVVSVVVSVVVVVVSVVLVVCVVPPDGRDMDGYVVSVVVCCVSPVPRTDIGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01851", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "N#CC(C#N)=NNc1cccc(Cl)c1", "output": "The molecule contains the following functional groups: aromatic ring, nitrile, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "N#CC(C#N)=NNc1cccc(Cl)c1", "compound_name": null, "pubchem_cid": null, "MW": 204.62, "LogP": 2.16, "TPSA": 71.97, "QED": 0.59, "functional_groups": ["aromatic ring", "nitrile", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": null, "NR-AhR": 0.0, "NR-Aromatase": 1.0, "NR-ER": null, "NR-ER-LBD": null, "NR-PPAR-gamma": null, "SR-ARE": null, "SR-ATAD5": null, "SR-HSE": 1.0, "SR-MMP": 1.0, "SR-p53": null}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "nitrile", "halide"]}}
{"id": "eval_01852", "category": "Cell", "instruction": "Infer the cell identity from this expression profile.", "input": "<cell>\nHAVCR2:high CD8A:medium TOX:high LAG3:medium\n</cell>", "output": "This cell is most consistent with exhausted CD8 T cell, an exhausted CD8+ T cell. Key markers such as LAG3, HAVCR2, TOX support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "exhausted CD8 T cell", "tissue": "spleen", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["HAVCR2", "CD8A", "TOX", "LAG3"]}}
{"id": "eval_01853", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "COCCCNc1nc(NC(C)C)nc(SC)n1", "output": "This molecule is medium-sized and moderately polar, with moderately lipophilic character. It contains ether. Its descriptor profile: MW=271.39, LogP=1.86, TPSA=71.96, HBD=2, HBA=7, QED=0.55.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "ESOL", "canonical_smiles": "COCCCNc1nc(NC(C)C)nc(SC)n1", "compound_name": null, "pubchem_cid": null, "MW": 271.39, "LogP": 1.86, "TPSA": 71.96, "QED": 0.55, "functional_groups": ["ether"], "moleculenet_labels": {"label": -2.928}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01854", "category": "Mutation", "instruction": "Describe the effect of mutation A0A068J840-H144Y on protein UDP-glycosyltransferase 1.", "input": "### Wild-type:\nMKSELIFLPAPAIGHLVGMVEMAKLFISRHENLSVTVLIAKFYMDTGVDNYNKSLLTNPTPRLTIVNLPETDPQNYMLKPRHAIFPSVIETQKTHVRDIISGMTQSESTQVVGLLADLLFINIMDIANEFNVPTYVYSPAGAGHLGLAFHLQTLNDKKQDVTEFRNSDTELLVPSFANPVPAEVLPSMYVDKEGGYDYLFSLFRRCRESKAIIINTFEELEPYAINSLRMDSMIPPIYPVGPILNLNGDGQNSDEAAVILGWLDDQPPSSVVFLCFGSYGSFQENQVKEIAMGLERSGHR\n\n### Mutant:\nMKSELIFLPAPAIGHLVGMVEMAKLFISRHENLSVTVLIAKFYMDTGVDNYNKSLLTNPTPRLTIVNLPETDPQNYMLKPRHAIFPSVIETQKTHVRDIISGMTQSESTQVVGLLADLLFINIMDIANEFNVPTYVYSPAGAGYLGLAFHLQTLNDKKQDVTEFRNSDTELLVPSFANPVPAEVLPSMYVDKEGGYDYLFSLFRRCRESKAIIINTFEELEPYAINSLRMDSMIPPIYPVGPILNLNGDGQNSDEAAVILGWLDDQPPSSVVFLCFGSYGSFQENQVKEIAMGLERSGHR\n\n### Answer:", "output": "Loss of glycosylase activity toward protopanaxatriol (PPT). The mutation in UGTPg1 may play an important role in determining its activity and substrate regio-specificity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01855", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "Cn1c(N)nc2c3cccnc3ccc21", "output": "This molecule is small and moderately polar, with moderately lipophilic character. It contains aromatic ring, pyridine. Its descriptor profile: MW=198.23, LogP=1.7, TPSA=56.73, HBD=1, HBA=3, QED=0.6.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "Cn1c(N)nc2c3cccnc3ccc21", "compound_name": null, "pubchem_cid": null, "MW": 198.23, "LogP": 1.7, "TPSA": 56.73, "QED": 0.6, "functional_groups": ["aromatic ring", "pyridine"], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": 0.0, "NR-AhR": 1.0, "NR-Aromatase": null, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01856", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAGNEKSRAKMEQAKGKAKEAAGRAVGNERMTAEGRAAQSKGDARQAKEKGKDVFRH\n\n### Sequence 2:\nMVAYPEISWTRNGCTVSKYPEISWTRNGCTVSKYPEISWTRNGCTVSKYPEISWTRNGCTVA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01857", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMKTVTDYQNKNILVLGIAKSGYAAANLLKSLGANVIVNDGKPLANNELAAELQAKGMDVVCGGHPLELLERNIALVVKNPGIPYSNPLLVAATEKQIPIITEIELAYRISEAPFIGITGSNGKTTTTMLTFEMLKEGEKHPAIAGNIGTVACEVAQAAKANEVLVTELSSFQLMGVETFQPKIAAFLNLFEAHLDYHGTKKEYGLAKANIFKNQTAADYSVINADDADVMELSANSKGQKILFSTTKEIEDGACIKENALYFKGEKVVEIKDIVLPGKHNLENILAAMSIAKLLGVANEAIVAVLKRFTGVKHRLEYVTTIHNRKFYNDSKATNILATEKALSAFTSPVILLAGGLDRGNEFDDLIPYFEQHVKAIVTYGQTAPKLVHAAEKAGLAIIEAVHHLEEAVERAYAHSADGDVILLSPACASWDQFKTFEERGDIFIQAVHKLI\n\n### Sequence 2:\nMKTVTEFQNKNILVLGIAKSGYAAATLLQKLGANVIVNDGKPLAENVLAAELQAKGMDVVCGGHPLELLERNISLVVKNPGIPYSNPILVAAKEKQIPIVTEVELAYRISEAPFVGITGSNGKTTTTMLTFEMLKEGQKHPVIAGNIGTVACEVAQDAKENEVVVTELSSFQLMGVELFQPKIAAFLNLFEAHLDYHGTKKEYGLAKANIFKNQTENDYSVINADDADVMALSAYSKGQKVLFSTTKEIEDGACIKDNALYFKAEKVVEVDDIVLPGQHNLENILAAMSIAKLLGVSNEAITAVLKRFTGVKHRLEYVTTINNRKFYNDSKATNMLATEKALSAFTQPTVLLAGGLDRGNEFDDLIPYF-KNVKAIVTFGQTAPKLVRAAEKAGLETIESVDTLDEAVVKAYAHSTDGDVILLSPACASWDQFKTFEERGDIFIQAVHKLI\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01858", "category": "Mutation", "instruction": "Describe the effect of mutation P94522-A181F on protein Extracellular endo-alpha-.", "input": "### Wild-type:\nMKKKKTWKRFLHFSSAALAAGLIFTSAAPAEAAFWGASNELLHDPTMIKEGSSWYALGTGLTEERGLRVLKSSDAKNWTVQKSIFTTPLSWWSNYVPNYGQNQWAPDIQYYNGKYWLYYSVSSFGSNTSAIGLASSTSISSGGWKDEGLVIRSTSSNNYNAIDPELTFDKDGNPWLAFGSAWSGIKLTKLDKSTMKPTGSLYSIAARPNNGGALEAPTLTYQNGYYYLMVSFDKCCDGVNSTYKIAYGRSKSITGPYLDKSGKSMLEGGGTILDSGNDQWKGPGGQDIVNGNILVRHAYD\n\n### Mutant:\nMKKKKTWKRFLHFSSAALAAGLIFTSAAPAEAAFWGASNELLHDPTMIKEGSSWYALGTGLTEERGLRVLKSSDAKNWTVQKSIFTTPLSWWSNYVPNYGQNQWAPDIQYYNGKYWLYYSVSSFGSNTSAIGLASSTSISSGGWKDEGLVIRSTSSNNYNAIDPELTFDKDGNPWLAFGSFWSGIKLTKLDKSTMKPTGSLYSIAARPNNGGALEAPTLTYQNGYYYLMVSFDKCCDGVNSTYKIAYGRSKSITGPYLDKSGKSMLEGGGTILDSGNDQWKGPGGQDIVNGNILVRHAYD\n\n### Answer:", "output": "Significantly more active than the wild-type arabinanase against both linear arabinan (50-fold) and the arabinooligosaccharides (1000-fold).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01859", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGSVAVEDQPDVSAVLSAYNQQGDPTMYEEYYSGLKHFIECSLDCHRAELSQLFYPLFVHMYLELVYNQHENEAKSFFEKFHGDQECYYQDDLRVLSSLTKKEHMKGNETMLDFRTSKFVLRISRDSYQLLKRHLQEKQNNQIWNIVQEHLYIDIFD\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCHHHHGGGSSCCCGGGHHHHHHHHHHHHHTSCHHHHHHHGGGHHHHHHHHHHHHHHTTCHHHHHHHHHHHSTTSCGGGHHHHHHHTTCCSHHHHTTCGGGGGGCGGGCEEEEEHHHHHHHHHHHTSSSSCHHHHHHHHTCEEEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01860", "category": "Cell", "instruction": "Infer the most likely cell identity from these markers.", "input": "CLCA1, MUC2, TFF3", "output": "This cell is most consistent with goblet cell, an intestinal goblet cell. Key markers such as MUC2, CLCA1, TFF3 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "goblet cell", "tissue": "lung", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["CLCA1", "MUC2", "TFF3"]}}
{"id": "eval_01861", "category": "Mutation", "instruction": "Describe the effect of mutation C8VSZ2-H149A on protein Alpha-ketoglutarate-dependent xanthine dioxygenase xanA.", "input": "### Wild-type:\nMPAITVKPLTPPAGSAIDFGAVITDVDLEHLTDGDFSTIRSALYTHLVVVLKNQHQLTPKAQYELTRRFDPSATQYGHGKTLDAKRSILHPDLKTIPHQPQVQVIGHGFIDSYEGLENITLKHPHHRTFHRDPIPQEDDYDSTRFYRWHIDAALYGLNPPIVTTLLAVKVPGGRRQTVRYDDGSGETMDVPLGTTAFASGERMFELLSEEDKEFALSSRVEYAPHPYIWMSPARSLPTGLGLHSDDLELPLSELPPIDESAIQILPMVWKNPATGKPALQIHPSAVRKIHCGDGTVIDDL\n\n### Mutant:\nMPAITVKPLTPPAGSAIDFGAVITDVDLEHLTDGDFSTIRSALYTHLVVVLKNQHQLTPKAQYELTRRFDPSATQYGHGKTLDAKRSILHPDLKTIPHQPQVQVIGHGFIDSYEGLENITLKHPHHRTFHRDPIPQEDDYDSTRFYRWAIDAALYGLNPPIVTTLLAVKVPGGRRQTVRYDDGSGETMDVPLGTTAFASGERMFELLSEEDKEFALSSRVEYAPHPYIWMSPARSLPTGLGLHSDDLELPLSELPPIDESAIQILPMVWKNPATGKPALQIHPSAVRKIHCGDGTVIDDL\n\n### Answer:", "output": "Impairs catalytic activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01862", "category": "Mutation", "instruction": "Describe the effect of mutation Q9Y375-R253K on protein Complex I intermediate-associated protein 30, mitochondrial.", "input": "### Wild-type:\nMALVHKLLRGTYFLRKFSKPTSALYPFLGIRFAEYSSSLQKPVASPGKASSQRKTEGDLQGDHQKEVALDITSSEEKPDVSFDKAIRDEAIYHFRLLKDEIVDHWRGPEGHPLHEVLLEQAKVVWQFRGKEDLDKWTVTSDKTIGGRSEVFLKMGKNNQSALLYGTLSSEAPQDGESTRSGYCAMISRIPRGAFERKMSYDWSQFNTLYLRVRGDGRPWMVNIKEDTDFFQRTNQMYSYFMFTRGGPYWQEVRIPFSKFFFSNRGRIRDVQHELPLDKISSIGFTLADKVDGPFFLEIDF\n\n### Mutant:\nMALVHKLLRGTYFLRKFSKPTSALYPFLGIRFAEYSSSLQKPVASPGKASSQRKTEGDLQGDHQKEVALDITSSEEKPDVSFDKAIRDEAIYHFRLLKDEIVDHWRGPEGHPLHEVLLEQAKVVWQFRGKEDLDKWTVTSDKTIGGRSEVFLKMGKNNQSALLYGTLSSEAPQDGESTRSGYCAMISRIPRGAFERKMSYDWSQFNTLYLRVRGDGRPWMVNIKEDTDFFQRTNQMYSYFMFTRGGPYWQEVKIPFSKFFFSNRGRIRDVQHELPLDKISSIGFTLADKVDGPFFLEIDF\n\n### Answer:", "output": "Decrease in MC1DN11; due to a nucleotide substitution located in the splice site consensus sequence at the end of exon 3; patient cells contain abnormally spliced transcripts corresponding to protein variant R-253 but also transcripts containing the final 6 base pairs of exon 3 and corresponding to protein variant 252-VK-253 del. Mutations in the NDUFAF1 gene at site result in enhanced complex I assembly at early stages, leading to increased levels and activity of complex I. This can prevent mitochondrial disease, presenting with cardioencephalomyopathy.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01863", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nIKVRENEPFDVALRRFKRSCEKAGVLAEVRRREFYEKPTTERKRAKASAVK\n\n### Answer:", "output": "<SEQ_3Di>DDDDDPDDDPDDPCCVVPPPDDPPNPVQPPDDDAVPNCVDVVVVVVVVVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01864", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "N#Cc1nc(NC2CCCC2)c(N)c(N2CCOCC2)n1", "output": "The molecule contains the following functional groups: ether, nitrile.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "N#Cc1nc(NC2CCCC2)c(N)c(N2CCOCC2)n1", "compound_name": null, "pubchem_cid": null, "MW": 288.36, "LogP": 1.12, "TPSA": 100.09, "QED": 0.86, "functional_groups": ["ether", "nitrile"], "moleculenet_labels": {"label": 2.59}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["ether", "nitrile"]}}
{"id": "eval_01865", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nADHI_CERS4\n\n### Answer:", "output": "Belongs to the zinc-containing alcohol dehydrogenase family. Class-III subfamily.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01866", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMGRVKLEIKRIENTTNRQVTFSKRRNGLIKKAYELSILCDIDIALIMFSPSDRLSLFSGKTRIEDVFSRFINLPKQERESALYFPDQNRRPDIQNKECLLRILQQLKTENDIALQVTNPAAINSDVEELEHEVCRLQQQLQMAEEELRRYEPDPIRFTTMEEYEVSEKQLLDTLTHVVQRRDHLMSNHLSSYEASTMQPNIGGPFVNDVVEGWLPENGTNQTHLFDASAHSNQLRELSSAMYEPLLQGSSSSSNQNNMSECHVTNHNGEMFPEWAQAYSSSALFASMQQQHEGVGPSIEEMMPAQQSDIPGVTAETQVDHEVSDYETKVPQLSSQ\n\n### Sequence 2:\nMGRVKLEIKRIENTTNRQVTFSKRRNGLIKKAYELSILCDIDIALLMFSPSDRLSLFSGKTRIEDVFSRYINLSDQERENALVFPDQSRRPDFQSKEYLLRTLQQLKAENDIALQLTNPTAINSDVEELEHEVYKLQQQLLMAEEELRKYEPDPIRFTTMEEYETCEKQLMDTLTRVNQRREHILSQDQLSSYEASALQQQQSMGGPFGNDVVGGWLTENGPNEAHLFDASAH--------SAMYETLLQGSSSSSNQNNIMGESNVSNHNGDMFQEWAQAYNSTTAHNPSTLFPPMQHQHGLVVDPNIEEI------EIPVMKKDAQADHEVSDYDIRMPQLSSQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01867", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nZFP60_MOUSE\n\n### Answer:", "output": "Belongs to the krueppel C2H2-type zinc-finger protein family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01868", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nSPEA_PHOLL\n\n### Answer:", "output": "Biosynthetic arginine decarboxylase, ADC.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01869", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRL23_ECOL5\n\n### Answer:", "output": "50S ribosomal protein L23.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01870", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: VEGFR2\nLigand SMILES: CO/N=C(\\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(C[N+]3(C)CCCC3)CS[C@H]12)c1csc(N)n1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and vascular endothelial growth factor receptor 2 (VEGFR2) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "CO/N=C(\\C(=O)N[C@@H]1C(=O)N2C(C(=O)O)=C(C[N+]3(C)CCCC3)CS[C@H]12)c1csc(N)n1", "compound_name": null, "pubchem_cid": null, "MW": 481.58, "LogP": 0.05, "TPSA": 147.21, "QED": 0.21, "functional_groups": ["amide", "hydroxyl"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "VEGFR2", "has_evidence": false}}
{"id": "eval_01871", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMARIAGVNIPTNKRVVIALQYIHGIGAKKAEEITQKVNIPAERRVNQLTDAEVLQIRETIDRDYLVEGDLRREVSMNIKRLMDLGAYRGLRHRKQLPVRGQRTHTNARTRKGKAKPIAGKKK\n\n### Sequence 2:\nMARIAGVNIPTNKRVVIALQYIHGIGAKKAEEITQKVNIPAERRVNQLTDAEVLQIRETIDRDYLVEGDLRREVSMNIKRLMDLGAYRGLRHRKQLPVRGQRTHTNARTRKGKAKPIAGKKK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01872", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nY329_MYCPN\n\n### Answer:", "output": "Uncharacterized protein MG236 homolog.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01873", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "PF4, PPBP, GP1BA, ITGA2B", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as ITGA2B, PPBP, PF4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "brain", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PF4", "PPBP", "GP1BA", "ITGA2B"]}}
{"id": "eval_01874", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nSYA_RHILO\n\n### Answer:", "output": "Belongs to the class-II aminoacyl-tRNA synthetase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01875", "category": "Cell", "instruction": "Given the expression strengths of marker genes, what cell type is this?", "input": "<cell>\nPDGFRA:low LUM:high COL3A1:low\n</cell>", "output": "This cell is most consistent with fibroblast, a fibroblast cell. Key markers such as COL3A1, PDGFRA, LUM support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "bin_to_type", "source": "panglao_cellmarker_panels", "cell_type": "fibroblast", "tissue": "lymph node", "disease": "normal", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_expression_bins", "top_genes": ["PDGFRA", "LUM", "COL3A1"]}}
{"id": "eval_01876", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMEPIIRIRNLHKWFGPLHVLKGIHLEVAPGEKLVIIGPSGSGKSTLIRTINRLEDFQEGEVVVDGLSVKDDRALREIRREVGMVFQQFNLFPHMTVLENVTLAPMRVRRWPREKAEKKALELLERVGILDQARKYPAQLSGGQQQRVAIARALAMEPKIMLFDEPTSALDPEMVGEVLDVMRDLAQGGMTMVVVTHEMGFAREVADRVVFMDGGQIVEEGRPEEIFTRPKEERTRSFLQRVLHH\n\n### Answer:", "output": "<SEQ_2D>CCCSEEEEEEEEEETTEEEEEEEEEEECTTCEEEEECSTTSSHHHHHHHHTTSSCCSEEEEEETTEETTCGGGHHHHHHHEEEECSSCCCCTTSCHHHHHHHHHHHTSCCCHHHHHHHHHHHHHHTTCGGGTTSCGGGSCHHHHHHHHHHHHHTTCCSEEEEESTTTTSCHHHHHHHHHHHHHHHTSSCEEEEECSCHHHHHHHCSEEEEEETTEEEEEECHHHHHHSCCSHHHHHHHHHHTCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01877", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMQDTIPVLTIDGPSGAGKGTAARAVAARLGWNFLDSGAIYRALAVAAVDRGVSREDESALAALAASLDLVFGADSTARILLWDADISGRIVTEECGNLASKLAAFPAVRQALLDKQRGFRRPPGLVADGRDMGTVVFPDAPYKVFLTASAEVRARRRYNQLKEKGMDVSLAHLTEEIEERDRRDRERQIAPLRAAADAVVIDSSDLSVDEVIQVCLSVVQSH\n\n### Sequence 2:\nMQESTPVITIDGPSGAGKGTVARIVADQLGWHLLDSGAIYRVLAVAIQHHQLSLDDEEPLIPMAAHLDVQFEINSQGEAKVILEGENVTEIIRTEEVGGLASKVAAFPRVREALLRRQRAFSVSPGLIADGRDMGTVVFPKAPVKIFLTASAEERADRRFNQLKEKGIDVNIGRLLDDIRQRDERDQNRKVAPLIPAEGALTIDSTDISITEVVNKIL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01878", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: ALK\nLigand SMILES: Cc1ccc(S(=O)(=O)NC(=O)NN2CCCCCC2)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "Cc1ccc(S(=O)(=O)NC(=O)NN2CCCCCC2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 311.41, "LogP": 1.77, "TPSA": 78.51, "QED": 0.89, "functional_groups": ["aromatic ring", "amide", "amine", "sulfonamide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01879", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMYNIYKSLNQQVQRALNTAFPEAASQVKESGDFLNPQLVAATKPEFGDFQINGALALARIIKKSPRQIAEILIKQLESNEVFKAICLPPEIAGPGFINLTLQNTCLINEITSRLNDDLLGVPLVNDDEITKKLKPVIVDFSSPNIAKEMHVGHLRSTIIGDSIARILNYRGYKVIRLNHVGDWGTQFGMLITHLKEVAPKALTTANVINLGNLVEFYKKAKQRFDEDEYFQQCSRNEVVNLQRGNKESLKAWELLCEQSRKEFNKIYDRLKIEISERGESFYNPFLQGVIDDLTRSGLLVEDDGAKCVFLNGINGKDGNPLPLIIQKADGGFNYATTDLAAIRYRLKDQPDGDGAGRIIYVTDSGQANHFAGVFQVAKRAKWLPSSSRIEHVPFGLVQGEDGKKLKTRSGETVRLKDLLDEAISRAKLDIERRLNEENRKESQAFIEKVSNTIGIAAVKYADLSQNRITNYQFSFDRMLALQGNTAPYLLYAVVRIAGINRKGGDLHSSVNKLNFSEPQEWRLIRELLKFDEVIIAVEEELLPNRLCNYLFELSQVFNRFYDQIPVLKAEEPSRSCRLALCQLTGDTLKKGLNLLGISTLERM\n\n### Sequence 2:\nMMRSHYCGQLNESLEGQEVTLCGWVHRRRDHGGVIFLDIRDREGLAQVVFDPDRAETFAKADRVRSEYVVRITGKVRPRPAGAVNPNMASGAIEVLGYELDVLNQAETPPFPLDEYSDVGEETRLRYRFIDLRRPEMAAKLKLRSSITSSIRRYLDENGFLDVETPILTRATPEGARDYLVPSRTHAGSFFALPQSPQLFKQLLMVAGFDRYYQIAKCFRDEDLRADRQPEFTQIDIETSFLDEADIMGITENMIRKLFKEVLDVEFGELPHMTFEEAMRRYGSDKPDLRIPLELVDVADQLKAVEFKVFSGPANDPKGRVAALRVPGAASMPRSQIDDYTRFVGIYGAKGLAYIKVNERAKGVEGLQSPIVKFIPKENLDVILDRVGAVDGDIVFFGADKAKIVSEALGALRIRLGHDLKLLTCEWAPLWVVDFPMFEENDDGSLSALHHPFTAPKCTPEELAANPAVALSRAYDMVLNGTELGGGSIRIHRKEMQQAVFRILGIDEAEQQEKFGFLLDALKYGAPPHGGLAFGLDRLVMLMTGASSIREVIAFPKTQSAACVMTQAPGAVDAKALRELHIRLREQPKAE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01880", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "SFTPA1, SFTPC, ABCA3, SFTPB", "output": "The expression pattern is consistent with alveolar type II cell in a context associated with colorectal carcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "alveolar type II cell", "tissue": "spleen", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["SFTPA1", "SFTPC", "ABCA3", "SFTPB"]}}
{"id": "eval_01881", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMPQSSRYSDEHVEQLLSELVSVLEKHRTPTDLSLMVLGNMVTNLINTSIAPAQRKVLARSFAEALQASVREDKAH\n\n### Sequence 2:\nMPQHSRYSDEHVEQLLSDLLNVLEKHKAPTDLSLMVLGNMVTNLINTSIAPAQRQAIANSFARALQSSINDDKAH\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01882", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "COc1cccc(C(=O)NC2C[C@H]3CC[C@@H](C2)N3Cc2ccccc2)c1OC", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, amine, ether. Its descriptor profile: MW=380.49, LogP=3.63, TPSA=50.8, HBD=1, HBA=4, QED=0.83.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "COc1cccc(C(=O)NC2C[C@H]3CC[C@@H](C2)N3Cc2ccccc2)c1OC", "compound_name": null, "pubchem_cid": null, "MW": 380.49, "LogP": 3.63, "TPSA": 50.8, "QED": 0.83, "functional_groups": ["aromatic ring", "amide", "amine", "ether"], "moleculenet_labels": {"label": 1}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01883", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nY751_MYCTU\n\n### Answer:", "output": "Exhibits S-adenosyl-L-methionine-dependent methyltransferase activity.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01884", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRL9_MYCLE\n\n### Answer:", "output": "Binds to the 23S rRNA.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01885", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nAVNGKGMNPDYKAYLMAPLKKIPEVTNWETFENDLRWAKQNGFYAITVDFWWGDMEKNGDQQFDFSYAQRFAQSVKNAGMKMIPIISTHQCGGNVGDDCNVPIPSWVWNQKSDDSLYFKSETGTVNKETLNPLASDVIRKEYGELYTAFAAAMKPYKDVIAKIFLSGGPAGELRYPSYTTSDGTGYPSRGKFQAYTEFAKSKFRLWVLNKYGSLNEVNKAWGTKLISELAILPPSDGEQFLMNGYLSMYGKDYLEWYQGILENHTKLIGELAHNAFDTTFQVPIGAKIAGVHWQYNNPTIPHGAEKPAGYNDYSHLLDAFKSAKLDVTFTCLEMTDKGSYPEYSMPKTLVQNIATLANEKGIVLNGENALSIGNEEEYKRVAEMAFNYNFAGFTLLRYQDVMYNNSLMGKFKDLLGVTPVMQTIVVKNVPTTIGDTVYITGNRAELGSWDTKQYPIQLYYDSHSNDWRGNVVLPAERNIEFKAFIKSKDGTVKSWQTIQQSWNPVPLKTTSHTSSW\n\n### Answer:", "output": "<SEQ_2D>CGGGCCCCTTCEEEEECCSSCGGGTSCHHHHHHHHHHHHHTTEEEEEEEEEHHHHTCSSTTCCCCHHHHHHHHHHHHHTCEEEEEEECSCBSSSTTCCCCBCCCGGGGGGCSSSCSSEECTTCCEECSSCCTTCHHHHHHHHHHHHHHHHHHHGGGGGGBCCEEECCSGGGBSSCCCSCTTTTCCTTSCCCCCCCSHHHHHHHHHHHHHHHSSHHHHHHHHTCCCSSGGGCCSCSCHHHHHHTGGGSHHHHHHHHHHHHHHHHHHHHHHHHHHHHHTTTTCCCEEEECCCCCTTTTCSSSTTTTHHHHTCSCHHHHHHHHHHHTCEEEECCTTCCCCCCTTTCCCHHHHHHHHHHHHHHHTCCEEEECSSCCCSHHHHHHHHHHHHHTTCSEEEESCTHHHHTCHHHHHHHHHHTSCCEEEEEEEEESCCCCTTCEEEEEESSGGGTTTCSSSSCEECEEETTTTEEEEEEEEETTCCEEEEEEEECTTSCEEEECSSCEEESSCCSSCCEEEEEC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01886", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAEKLPPEVQAQLAKFQQLKDQLDRLLLEKSTIENELREINKVLEELSVLNADATIYKIVGNLLVKSDKTSVEKELNDRKELLELRSRTYQKQESILRKQLEDLQAKINEMLSKYYPQGGQTGIKA\n\n### Sequence 2:\nQIPPSLQDLVNRFNQVQAQLQNVLLRKQQYEAELREIEKALTEIEKLPQDAKIYKSVGNFLIPQNRDAALQELRERKELLELHTKTLARQESLLREQIEKLREEINKELSR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01887", "category": "Mutation", "instruction": "Describe the effect of mutation B1PVZ9-D62G on protein Arginine kinase Met e 2.", "input": "### Wild-type:\nMADAAVIEKLEAGFKKLEAATDCKSLLKKYLTKEVFDKLKDKKTSLGATLLDVIQSGVENLDSGVGIYAPDAEAYTLFAPLFDPIIEDYHVGFKQTDKHPNKDFGDVNSFVNVDPEGKFVISTRVRCGRSMQGYPFNPCLTESQYKEMEAKVSSTLSSLEGELKGTYYPLTGMSKEVQQKLIDDHFLFKEGDRFLQAANACRYWPAGRGIYHNDNKTFLVWVNEEDHLRIISMQMGGDLGQVFRRLTSAVNEIEKRIPFSHHDRLGFLTFCPTNLGTTVRASVHIKLPKLAANREKLEEV\n\n### Mutant:\nMADAAVIEKLEAGFKKLEAATDCKSLLKKYLTKEVFDKLKDKKTSLGATLLDVIQSGVENLGSGVGIYAPDAEAYTLFAPLFDPIIEDYHVGFKQTDKHPNKDFGDVNSFVNVDPEGKFVISTRVRCGRSMQGYPFNPCLTESQYKEMEAKVSSTLSSLEGELKGTYYPLTGMSKEVQQKLIDDHFLFKEGDRFLQAANACRYWPAGRGIYHNDNKTFLVWVNEEDHLRIISMQMGGDLGQVFRRLTSAVNEIEKRIPFSHHDRLGFLTFCPTNLGTTVRASVHIKLPKLAANREKLEEV\n\n### Answer:", "output": "Significantly reduced catalytic activity in the forward reaction. Altered three-dimensional conformation and impaired structural stability of the protein. The mutant had a pronounced loss in activity. The conformation of the mutant had been changed and the stability in a urea solution was also reduced. We speculated that the hydrogen bond and electrostatic interactions formed between residues 62 and 193 play a key role in stabilizing the structure and mediating the synergism in substrate binding of arginine kinase from greasyback shrimp (Mapenaeus ensis).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01888", "category": "Mutation", "instruction": "Describe the effect of mutation Q12834-A72S on protein Cell division cycle protein 20 homolog.", "input": "### Wild-type:\nMAQFAFESDLHSLLQLDAPIPNAPPARWQRKAKEAAGPAPSPMRAANRSHSAGRTPGRTPGKSSSKVQTTPAKPGGDRYIPHRSAAQMEVASFLLSKENQPENSQTPTKKEHQKAWALNLNGFDVEEAKILRLSGKPQNAPEGYQNRLKVLYSQKATPGSSRKTCRYIPSLPDRILDAPEIRNDYYLNLVDWSSGNVLAVALDNSVYLWSASSGDILQLLQMEQPGEYISSVAWIKEGNYLAVGTSSAEVQLWDVQQQKRLRNMTSHSARVGSLSWNSYILSSGSRSGHIHHHDVRVAEH\n\n### Mutant:\nMAQFAFESDLHSLLQLDAPIPNAPPARWQRKAKEAAGPAPSPMRAANRSHSAGRTPGRTPGKSSSKVQTTPSKPGGDRYIPHRSAAQMEVASFLLSKENQPENSQTPTKKEHQKAWALNLNGFDVEEAKILRLSGKPQNAPEGYQNRLKVLYSQKATPGSSRKTCRYIPSLPDRILDAPEIRNDYYLNLVDWSSGNVLAVALDNSVYLWSASSGDILQLLQMEQPGEYISSVAWIKEGNYLAVGTSSAEVQLWDVQQQKRLRNMTSHSARVGSLSWNSYILSSGSRSGHIHHHDVRVAEH\n\n### Answer:", "output": "Partially gain of BUB1-mediated phosphorylation and inhibition and fully functional spindle-assembly checkpoint; when associated with A-41; A-92; A-153; A-157 and A-161.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01889", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "MRC1, APOE, SPP1, CD163, CD206", "output": "This cell is most consistent with tumor-associated macrophage, a tumor-associated macrophage (TAM). Key markers such as CD206, MRC1, APOE support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "tumor-associated macrophage", "tissue": "ovary", "disease": "COVID-19", "organism": "Homo sapiens", "assay": "Smart-seq2", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["MRC1", "APOE", "SPP1", "CD163", "CD206"]}}
{"id": "eval_01890", "category": "Mutation", "instruction": "Describe the effect of mutation Q9C6B8-S290A on protein Auxin efflux carrier component 1.", "input": "### Wild-type:\nMITAADFYHVMTAMVPLYVAMILAYGSVKWWKIFTPDQCSGINRFVALFAVPLLSFHFIAANNPYAMNLRFLAADSLQKVIVLSLLFLWCKLSRNGSLDWTITLFSLSTLPNTLVMGIPLLKGMYGNFSGDLMVQIVVLQCIIWYTLMLFLFEYRGAKLLISEQFPDTAGSIVSIHVDSDIMSLDGRQPLETEAEIKEDGKLHVTVRRSNASRSDIYSRRSQGLSATPRPSNLTNAEIYSLQSSRNPTPRGSSFNHTDFYSMMASGGGRNSNFGPGEAVFGSKGPTPRPSNYEEDGGPAK\n\n### Mutant:\nMITAADFYHVMTAMVPLYVAMILAYGSVKWWKIFTPDQCSGINRFVALFAVPLLSFHFIAANNPYAMNLRFLAADSLQKVIVLSLLFLWCKLSRNGSLDWTITLFSLSTLPNTLVMGIPLLKGMYGNFSGDLMVQIVVLQCIIWYTLMLFLFEYRGAKLLISEQFPDTAGSIVSIHVDSDIMSLDGRQPLETEAEIKEDGKLHVTVRRSNASRSDIYSRRSQGLSATPRPSNLTNAEIYSLQSSRNPTPRGSSFNHTDFYSMMASGGGRNSNFGPGEAVFGSKGPTPRPANYEEDGGPAK\n\n### Answer:", "output": "Apical-to-basal shift in polar targeting and increased auxin accumulation in the root tips; when associated with A-231 and A-252.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01891", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMRNYDLSPLLRQWIGFDKLASTMQGGQEPQGFPPYNIEKTDDNHYRISLALAGFKQSELDIEVEGPRLTVRGKPTPVEKQVEYLHQGLVRKEFSLTFTLAEHLNVDNAQFENGLLHIDLLRQVPEALQPQRIAIGSATPQERQVLESPEAPDQQ\n\n### Sequence 2:\nMRNYDLSPLLRQWIGFDKLASSM-GGQEPQGFPPYNIEKSDDNHYRISLALAGFKQSELDIEVEGPRLTVRGKPTPSEKQVEYLHQGLVCKEFALTFTLAEHLQVSEAQFENGLLHIDLVRQVPEALQPQRIAIGTT\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01892", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nFABA_XANOM\n\n### Answer:", "output": "3-hydroxydecanoyl-[acyl-carrier-protein] dehydratase, 3-hydroxyacyl-[acyl-carrier-protein] dehydratase FabA, Beta-hydroxydecanoyl thioester dehydrase, Trans-2-decenoyl-[acyl-carrier-protein] isomerase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01893", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nMESPRIELRSDITVELVDSSASDLAVVKAARVSTAGEDANDELYDGGSTRGLIRYLMRSRHGSPFEHNSMTFLVRAPIFTVRHLMRHRTWSFNEESARYREVGAAFYVPDATRLLRQEGKPGDYRYVGGSTDDHQQVVRSATRAYEVAFEEYQRLLDSGIAREIARLVLPVSTYSVLYATCNARALMHFLSLRTHRPDAAYVSHPQREIEMVAEQMETAWAKLMPVTHEAFTAFGRVSP\n\n### Answer:", "output": "<SEQ_2D>CCCCCCEEECCCEEEEEEEECCHHHHHHHHTTCCCCCCCCCCCCCCCCHHHHHHHHHHTTCCGGGGSCEEEEEEEEEHHHHHHHTTCTTSEEEECCTTTSCCCSEEECCCTTSCCEEESCGGGCEEECCCHHHHHHHHHHHHHHHHHHHHHHHHHHHTTCCHHHHGGGSBTTBEEEEEEEECHHHHHHHHHHHSCCTTSSSCCCCCHHHHHHHHHHHHHHHHHCHHHHHHHHHTTSCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01894", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMNFIYQKKSPKTKEIEQLAKTQNIKIIRINTNELDKILKNKDHRGFALKLKLEKNKNVKTQTKDFENLLETFKKKENAFILLLDEIEDPQNFGAILRTAEQFSIDLVITTQKRSAKDNSTVLRTSSGASQYVKKMTVTNINNTINLLKNYGFWIYTGDIKGQDINKIKINDKKIALILGNEGKGVHKLIKENSDFLIRIPTSGKIDSLNVSVSTGILIFEIKRQLNLL\n\n### Sequence 2:\nYIQEGKSDARTREVLEKAANENIRVYFADADRLNAISKGARHQGVVGFIDASKNH------VHLEDVLENLS--EPPLLLILDGITDPHNLGACLRTADAMGVHAVIAPKDKSAGLNATVSKVACGAAETVPYITVTNLARTLRELKEYGIWIIGTDMSGESDLYHCNLPDSA-AWVMGNEGDGMRRLTREHCDMLVSIPMFGTVESMNVSVSAGMVLSETRRQ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01895", "category": "Cell", "instruction": "What disease context is this cell most associated with?", "input": "VIL1, APOB, FABP2", "output": "The expression pattern is consistent with enterocyte in a context associated with lung adenocarcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "enterocyte", "tissue": "heart", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x 3' v3 scRNA-seq", "split": "train", "generation_method": "template_disease_state", "top_genes": ["VIL1", "APOB", "FABP2"]}}
{"id": "eval_01896", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: ALK\nLigand SMILES: O=[N+]([O-])c1cccnc1-n1cnc2ccccc21\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "O=[N+]([O-])c1cccnc1-n1cnc2ccccc21", "compound_name": null, "pubchem_cid": null, "MW": 240.22, "LogP": 2.33, "TPSA": 73.85, "QED": 0.51, "functional_groups": ["aromatic ring", "nitro", "pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01897", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: EGFR\nLigand SMILES: OC1(c2ccccc2)CSC2=Nc3ccccc3C3CCN1N23\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "HIV", "canonical_smiles": "OC1(c2ccccc2)CSC2=Nc3ccccc3C3CCN1N23", "compound_name": null, "pubchem_cid": null, "MW": 323.42, "LogP": 3.24, "TPSA": 39.07, "QED": 0.87, "functional_groups": ["aromatic ring", "amine", "hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01898", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "NCCCNCCSP(=O)(O)O", "output": "This molecule is small and polar, with hydrophilic character. It contains amine, hydroxyl. Its descriptor profile: MW=214.23, LogP=-0.25, TPSA=95.58, HBD=4, HBA=4, QED=0.34.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "BBBP", "canonical_smiles": "NCCCNCCSP(=O)(O)O", "compound_name": null, "pubchem_cid": null, "MW": 214.23, "LogP": -0.25, "TPSA": 95.58, "QED": 0.34, "functional_groups": ["amine", "hydroxyl"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01899", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMSRVLRLGVNIDHVATIRNARGGDHPDPLRAAKIAAEAGADGITAHLREDRRHIGDRDIARLRDEIDLPLNLEMAATAEMLAIALRHRPHAACIVPEKREERTTEGGLDVVGGAVHLAPIITALGEAGVRVSLFIEPDLRQLDAARALGAPVVELHTGAYCDAAAGPAREAQFLRIDKAARHAQALGLECHAGHGLTYDTVEPVAALPAIAELNIGHFLIGEAIFVGLHQAIARMRAHMDAALRGGVAA\n\n### Sequence 2:\nLRLGVNIDHVATLRNARGGALPDPVRAAVLAEQAGADGITAHLREDRRHIRDADIAAIMEAITIPLNFEMAATDEMQEIALGHLPHAACIVPERREERTTEGGLEVAGDDNRLAQYIAPLKDAGIRVSLFIAAERSQIEAAARIGAPVIELHTGAYCDYDAEGDVAARDAELARLIAGAKLGVELGLEVHMGHGLNYDTVAPIAALPEVAELNIGHFLIGESIFVGLEAAMTEMRARMDAA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01900", "category": "Mutation", "instruction": "Describe the effect of mutation P0A794-G194S on protein Pyridoxine 5'-phosphate synthase.", "input": "### Wild-type:\nMAELLLGVNIDHIATLRNARGTAYPDPVQAAFIAEQAGADGITVHLREDRRHITDRDVRILRQTLDTRMNLEMAVTEEMLAIAVETKPHFCCLVPEKRQEVTTEGGLDVAGQRDKMRDACKRLADAGIQVSLFIDADEEQIKAAAEVGAPFIEIHTGCYADAKTDAEQAQELARIAKAATFAASLGLKVNAGHGLTYHNVKAIAAIPEMHELNIGHAIIGRAVMTGLKDAVAEMKRLMLEARG\n\n### Mutant:\nMAELLLGVNIDHIATLRNARGTAYPDPVQAAFIAEQAGADGITVHLREDRRHITDRDVRILRQTLDTRMNLEMAVTEEMLAIAVETKPHFCCLVPEKRQEVTTEGGLDVAGQRDKMRDACKRLADAGIQVSLFIDADEEQIKAAAEVGAPFIEIHTGCYADAKTDAEQAQELARIAKAATFAASLGLKVNAGHSLTYHNVKAIAAIPEMHELNIGHAIIGRAVMTGLKDAVAEMKRLMLEARG\n\n### Answer:", "output": "In pdxH null mutation suppressor. The mutation in the PdxJ polypeptide allows growth of a delta pdxH::omega null mutant in the absence of pyridoxal. This suggests that PdxJ may participate in formation of the pyridine ring of pyridoxine 5'-phosphate.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01901", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMFHILRLESTVDLSEPLKDNGIIVFQSDKLDLEPSPNLGPTGIDNTNVNLINAKGDVLLHIGIRRRENAFVFNSIPYGESRGPEERIPLEGTFGDRRDPSITIFDHPDRYQIMIDYKTVYYYKKRLEGRCEKVSYKINEGQTPPFSDVLGVTVLYFANVMPRA\n\n### Answer:", "output": "<SEQ_3Di>DKDKAFEFDKDFDPDFADAFKKKKKKFQAFALDQWPDCPPVNFWWKWKFFAFPVGWTQWIWIDGSVVQWIWIWIGGVPGDIDDTDIDGNPPQDDPDGIKMWMWTGHDFWIFIHIVPHTDDIGGTDDDHTGGMMHIYTPPPTDDRGDSMIMMDMGHPVVVDDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01902", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nVHEIVLTQSPGTLSLSPGEKATLSCRASQSIASTYLAWYQQKPGQAPRLLLHQGYNRATGIPDRFSGSGSGTVYTLTISGLEPDDFAVYYCQHLGTSPPYTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGE\n\n### Answer:", "output": "<SEQ_3Di>DPDKAKAKPDQEAAEAFFAKDKIKIFIPWFDQLQFKWKWWDAPPDDIDTQDHSFFHGDPPHDPQWGKDDGTGMIITMGRGGDPVRQTKMKMKGGDPDDDIDIYPIYGYAHDDDFDFWDKDKDWADPVVLVVFKTKIKMKTPFGPDPDKDKWKAFPNRTDDDQKDKDKDQQDPPSRGIMMMMIGMDTSVVVVPTFKIWIWMDDPNDPDIDIDIGTPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01903", "category": "Mutation", "instruction": "Describe the effect of mutation Q8N307-T442I on protein Mucin-20.", "input": "### Wild-type:\nMGCLWGLALPLFFFCWEVGVSGSSAGPSTRRADTAMTTDDTEVPAMTLAPGHAALETQTLSAETSSRASTPAGPIPEAETRGAKRISPARETRSFTKTSPNFMVLIATSVETSAASGSPEGAGMTTVQTITGSDPREAIFDTLCTDDSSEEAKTLTMDILTLAHTSTEAKGLSSESSASSDSPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPV\n\n### Mutant:\nMGCLWGLALPLFFFCWEVGVSGSSAGPSTRRADTAMTTDDTEVPAMTLAPGHAALETQTLSAETSSRASTPAGPIPEAETRGAKRISPARETRSFTKTSPNFMVLIATSVETSAASGSPEGAGMTTVQTITGSDPREAIFDTLCTDDSSEEAKTLTMDILTLAHTSTEAKGLSSESSASSDSPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPVITPSRASESSASSDGPHPV\n\n### Answer:", "output": "The production of MUC20 is correlated with development and progression of IgAN and other renal injuries.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01904", "category": "Literature", "instruction": "What are the subcellular locations of this protein?", "input": "### Protein Sequence:\nSYL_STRMK\n\n### Answer:", "output": "Cytoplasm.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01905", "category": "Mutation", "instruction": "Describe the effect of mutation P77754-L119H on protein Periplasmic chaperone Spy.", "input": "### Wild-type:\nMRKLTALFVASTLALGAANLAHAADTTTAAPADAKPMMHHKGKFGPHQDMMFKDLNLTDAQKQQIREIMKGQRDQMKRPPLEERRAMHDIIASDTFDKVKAEAQIAKMEEQRKANMLALMETQNKIYNILTPEQKKQFNANFEKRLTERPAAKGKMPATAE\n\n### Mutant:\nMRKLTALFVASTLALGAANLAHAADTTTAAPADAKPMMHHKGKFGPHQDMMFKDLNLTDAQKQQIREIMKGQRDQMKRPPLEERRAMHDIIASDTFDKVKAEAQIAKMEEQRKANMLAHMETQNKIYNILTPEQKKQFNANFEKRLTERPAAKGKMPATAE\n\n### Answer:", "output": "Decreased chaperone activity, substrate affinity and chaperone flexibility.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01906", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAEPTLASGWYPRLVVGLGNPGKNYGRTRHNVGFMVANLLAVRLGSKFEVHKRSGADVVNGRLAGCSMLVAKPRNYMNESGQQVGLLAKLYSVTPADIIIVHDDLDLDFGRIRLKLGGGEGGHNGLRSVAAALGTKDFQRVRIGIGRPTGRKDPASFVLENFTTAERMQVPTICKRAADATELLVGLGLEPAQNHVHAW\n\n### Sequence 2:\nMAKATTIKEALAKWEERTGQKAGEAKEVKLYAQIPPLEKMDASLSTLVNCEKLSLSTNCIEKIANLNGLKYLKILSLGRNNIKNLNGLEAVGETLEELWISYNLIEKLKGIHVMKKLKVLYMSNNLVKDWAEFSKLGELPLLGDIVFVGNPLEEKHTAEGNWMEEAVKRLPKLKKLDGNPVIKQEEEEGDES\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01907", "category": "Mutation", "instruction": "Describe the effect of mutation Q2LKW6-I744A on protein NACHT, LRR and PYD domains-containing protein 1b allele 1.", "input": "### Wild-type:\nMEESPPKQKSNTKVAQHEGQQDLNTTRHMNVELKHRPKLERHLKLGMIPVVYMKQGEEILYPAQSLREENLIQNFTSLLLLQKLCPKDPENMIRKSWASCVPEEGGHMINIQDLFGPNIGTQKEPQLVIIEGAAGIGKSTLARLVKRAWKEGQLYRDHFQHVFFFSCRELAQCKKLSLAELIAQGQEVPTAPINQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQRQPVHTLLGSLLGKSILPEAFFLLTTRTTALQKFIPSLPMPCQVEVLGFSGIERENYFYKYFANQRHAIT\n\n### Mutant:\nMEESPPKQKSNTKVAQHEGQQDLNTTRHMNVELKHRPKLERHLKLGMIPVVYMKQGEEILYPAQSLREENLIQNFTSLLLLQKLCPKDPENMIRKSWASCVPEEGGHMINIQDLFGPNIGTQKEPQLVIIEGAAGIGKSTLARLVKRAWKEGQLYRDHFQHVFFFSCRELAQCKKLSLAELIAQGQEVPTAPINQILSHPEKLLFILDGIDEPAWVLADQNPELCLHWSQRQPVHTLLGSLLGKSILPEAFFLLTTRTTALQKFIPSLPMPCQVEVLGFSGIERENYFYKYFANQRHAIT\n\n### Answer:", "output": "Spontaneous IL1B release under basal conditions, loss of response to metabolic inhibitors and to LT; when associated with A-746.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01908", "category": "Mutation", "instruction": "Describe the effect of mutation O75340-A47E on protein Programmed cell death protein 6.", "input": "### Wild-type:\nMAAYSYRPGPGAGPGPAAGAALPDQSFLWNVFQRVDKDRSGVISDTALQQALSNGTWTPFNPVTVRSIISMFDRENKAGVNFSEFTGVWKYITDWQNVFRTYDRDNSGMIDKNELKQALSGFGYRLSDQFHDILIRKFDRQGRGQIAFDDFIQGCIVLQRLTDIFRRYDTDQDGWIQVSYEQYLSMVFSIV\n\n### Mutant:\nMAAYSYRPGPGAGPGPAAGAALPDQSFLWNVFQRVDKDRSGVISDTELQQALSNGTWTPFNPVTVRSIISMFDRENKAGVNFSEFTGVWKYITDWQNVFRTYDRDNSGMIDKNELKQALSGFGYRLSDQFHDILIRKFDRQGRGQIAFDDFIQGCIVLQRLTDIFRRYDTDQDGWIQVSYEQYLSMVFSIV\n\n### Answer:", "output": "Increases interaction with SEC31A and PLSCR3, and localization to the endoplasmic reticulum; when associated with A-114.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01909", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMATIFDILNSINNKNNNHSNYYSCKRLKSNNNNNNNNFRNNSIKRIDIIPSMDVTITDNNIIIETELAGISKDNIEIDIKDSILTIQGEKKINNNYKQQQQQHSDKTKTNNELPLIEENKKEFRKYLSERSFGNFKRCLDLTSILYQLDLSTIKSNFENGLLIITINKKSDLSNSSFKININ\n\n### Sequence 2:\nMATLFDILNTLNN-----NNYFENCKRQCSINKSNKT--------IIDIIPPMDVTMTNDKLIIETELAGISKDHIEIDIKDSILTIQGEKKKNLNKQQQQLVIEKSTTSSTTLDSKEDEASIEEFEDDIKPKSKSTVTTTATKENKEDENKTKLSDKKFISERSFGNFKRYLDLTKVLYQLDLNSINTQFENGLLTITINKKLHYSN-TIKININ\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01910", "category": "Mol", "instruction": "Identify the functional groups in the given SMILES.", "input": "O=[N+]([O-])c1cccc2ccc(C(Cl)(Cl)Cl)nc12", "output": "The molecule contains the following functional groups: aromatic ring, nitro, halide, pyridine.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Lipophilicity", "canonical_smiles": "O=[N+]([O-])c1cccc2ccc(C(Cl)(Cl)Cl)nc12", "compound_name": null, "pubchem_cid": null, "MW": 291.52, "LogP": 3.97, "TPSA": 56.03, "QED": 0.45, "functional_groups": ["aromatic ring", "nitro", "halide", "pyridine"], "moleculenet_labels": {"label": 3.7}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "nitro", "halide", "pyridine"]}}
{"id": "eval_01911", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLLSDRDIRAEIAAGRLGVEPLLENLIQPSSIDVRLDRMFRVFDNSRYTHIDPAQRQDELTSLVEPGEGEPFVLHPGEFVLGSTLEVCTLPDDLAGRLEGKSSLGRLGLLTHSTAGFIDPGFSGHITLELSNVANLPITLWPGMKIGQLCLFRLSSPAEHPYGSAAAGSKYQGQRGPTPSRSYLNFPLPSDAVDAVESR\n\n### Sequence 2:\nMAATTSPMASQLKSGFTTKALVVPKGISGPALRGFPSPRRHTSFTVRAIKTEKPTYQVIQPLNGDPFIGGLETPVTSSPLIAWYLSNLPAYRTAVNPLLRGVEVGLAHGFLLVGPFVKAGPLRNTEYAGAAGSLAAAGLVVILSMCLTMYGIASFKEGEPSIAPALTLTGRKKQPDQLQSADGWAKFTGGFFFGGVSGVTWACFLMYVLDLPYYFK\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01912", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nSMRPSLKTLQEKGLIKDQIFGSHLHKVCERENSTVPWFVKQCIEAVEKRGLDVDGIYRVSGNLATIQKLRFIVNQEEKLNLDDSQWEDIHVVTGALKMFFRELPEPLFPYSFFEQFVEAIKKQDNNTRIEAVKSLVQKLPPPNRDTMKVLFGHLTKIVAKASKNLMSTQSLGIVFGPTLLRAENETGNMAIHMVYQNQIAELMLSEYSKIFGSE\n\n### Answer:", "output": "<SEQ_2D>CCCCCHHHHHHTTSCCCCSTTSCHHHHHHHHTSSSCHHHHHHHHHHHHHTTTSTTTTTSCCCHHHHHHHHHHHHTTCCCCTTSGGGCCHHHHHHHHHHHHHHSSSCSSCHHHHHHHHHHHTSSSHHHHHHHHHHHHTTSCHHHHHHHHHHHHHHHHHHHTHHHHCCCHHHHHHHHHHHHTCCTTSSSCHHHHHHHHHHHHHHHHHTHHHHHCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01913", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMHKSSNFSGLNLVKSAPICLLDIFHRCVDSETALDTSLLYLIDMFGTAVFAVSGVLLAGRLKMDPFGVMVLASVTAIGGGTIRDMALGATPVFWIKDTNYLWVIMITCALTMLLVRRPKRLAWWILPVCDAIGLAVFVGIGVEKALIYQDSALIAIIMGVITGCGGGIIRDVLAREIPMVLRSEVYATACIIGGIFHTTALAMGYSSSTALLSGVFSTLLIRLGAIRWHLSLPTFALTK\n\n### Sequence 2:\nMKRPIFIGITGGTGSGKSTIAKEIYRQFGEDCIAMIEQDSYYKDQSHLSMEDRVKTNYDHPNAFDNNLLVSHLESLLNGHSIQKPSYDFSIHNRIEDTTKVEPKEIVIVEGILILEDPRIRELLDIKIYVDTDADVRIIRRMVRDINERGRTMESVINQYLNVVKPMHNQFTEPTKKFADIIIPEGGHNKVAIDIIVAKIKEVLGKYE\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01914", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NP58-W546A on protein ATP-binding cassette sub-family B member 6.", "input": "### Wild-type:\nMVTVGNYCEAEGPVGPAWMQDGLSPCFFFTLVPSTRMALGTLALVLALPCRRRERPAGADSLSWGAGPRISPYVLQLLLATLQAALPLAGLAGRVGTARGAPLPSYLLLASVLESLAGACGLWLLVVERSQARQRLAMGIWIKFRHSPGLLLLWTVAFAAENLALVSWNSPQWWWARADLGQQVQFSLWVLRYVVSGGLFVLGLWAPGLRPQSYTLQVHEEDQDVERSQVRSAAQQSTWRDFGRKLRLLSGYLWPRGSPALQLVVLICLGLMGLERALNVLVPIFYRNIVNLLTEKAPWN\n\n### Mutant:\nMVTVGNYCEAEGPVGPAWMQDGLSPCFFFTLVPSTRMALGTLALVLALPCRRRERPAGADSLSWGAGPRISPYVLQLLLATLQAALPLAGLAGRVGTARGAPLPSYLLLASVLESLAGACGLWLLVVERSQARQRLAMGIWIKFRHSPGLLLLWTVAFAAENLALVSWNSPQWWWARADLGQQVQFSLWVLRYVVSGGLFVLGLWAPGLRPQSYTLQVHEEDQDVERSQVRSAAQQSTWRDFGRKLRLLSGYLWPRGSPALQLVVLICLGLMGLERALNVLVPIFYRNIVNLLTEKAPWN\n\n### Answer:", "output": "Loss of substrate-stimulate ATPase activity. Impairs protein expression. The mutation in ABCB6 may affect the substrate binding pocket and the translocation mechanism of the transporter.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01915", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMSKIRGNRKTRVGVVISDKMDKTVVVKVDQMVKHPIYKKYIKRRVTFKAHDEENRCNVGDKVSVVETRPLSRDKRWRVREILEKNVIL\n\n### Sequence 2:\nMAPEVIRQDFQAGEVAFRRTGYLRGRSVLTQTKHSLAGNGRHPVALRTRLGSLALGAVPTWTKLWAQSTTWQTRNHTRTGHAYPRFTRPSFPSCNRNGKRRKLRLGLPY\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01916", "category": "Mutation", "instruction": "Describe the effect of mutation P49682-A293E on protein C-X-C chemokine receptor type 3.", "input": "### Wild-type:\nMVLEVSDHQVLNDAEVAALLENFSSSYDYGENESDSCCTSPPCPQDFSLNFDRAFLPALYSLLFLLGLLGNGAVAAVLLSRRTALSSTDTFLLHLAVADTLLVLTLPLWAVDAAVQWVFGSGLCKVAGALFNINFYAGALLLACISFDRYLNIVHATQLYRRGPPARVTLTCLAVWGLCLLFALPDFIFLSAHHDERLNATHCQYNFPQVGRTALRVLQLVAGFLLPLLVMAYCYAHILAVLLVSRGQRRLRAMRLVVVVVVAFALCWTPYHLVVLVDILMDLGALARNCGRASRVDVAK\n\n### Mutant:\nMVLEVSDHQVLNDAEVAALLENFSSSYDYGENESDSCCTSPPCPQDFSLNFDRAFLPALYSLLFLLGLLGNGAVAAVLLSRRTALSSTDTFLLHLAVADTLLVLTLPLWAVDAAVQWVFGSGLCKVAGALFNINFYAGALLLACISFDRYLNIVHATQLYRRGPPARVTLTCLAVWGLCLLFALPDFIFLSAHHDERLNATHCQYNFPQVGRTALRVLQLVAGFLLPLLVMAYCYAHILAVLLVSRGQRRLRAMRLVVVVVVAFALCWTPYHLVVLVDILMDLGALARNCGRESRVDVAK\n\n### Answer:", "output": "Highly enhances binding to CXCL10 and CXCL9- and CXCL11-induced chemotaxis. Enables binding to CXCL11 and CXCL10-induced chemotaxis. The residue E293 in CXCR3 is mutated to A. This mutation is prohibited for ligand binding and activation by all three ligands (CXCL9, CXCL10, and CXCL11).", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01917", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMIKLTARQEQILNLIRDAIENTGFPPTRAEIATELGFRSANAAEEHLQALARKGAIEISPGTSRGIRLRDMGNGDERFPGRQMALPHPALMQLSLPLVGRVAAGSPILAQEHIEATYDVDRSLFSAKPDFLLKVRGMSMRDAGILDGDLLAVKKIDTAKNGQIVVARLGEEVTVKRYKKSGSLIELLPENPDFEPIRVDLAHDEFALEGLAVGLMRTWN\n\n### Sequence 2:\nMIKLTARQEQILNLIRDAIENTGFPPTRAEIATELGFRSANAAEEHLQALARKGAIEISPGTSRGIRLRDMGNGDERFPGRQMALPHPALMQLSLPLVGRVAAGSPILAQEHIEATYDVDRSLFSAKPDFLLKVRGMSMRDAGILDGDLLAVKKIDTAKNGQIVVARLGEEVTVKRYKKSGSLIELLPENPDFEPIRVDLAHDEFALEGLAVGLMRTWN\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01918", "category": "Mutation", "instruction": "Describe the effect of mutation P0AB58-C357S on protein Lipopolysaccharide assembly protein B.", "input": "### Wild-type:\nMLELLFLLLPVAAAYGWYMGRRSAQQNKQDEANRLSRDYVAGVNFLLSNQQDKAVDLFLDMLKEDTGTVEAHLTLGNLFRSRGEVDRAIRIHQTLMESASLTYEQRLLAIQQLGRDYMAAGLYDRAEDMFNQLTDETDFRIGALQQLLQIYQATSEWQKAIDVAERLVKLGKDKQRVEIAHFYCELALQHMASDDLDRAMTLLKKGAAADKNSARVSIMMGRVFMAKGEYAKAVESLQRVISQDRELVSETLEMLQTCYQQLGKTAEWAEFLQRAVEENTGADAELMLADIIEARDGSEA\n\n### Mutant:\nMLELLFLLLPVAAAYGWYMGRRSAQQNKQDEANRLSRDYVAGVNFLLSNQQDKAVDLFLDMLKEDTGTVEAHLTLGNLFRSRGEVDRAIRIHQTLMESASLTYEQRLLAIQQLGRDYMAAGLYDRAEDMFNQLTDETDFRIGALQQLLQIYQATSEWQKAIDVAERLVKLGKDKQRVEIAHFYCELALQHMASDDLDRAMTLLKKGAAADKNSARVSIMMGRVFMAKGEYAKAVESLQRVISQDRELVSETLEMLQTCYQQLGKTAEWAEFLQRAVEENTGADAELMLADIIEARDGSEA\n\n### Answer:", "output": "Lack of activity; when associated with S-360; S-371 and S-374. The mutation in the YciM protein alters its function and is required for envelope integrity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01919", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nDTELLSIKILNAGANGDKVVEGTIDEAKKTINFPRLDVETDFSALSIEAELSEGAALQSEVMDYSMDAETNEKTQVLRIINHNRYKDYLMKVRKRVPVFGADFEKPTVYNFSGDNIYSDFATNYTRCASYDGEHVLVVSRPTTPNFHTPHLLKVSDLKRGEIKPIMLDVTGVKGGTYDYNMGALINGHVYLSSLSGGKVSPFKIYYWETPTSNPEVIANINVGNIPGAGNRHGDNASYNIDENGNGFIFFGDNAATEFLKVPISGHKTVDIGNIKVLPSKSDATMVTNVYRVGDTDQYLWSGIRVPVTLVDESLGEKYKSKIAGEAVAPKVVTFNEERYLLVCTAGQGAASKASIALEVYDLTKGETIEDALKKFDEGENHNPIYQFKLGGSGNGNALAQTDYYIEKDENGKDAKLCLFASRTQSGFVICEFPIKQEEMD\n\n### Answer:", "output": "<SEQ_3Di>DWDWPWKWFQQFDLRGGDIWTWDADVVQQEIETWATEPNTQQQWTFMDTDIDPFKDWPDRTDGFDDPPVDFKDWDWTWIDGPPDIDTGIYMYGHFDDAFDFDQVDKDKLDCFDVNPDVLLQALQWFEWEDAPQWIWTQGCADPPRGDHIFIAGVVCSVVVHGDTLDAACVVLDWAPASWHYWYQDLQKIKTWHKHQFQATWTWIWIDPHRHRHIDTLDGDRNVPDPPDDTIWRLDWEDEADNQLWDWIWIAGQQQQWIFTWTTGNNRHTPPVPTDTAGGDNQQHNAKYWYDADPDQWIWIHHQRAATFIAGPRRHTDGGAPHHLQWHAWYWDAARQWIKTWTARFQADQPRPRQRKTFIATQRPDDDSVRSRVVCRPDPDNGTPDIDRLDDHHDNSWHKHKDWYFDADPVRHTFKIWIKMATRSTHMMIIMTGIDDTPSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01920", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMSETPAQCSIKQERISYTPPESPVPSYASSTPLHVPVPRALRMEEDSIRLPAHLRLQPMFWSRDDVAQWLKWAENEFSLRPIDSNTFEMNGKALLLLTKEDFRYRSPHSGDVLYELLQHILKQRKPRILFSPFFHPGNSIHTQQEVILHQNHEEDNGVQRTSRPSAENVHQNPPTIELLHRSRSPITTNHRPSPDPEQRPLRSPLDNMIRRLSPAERAQGPRLHQENNHQEPYPLSVSPMENNHCPPSSEPHPRPSSPRQEGTRVIQLMPSPIMHPLILNPRHSVDFKQPRLSEDGLHREGKPINLSHREDLAYMNHIMVSVSPPEEHAVPIGRIADCRLLWDYVYQLLSDSRYENFIRWEDKESKIFRIVDPNGLARLWGNHKNRTNMTYEKMSRALRHYYKLNIIRKEPGQRLLFRFMKTPDEIMSGRTDRLEHLESQELDEQIYQEDEC\n\n### Sequence 2:\nMSETPAQCSIKQERISYTPPESPVPSYASSTPLHVPVPRALRMEEDSIRLPAHLRLQPIYWSRDDVAQWLKWAENEFSLRPIDSNTFEMNGKALLLLTKEDFRYRSPHSGDVLYELLQHILKQRKPRILFSPFFHPGNSIHTQPEVILHQNHEEDNCVQRTPRPSVDNVHHNPPTIELLHRSRSPITTNHRPSPDPEQRPLRSPLDNMIRRLSPAERAQGPRPHQENNHQESYPLSVSPMENNHCPASSESHPKPSSPRQESTRVIQLMPSPIMHPLILNPRHSVDFKQSRLSEDGLHREGKPINLSHREDLAYMNHIMVSVSPPEEHAMPIGRIADCRLLWDYVYQLLSDSRYENFIRWEDKESKIFRIVDPNGLARLWGNHKNRTNMTYEKMSRALRHYYKLNIIRKEPGQRLLFRFMKTPDEIMSGRTDRLEHLESQELDEQIYQEDEC\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01921", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nIQKTPQIQVYSRHPPENGKPNILNCYVTQFHPPHIEIQMLKNGKKIPKVEMSDMSFSKDWSFYILAHTEFTPTETDTYACRVKHDSMAEPKTVYWDRDM\n\n### Answer:", "output": "<SEQ_2D>CCBCCEEEEEESSCCCTTSCEEEEEEEEEEBSSCCEEEEEETTEECSCCEECCCEECTTSCEEEEEEEEECCCSSCCEEEEEECTTSSSCEEEECCSCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01922", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "COCC(O)CO", "output": "The molecule contains the following functional groups: hydroxyl, ether.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "COCC(O)CO", "compound_name": null, "pubchem_cid": null, "MW": 106.12, "LogP": -1.01, "TPSA": 49.69, "QED": 0.48, "functional_groups": ["hydroxyl", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["hydroxyl", "ether"]}}
{"id": "eval_01923", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nPPBP PF4 ITGA2B GP1BA\n</cell>", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as GP1BA, ITGA2B, PF4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "prostate", "disease": "breast carcinoma", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["PPBP", "PF4", "ITGA2B", "GP1BA"]}}
{"id": "eval_01924", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMKCPYCAYGESKVVDSRSTEDGSSIRRRRECLKCNRRYTTYEKIETTPILVIKKNMSREYFDRNKIVNGLMKACQKRPVSRKQIEQIADEVERHISNEMLTEVNTDKIGQIIMKNLKKIDEVSYVRFASVYRQFKDINTFMKEIKNLMDKN\n\n### Sequence 2:\nMSLIVYFSSRSGNTHRFVERLGVRSSRIPLEASGALQVREPFVLVTPTYGGGSTKGAVPNPVIRFLNDADNRALIRGVIAAGNSNFGEAFCIAGNIISAKCGVPYLYRFELLGTAEDVGNVRNGMEQFWTRQTQA\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01925", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nATPG_ARTPT\n\n### Answer:", "output": "Belongs to the ATPase gamma chain family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01926", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nIKBL_ASFB7\n\n### Answer:", "output": "Belongs to the asfivirus A238L family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01927", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nPHMERASLIQKAKLAEQAERYEDMAAFMKGAVEKGEELSCEERNLLSVAYKNVVGGQRAAWRVLSSIEQKSNEEGEKGPEVREYREKVETELQGVCDTVLGLLDSHLIKEAGDAESRVFYLKMKGDYYRYLAEVATGDDKKRIIDSARSAYQEAMDISAAAMPPTNPIRLGLALNFSVFHYEIANSPEEAISLAKTTFDEAMADLHTLSEDSYKDSTLIMQLLRDNLTLWT\n\n### Answer:", "output": "<SEQ_3Di>DDDDLVVLVVVLVVCVVVVVLVSNLVSLLVNQVVQDAADPVSLVSNVVSLCVVLVVLLVVLVVLVVVVVCCVDPVDDDCVSVVVSVVSLVVNLVSLVSLLVSLVPHQCVPDDDLVSNLSSLLSQLVSLVSNLVPDDPPVNVVSLVRSVVSLVVSQVSCVVPHQLLPLSNLVSLLVVLVCVVPRVVNLVVSLVSLVVSLVRNVVPLVVDDPVSNVSSVVSNVSSVVVNVVSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01928", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "CC(C)C(C)C", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains no standard functional groups detected. Its descriptor profile: MW=86.18, LogP=2.3, TPSA=0.0, HBD=0, HBA=0, QED=0.46.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "CC(C)C(C)C", "compound_name": null, "pubchem_cid": null, "MW": 86.18, "LogP": 2.3, "TPSA": 0.0, "QED": 0.46, "functional_groups": [], "moleculenet_labels": {"NR-AR": null, "NR-AR-LBD": null, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": null, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01929", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMAYYSTVLKRKVRKMLAVHFIAIGGIGMSGLARILQSKGYRVSGSDLKETELTKKLRAEGITVFIGHREENLASDVSLVVVSTAVSQDNPELLKAKRLGIPVMHRGELLARLMQEKKGIAVAGTHGKTTTSSMIAYVLEKEGFDPVIAVGGEIVDLGYNAKAGQGEYMVAEADESDGSFLKLLPYAAVITNIEADHLDYYQSFEEIKKAFKKFADNIRPEGFGVFCWDNLQVREMLKGYKKRKFTYGFSPGSDFMLRDYREEQNQLVANIYYKNTLEGELRLKVPGKHNILNAAAATAVLRNIGLSFKAISERLLEFNGAKRRFQILGERNGALIVDDYAHHPTEVEATLRAAKLYKDRDVLVVFQPHRYTRTHFFYKEFARVLVDAEKVVLTGIYSAGEKPIPGVSGEMIAEEMKKLGKNPLYLESLDEVYNYLEQNLKPGLLVLLLGAGNINQVGYKLLGKA\n\n### Sequence 2:\nMLKSLSKLTPSIRGVVSINRSFCTKKFLPTNRSVSESDPEIYDLMMKEKQRQFTGLELIASENFTSRAVMESIGSCFTNKYAEGLPGARYYGGNEVVDQLENLCIKRALETFNLNPEEWGVNVQPYSGSTANFAAFTGLLKPHDRIMGLDLPSGGHLTHGYQTDKKKISATSIFFESMPYQVNETGYVDYNKMEANAALFRPKLLIAGASAYPREWDYERMRKIADKHGAYLLCDMAHISGMVAGKQAISPFLFCDVVTTTTHKTLRGPRAGLIFFRKTKRRDAKGNIIDDDLENRINFAVFPSCQGGPHENTIAGIAVALKEASSPDFQEYTKQVRRNSQTMGEELKKRGYSLVTEGTDNHLVLWDLRPQGITGSKIEKACDEAHITVNKNAVYGDTNAIAPGGVRLGAPALTSRGLKEQDFVKVVDFLDRVVKISLDIQSKVGKKMPDFQRAIADNQDLKQIRQEVKEFSTKFGMPGEL\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01930", "category": "Mol", "instruction": "Provide a drug-likeness assessment for this compound.", "input": "SMILES: Oc1nc2[nH]c(-c3ccc(Br)cc3)cc2c2ccccc12\nDescriptors: MW=339.19, LogP=4.85, TPSA=48.91, HBD=2, HBA=2", "output": "Descriptors: MW=339.19, LogP=4.85, TPSA=48.91, HBD=2, HBA=2, QED=0.52, Lipinski violations=0. Overall the molecule is broadly compatible with common oral drug-like descriptors, and a moderate QED score suggests mixed drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "Oc1nc2[nH]c(-c3ccc(Br)cc3)cc2c2ccccc12", "compound_name": null, "pubchem_cid": null, "MW": 339.19, "LogP": 4.85, "TPSA": 48.91, "QED": 0.52, "functional_groups": ["aromatic ring", "hydroxyl", "halide", "phenol", "pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}
{"id": "eval_01931", "category": "Mutation", "instruction": "Describe the effect of mutation P63092-Q170A on protein Guanine nucleotide-binding protein G.", "input": "### Wild-type:\nMGCLGNSKTEDQRNEEKAQREANKKIEKQLQKDKQVYRATHRLLLLGAGESGKSTIVKQMRILHVNGFNGEGGEEDPQAARSNSDGEKATKVQDIKNNLKEAIETIVAAMSNLVPPVELANPENQFRVDYILSVMNVPDFDFPPEFYEHAKALWEDEGVRACYERSNEYQLIDCAQYFLDKIDVIKQADYVPSDQDLLRCRVLTSGIFETKFQVDKVNFHMFDVGGQRDERRKWIQCFNDVTAIIFVVASSSYNMVIREDNQTNRLQEALNLFKSIWNNRWLRTISVILFLNKQDLLAEK\n\n### Mutant:\nMGCLGNSKTEDQRNEEKAQREANKKIEKQLQKDKQVYRATHRLLLLGAGESGKSTIVKQMRILHVNGFNGEGGEEDPQAARSNSDGEKATKVQDIKNNLKEAIETIVAAMSNLVPPVELANPENQFRVDYILSVMNVPDFDFPPEFYEHAKALWEDEGVRACYERSNEYALIDCAQYFLDKIDVIKQADYVPSDQDLLRCRVLTSGIFETKFQVDKVNFHMFDVGGQRDERRKWIQCFNDVTAIIFVVASSSYNMVIREDNQTNRLQEALNLFKSIWNNRWLRTISVILFLNKQDLLAEK\n\n### Answer:", "output": "Increases GDP release but does not affect receptor-mediated activation. decreased activation by AlF4-, increased thermolability, increased rate of GDP release, not impaired receptor-mediated activation", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01932", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAFSKLLEQAGGVGLFQTLQVLTFILPCLMIPSQMLLENFSAAIPGHRCWTHMLDNGSAVSTNMTPKALLTISIPPGPNQGPHQCRRFRQPQWQLLDPNATATSWSEADTEPCVDGWVYDRSVFTSTIVAKWDLVCSSQGLKPLSQSIFMSGILVGSFIWGLLSYRFGRKPMLSWCCLQLAVAGTSTIFAPTFVIYCGLRFVAAFGMAGIFLSSLTLMVEWTTTSRRAVTMTVVGCAFSAGQAALGGLAFALRDWRTLQLAASVPFFAISLISWWLPESARWLIIKGKPDQALQELRKVARINGHKEAKNLTIEVLMSSVKEEVASAKEPRSVLDLFCVPVLRWRSCAMLVVNFSLLISYYGLVFDLQSLGRDIFLLQALFGAVDFLGRATTALLLSFLGRRTIQAGSQAMAGLAILANMLVPQDLQTLRVVFAVLGKGCFGISLTCLTIYKAELFPTPVRMTADGILHTVGRLGAMMGPLILMSRQALPLLPPLLYGVISIASSLVVLFFLPETQGLPLPDTIQDLESQKSTAAQGNRQEAVTVESTSL\n\n### Sequence 2:\nMAFSELLDLVGGLGRFQVLQTMALMVSIMWLCTQSMLENFSAAVPSHRCWAPLLDNSTAQASILGSLSPEALLAISIPPGPNQRPHQCRRFRQPQWQLLDPNATATSWSEADTEPCVDGWVYDRSIFTSTIVAKWNLVCDSHALKPMAQSIYLAGILVGAAACGPASDRFGRRLVLTWSYLQMAVMGTAAAFAPAFPVYCLFRFLLAFAVAGVMMNTGTLLMEWTAARARPLVMTLNSLGFSFGHGLTAAVAYGVRDWTLLQLVVSVPFFLCFLYSWWLAESARWLLTTGRLDWGLQELWRVAAINGKGAVQDTLTPEVLLSAMREELSMGQPPASLGTLLRMPGLRFRTCISTLCWFAFGFTFFGLALDLQALGSNIFLLQMFIGVVDIPAKMGALLLLSHLGRRPTLAASLLLAGLCILANTLVPHEMGALRSALAVLGLGGVGAAFTCITIYSSELFPTVLRMTAVGLGQMAARGGAILGPLVRLLGVHGPWLPLLVYGTVPVLSGLAAL-LLPETQSLPLPDTIQDVQNQAVKKATHGTLGNSVLKST\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01933", "category": "Mol", "instruction": "List the chemically relevant functional groups of this molecule.", "input": "CCCC(O)CC", "output": "The molecule contains the following functional groups: hydroxyl.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "ESOL", "canonical_smiles": "CCCC(O)CC", "compound_name": null, "pubchem_cid": null, "MW": 102.18, "LogP": 1.56, "TPSA": 20.23, "QED": 0.57, "functional_groups": ["hydroxyl"], "moleculenet_labels": {"label": -0.8}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["hydroxyl"]}}
{"id": "eval_01934", "category": "Mol", "instruction": "Which functional groups are present in this SMILES?", "input": "O=C(O)Cc1sc(-c2ccccc2)nc1-c1ccc(Cl)cc1", "output": "The molecule contains the following functional groups: aromatic ring, hydroxyl, halide.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "functional_group", "source": "Tox21", "canonical_smiles": "O=C(O)Cc1sc(-c2ccccc2)nc1-c1ccc(Cl)cc1", "compound_name": null, "pubchem_cid": null, "MW": 329.81, "LogP": 4.76, "TPSA": 50.19, "QED": 0.75, "functional_groups": ["aromatic ring", "hydroxyl", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": null, "NR-AhR": 1.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 1.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 1.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_smarts", "fgs": ["aromatic ring", "hydroxyl", "halide"]}}
{"id": "eval_01935", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMAVEEKNSPTGAAMTNTGILVPSSQQSLPEWWTKTQKFFSRENTITPTFGYFRLLFGTQPGKTDIALIVIGTIAGIGAGIPFPLLGILFGELVDDLNSSTCSTTQAPPGGYQAAITTKVLQVIYVSILNFVCMYIHTGCWSMVGERLVRRLRTKYFHSLLRQEIAFTDTLPSGDVTSRLVSDIEVIQAGTSEKVGLFIGTISYFVAAYIVAFLKVATIAAMLMSVVPIYFLMAFGGGHYIKKYSGRISTHINAATSIVSSSLSHMSIVHAFNANARLEALFAQHLVSARMDALKKAITHSIQFGMLYFVAYASNALAFWQGSRMIADLAEGKPSKVSVGAVYTVIFVLLDASFVLSQMAPFMHIFASAASAGDRLMTTIKRQSAIDGTSSEGDSTISLASEEIELQDVTFNYPARPEVPVLQGVSFKIPPNKHTAIVGTSGSGKSTVVALLERLYDPITGCVRVGNRDLKEINVRHLRGSIGLVQQEPNLLDRSILENIAHGLVSSSQEKHKHLLPTLLGPSLSELTEKIRQGASEDEAVTEQGDVVREIVNLTRHAATLSNAIDFINALPDGLATRVGSSGAELSGGQKQRIALARALIRDPPVLLLDEATAALDSTSERLIQAALNKVSENVTTVSIAHRLATAKDADNIIVMQKGRVMEQGTHMDLVARDGVYAGMVRLQNIGKFSSSSSIMTESTQVDANIDRSLTTDTLLNKEEKLSLEQGVLDEKEKPAQLYMPEEADSLPTEPENEKEKPKQTLWATMKGSFPLIRPNILLISLGLITSIMIGVSYTGEAVIFGHTVGSLSVCRGGPSIRSSGMLFGLLFFILAIVKFAAVIVNGAAFGWAAEKTLYRTRVLSLRSLLRQPLEWHNADGRTPGLLVALVTSDASALSSLTGTTIGVLFSTVANLFAGVILSQCHRMEDSRSPSGYLTRIKHQKAYAKATAITVESVDNIKSIAAFSLEQEAYSVFNRSLKAPYKSNMKSVLHGNFWLSLAYSISTLVYALAYWWGSQQILAGMYTQVQFFIVLPALLFSTQSCGQMFALVPDISKARIAASNIVDLLSIKHEGDEEYDKTGSKASAKHTDPRFNMLEDKPRDVEAQLITTTPSSFPTKGMGVQFRNVHFRYPSRPNQPALDDLSINISPGQFCALVGPSGSGKSTTFALLEKFYNPASGSIIIDGVDITKQSGAAFRDTIALVPQENVMFEGTVAFNIGLGARPDVEATQEEIEEACRLANIHDTIAALPDGYNTVCSQDGKQFSGGQRQRLSIARALVRKPRLLLLDESTSALDVESEKHVQDALAKVARKTTIVAIAHRLNTIHRADRIFMIERGRCVDQGTHAELVERCESYRANVIHQSLDA\n\n### Sequence 2:\nMAVEEKNSPTGAAMTNTGILVPSSQQSLPEWWTKTQKFFSRENTITPTFGYFRLLFGTQPGKTDIALIVIGTIAGIGAGIPFPLLGILFGELVDDLNSSTCSTTQAPPGGYQAAITTKVLQVIYVSILNFVCMYIHTGCWSMVGERLVRRLRTKYFHSLLRQEIAFTDTLPSGDVTSRLVSDIEVIQAGTSEKVGLFIGTISYFVAAYIVAFLKVATIAAMLMSVVPIYFLMAFGGGHYIKKYSGRISTHINAATSIVSSSLSHMSIVHAFNANARLEALFAQHLVSARMDALKKAITHSIQFGMLYFVAYASNALAFWQGSRMIADLAEGKPSKVSVGAVYTVIFVLLDASFVLSQMAPFMHIFASAASAGDRLMTTIKRQSAIDGTSSEGDSTISLASEEIELQDVTFNYPARPEVPVLQGVSFKIPPNKHTAIVGTSGSGKSTVVALLERLYDPITGCVRVGNRDLKEINVRHLRGSIGLVQQEPNLLDRSILENIAHGLVSSSQEKHKHLLPTLLGPSLSELTEKIRQGASEDEAVTEQGDVVREIVNLTRHAATLSNAIDFINALPDGLATRVGSSGAELSGGQKQRIALARALIRDPPVLLLDEATAALDSTSERLIQAALNKVSENVTTVSIAHRLATAKDADNIIVMQKGRVMEQGTHMDLVARDGVYAGMVRLQNIGKFSSSSSIMTESTQVDANIDRSLTTDTLLNKEEKLSLEQGVLDEKEKPAQLYMPEEADSLPTEPENEKEKPKQTLWATMKGSFPLIRPNILLISLGLITSIMIGVSYTGEAVIFGHTVGSLSVCRGGPSIRSSGMLFGLLFFILAIVKFAAVIVNGAAFGWAAEKTLYRTRVLSLRSLLRQPLEWHNADGRTPGLLVALVTSDASALSSLTGTTIGVLFSTVANLFAGVILSQCHRMEDSRSPSGYLTRIKHQKAYAKATAITVESVDNIKSIAAFSLEQEAYSVFNRSLKAPYKSNMKSVLHGNFWLSLAYSISTLVYALAYWWGSQQILAGMYTQVQFFIVLPALLFSTQSCGQMFALVPDISKARIAASNIVDLLSIKHEGDEEYDKTGSKASAKHTDPRFNMLEDKPRDVEAQLITTTPSSFPTKGMGVQFRNVHFRYPSRPNQPALDDLSINISPGQFCALVGPSGSGKSTTFALLEKFYNPASGSIIIDGVDITKQSGAAFRDTIALVPQENVMFEGTVAFNIGLGARPDVEATQEEIEEACRLANIHDTIAALPDGYNTVCSQDGKQFSGGQRQRLSIARALVRKPRLLLLDESTSALDVESEKHVQDALAKVARKTTIVAIAHRLNTIHRADRIFMIERGRCVDQGTHAELVERCESYRANVIHQSLDA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01936", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nTITIN_MOUSE\n\n### Answer:", "output": "Belongs to the protein kinase superfamily. CAMK Ser/Thr protein kinase family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01937", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nATPF2_SYNP1\n\n### Answer:", "output": "Component of the F(0) channel, it forms part of the peripheral stalk, linking F(1) to F(0).  The b'-subunit is a diverged and duplicated form of b found in plants and photosynthetic bacteria.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01938", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nRPO5_THEGJ\n\n### Answer:", "output": "DNA-directed RNA polymerase subunit Rpo5, DNA-directed RNA polymerase subunit H.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01939", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTWLSDFPQAWAETGGMGLAVRQAPLIIPLKATSTPVSIKQYPMSQEARLGIKPHIQRLLDQGILVPCQSPWNTPLLPVKKPGTNDYRPVQDLREVNKRVEDIHPTVPNPYNLLSGLPPSHQWYTVLDLKDAFFCLRLHPTSQPLFAFEWRDPEMGISGQLTWTRLPQGFKNSPTLFNEALHRDLADFRIQHPDLILLQYVDDLLLAATSELDCQQGTRALLQTLGNLGYRASAKKAQICQKQVKYLGYLLKEGQRWLTEARKETVMGQPTPKTPRQLREFLGKAGFCRLFIPGFAEMAAPLYPLTKPGTLFNWGPDQQKAYQEIKQALLTAPALGLPDLTKPFELFVDEKQGYAKGVLTQKLGPWRRPVAYLSKKLDPVAAGWPPCLRMVAAIAVLTKDAGKLTMGQPLVILAPHAVEALVKQPPMTHYQALLLDTDRVQFGPVVALNPAT\n\n### Answer:", "output": "<SEQ_3Di>DVCVLQLLAACVRNNDWAQVPDDADEFAWDPPQDQDEDAADDDDPVLLVLCVVVVVVCVVVVFWDFDAASHAFYWFQCADPPGGNGHTHTPCPSVLVGTDDDDADADDPLELLLPDALQQQWKKKWFFPPLQQHHFHDPVCQSNYWHWDDDVVVPDTHIITGRTGGPDNPCSLRVSCVVVVVLCVVVCVVPVQWDWRDHSGMIMITHRDLVRQVVSVSVVSVSSSVSTTHTHPVDIDGRDSWDADPQFTHHSQWGHHDPVLLCVLLVDDDDQDLVSLVVNVVSLVVVQFQQWLLCVLQVVVPQCNPPPDDVPCDPSNVSSSVSSNVRSVPGDTAGRHHLVDAKAKEWDDDQQKIWIFIWDDDPPDIGTHTIDIDGHDPVLNPPDPLVNRLSSCVVRLVVVCSNSVQAEYAYEAQDDCLVCCVPPCDVVSNVSAVPPRRYDYDDHHHDDPSD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01940", "category": "Mutation", "instruction": "Describe the effect of mutation P82610-Y660F on protein 5-methyltetrahydropteroyltriglutamate--homocysteine methyltransferase.", "input": "### Wild-type:\nMVQSSVLGFPRIGGQRELKKITEAYWSGKATVEELLAKGKELREHNWKLQQKAGVDIIPSNDFSYYDQVLDLSLLFNAIPERYTKFDLAPIDVLFAMGRGLQKKATETQAAVDVTALEMVKWFDSNYHYVRPTFSHSTEFKLNTAAGIKPVDEFNEAKALGVQTRPVILGPVSYLYLGKADKDSLDLEPISLLPKILPVYKELLQKLKEAGAEQVQIDEPVLVLDLPEAVQSKFKEAYDALVGADVPELILTTYFGDVRPNLKAIENLPVAGFHFDFVRVPEQLDEVASILKDGQTLSAG\n\n### Mutant:\nMVQSSVLGFPRIGGQRELKKITEAYWSGKATVEELLAKGKELREHNWKLQQKAGVDIIPSNDFSYYDQVLDLSLLFNAIPERYTKFDLAPIDVLFAMGRGLQKKATETQAAVDVTALEMVKWFDSNYHYVRPTFSHSTEFKLNTAAGIKPVDEFNEAKALGVQTRPVILGPVSYLYLGKADKDSLDLEPISLLPKILPVYKELLQKLKEAGAEQVQIDEPVLVLDLPEAVQSKFKEAYDALVGADVPELILTTYFGDVRPNLKAIENLPVAGFHFDFVRVPEQLDEVASILKDGQTLSAG\n\n### Answer:", "output": "No effect on catalytic activity. The document does not mention any functional changes caused by the mutation.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01941", "category": "Protein", "instruction": "Judge if the following two protein sequences share an evolutionary origin.", "input": "### Sequence 1:\nMNLIAILEQEEIARLTGGNAVTEFAPGDTVVVSVNVVEGTRKRVQAFEGVVIAKRNRGLNSSFIVRKISSGEAVERTFQLYSPQIAGIEVKRRGDVRRAKLYYLRSRSGKSARIKEKLVLKKAKSA\n\n### Sequence 2:\nMRTLWIVALLLVGVEGSVLELGKMILQETGKNAITSYGSYGCNCGWGHRGQPKDATDRCCFVHKCCYKKLTDCNHKTDRYSYSWKNKAIICEEKNPCLKEMCECDKAVAICLRENLDTYNKKYKAYFKLKCKKPDTC\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01942", "category": "Cell", "instruction": "Suggest a plausible disease or physiological state consistent with this cell profile.", "input": "APOB, HNF4A, ALB", "output": "The expression pattern is consistent with hepatocyte in a context associated with lung adenocarcinoma, if the dataset label supports it.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_disease", "source": "panglao_cellmarker_panels", "cell_type": "hepatocyte", "tissue": "kidney", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_disease_state", "top_genes": ["APOB", "HNF4A", "ALB"]}}
{"id": "eval_01943", "category": "Cell", "instruction": "Describe the cell state from this top-gene list.", "input": "<cell>\nS100A4 FAP ACTA2 COL1A1 PDGFRB\n</cell>", "output": "This cell is most consistent with cancer-associated fibroblast, a cancer-associated fibroblast (CAF). Key markers such as ACTA2, PDGFRB, S100A4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "bone marrow", "disease": "Alzheimer disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["S100A4", "FAP", "ACTA2", "COL1A1", "PDGFRB"]}}
{"id": "eval_01944", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTYPIFTVRWLAIHGLAVPTVFFLGSISAMQFIQR\n\n### Answer:", "output": "<SEQ_3Di>DPDDDDPVVVVVCVPPVVVVVVVVVVVVVVPDDD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01945", "category": "Cell", "instruction": "Describe the likely cell identity from the ranked expressed genes.", "input": "<cell>\nCDKN1C MS4A7 LST1 FCGR3A\n</cell>", "output": "This cell is most consistent with CD16+ monocyte, a non-classical CD16+ monocyte. Key markers such as FCGR3A, MS4A7, CDKN1C support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "top_to_identity", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "prostate", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_from_top_genes", "top_genes": ["CDKN1C", "MS4A7", "LST1", "FCGR3A"]}}
{"id": "eval_01946", "category": "Mol", "instruction": "Based on the evidence, describe the ligand-target relationship.", "input": "Target: EGFR\nLigand SMILES: Cc1c(CN2CCN(C(=O)[C@H](C)O)CC2)sc2c(N3CCOCC3)nc(-c3cnc(N)nc3)nc12\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and epidermal growth factor receptor (EGFR) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Lipophilicity", "canonical_smiles": "Cc1c(CN2CCN(C(=O)[C@H](C)O)CC2)sc2c(N3CCOCC3)nc(-c3cnc(N)nc3)nc12", "compound_name": null, "pubchem_cid": null, "MW": 498.61, "LogP": 0.9, "TPSA": 133.83, "QED": 0.52, "functional_groups": ["amide", "amine", "hydroxyl", "ether"], "moleculenet_labels": {"label": 1.9}, "split": "train", "generation_method": "template_protein_ligand", "target": "EGFR", "has_evidence": false}}
{"id": "eval_01947", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nRS11_ECO7I\n\n### Answer:", "output": "Belongs to the universal ribosomal protein uS11 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01948", "category": "Mol", "instruction": "Analyze the molecular properties of the given SMILES.", "input": "O=C(Nc1ccc(Cl)cc1)Nc1ccc(Cl)c(C(F)(F)F)c1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains aromatic ring, amide, halide. Its descriptor profile: MW=349.14, LogP=5.66, TPSA=41.13, HBD=2, HBA=1, QED=0.72.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=C(Nc1ccc(Cl)cc1)Nc1ccc(Cl)c(C(F)(F)F)c1", "compound_name": null, "pubchem_cid": null, "MW": 349.14, "LogP": 5.66, "TPSA": 41.13, "QED": 0.72, "functional_groups": ["aromatic ring", "amide", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": null, "NR-ER": null, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": null, "SR-ARE": null, "SR-ATAD5": 0.0, "SR-HSE": null, "SR-MMP": 1.0, "SR-p53": 1.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01949", "category": "Mutation", "instruction": "Describe the effect of mutation P20292-V20A on protein Arachidonate 5-lipoxygenase-activating protein.", "input": "### Wild-type:\nMDQETVGNVVLLAIVTLISVVQNGFFAHKVEHESRTQNGRSFQRTGTLAFERVYTANQNCVDAYPTFLAVLWSAGLLCSQVPAAFAGLMYLFVRQKYFVGYLGERTQSTPGYIFGKRIILFLFLMSVAGIFNYYLIFFFGSDFENYIKTISTTISPLLLIP\n\n### Mutant:\nMDQETVGNVVLLAIVTLISAVQNGFFAHKVEHESRTQNGRSFQRTGTLAFERVYTANQNCVDAYPTFLAVLWSAGLLCSQVPAAFAGLMYLFVRQKYFVGYLGERTQSTPGYIFGKRIILFLFLMSVAGIFNYYLIFFFGSDFENYIKTISTTISPLLLIP\n\n### Answer:", "output": "Increased affinity for the inhibitor MK-591.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01950", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRIMO_PELTS\n\n### Answer:", "output": "Catalyzes the methylthiolation of an aspartic acid residue of ribosomal protein S12.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01951", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: ALK\nLigand SMILES: COCCc1ccc(OCC(O)CNC(C)C)cc1.COCCc1ccc(OCC(O)CNC(C)C)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "Tox21", "canonical_smiles": "COCCc1ccc(OCC(O)CNC(C)C)cc1.COCCc1ccc(OCC(O)CNC(C)C)cc1", "compound_name": null, "pubchem_cid": null, "MW": 534.74, "LogP": 3.23, "TPSA": 101.44, "QED": 0.23, "functional_groups": ["aromatic ring", "amine", "hydroxyl", "ether"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01952", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "IAPP, NKX6-1, MAFA, INS", "output": "This cell is most consistent with pancreatic beta cell, a pancreatic beta cell secreting insulin. Key markers such as NKX6-1, IAPP, MAFA support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "pancreatic beta cell", "tissue": "tumor", "disease": "lung adenocarcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["IAPP", "NKX6-1", "MAFA", "INS"]}}
{"id": "eval_01953", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nDIP2B_MOUSE\n\n### Answer:", "output": "Disco-interacting protein 2 homolog B, DIP2 homolog B.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01954", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CNC1=Nc2ncccc2C(c2ccns2)=NC1c1cccs1", "output": "This molecule is medium-sized and moderately polar, with lipophilic character. It contains amine, pyridine. Its descriptor profile: MW=339.45, LogP=3.44, TPSA=62.53, HBD=1, HBA=7, QED=0.78.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "CNC1=Nc2ncccc2C(c2ccns2)=NC1c1cccs1", "compound_name": null, "pubchem_cid": null, "MW": 339.45, "LogP": 3.44, "TPSA": 62.53, "QED": 0.78, "functional_groups": ["amine", "pyridine"], "moleculenet_labels": {"label": 2.15}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01955", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nAPBC_RICBR\n\n### Answer:", "output": "Binds and transfers iron-sulfur (Fe-S) clusters to target apoproteins.  Can hydrolyze ATP.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01956", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMATVTSSAAVAIPSFAGLKASSTTRAATVKVAVATPRMSIKASLKDVGVVVAATAAAGILAGNAMAAEVLLGSSDGGLAFVPSDLSIASGEKITFKNNAGFPHNVVFDEDEVPAGVDVTKISMPEEDLLNAPGEEYSVTLTEKGTYKFYCAPHAGAGMVGKVTVN\n\n### Sequence 2:\nMESSSSTFLTTTSLDKKKPSPVSRKSPKQKKKTTSTNKPIKVRYISNPMRVQTCASKFRELVQELTGQDAVDLQPEPIYSPSSDDHNLSPPAENLAPRVLHQEPFGERDSDCYEPLNAEDMFLPDQMSAGFSGFFSNGFYNVNDFGSIDSM\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01957", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nPHO91_YEAST\n\n### Answer:", "output": "Vacuolar phosphate transporter that probably exports phosphate from the vacuolar lumen to the cytosol.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01958", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMFCTFCLKDRNYYMFKLFKDYWPTCTTECQICLEKIDDNGGIVAMPDTGMLNLEKMFHEQCIQRWRREHTRDPFNRVIKYYFNFPPKTLEECNVMLRETKGSIGDHEIDRVYKRVYQRVTQEDALDIELDFRHFFKMQS\n\n### Sequence 2:\nMFCTFCLKDRNYYMFKLFKDYWPTCTTECQICLEKIDDNGGIVAMPDTGMLNLEKMFHEQCIQRWRREHTRDPFNRVIKYYFNFPPKTLEECNVMLRETKGSIGDHEIDRVYKRVYQRVTQEDALDIELDFRHFFKMQS\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01959", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nRPOC_WEIHE\n\n### Answer:", "output": "DNA-dependent RNA polymerase catalyzes the transcription of DNA into RNA using the four ribonucleoside triphosphates as substrates.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01960", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMFLGTHTPRLDDKGRLILPAKFRDELAGGVVITKGQERCLYVFPTPEFQHIADQLRAQPMTHKAARAYSRVFFASAHDEVPDKQGRVTIPGHLREYAALDRDLVVIGAHTRVEIWDRVAWESYLAESEDEFADIEEGVLPGL\n\n### Sequence 2:\nMWSDPIADMLTRIRNANMVFKEYTDIPASNLKKKICEILKREGFIADYKYIEDGKQGILRVYLKYKGGRKNRERVIHGIVRVSHSGRRIYVDKDHIPKVKNGLGIAILTTSKGVLTDKEARQLGVGGEVIAYVW\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01961", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "COC(=O)CN1CCSC1=NC#N", "output": "This molecule is small and moderately polar, with hydrophilic character. It contains ester, amine, ether, nitrile. Its descriptor profile: MW=199.24, LogP=0.05, TPSA=65.69, HBD=0, HBA=5, QED=0.46.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "COC(=O)CN1CCSC1=NC#N", "compound_name": null, "pubchem_cid": null, "MW": 199.24, "LogP": 0.05, "TPSA": 65.69, "QED": 0.46, "functional_groups": ["ester", "amine", "ether", "nitrile"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01962", "category": "Mutation", "instruction": "Describe the effect of mutation Q96EU7-K152E on protein C1GALT1-specific chaperone 1.", "input": "### Wild-type:\nMLSESSSFLKGVMLGSIFCALITMLGHIRIGHGNRMHHHEHHHLQAPNKEDILKISEDERMELSKSFRVYCIILVKPKDVSLWAAVKETWTKHCDKAEFFSSENVKVFESINMDTNDMWLMMRKAYKYAFDKYRDQYNWFFLARPTTFAIIKNLKYFLLKKDPSQPFYLGHTIKSGDLEYVGMEGGIVLSVESMKRLNSLLNIPEKCPEQGGMIWKISEDKQLAVCLKYAGVFAENAEDADGKDVFNTKSVGLSIKEAMTYHPNQVVEGCCSDMAVTFNGLTPNQMHVMMYGVYRLRAFG\n\n### Mutant:\nMLSESSSFLKGVMLGSIFCALITMLGHIRIGHGNRMHHHEHHHLQAPNKEDILKISEDERMELSKSFRVYCIILVKPKDVSLWAAVKETWTKHCDKAEFFSSENVKVFESINMDTNDMWLMMRKAYKYAFDKYRDQYNWFFLARPTTFAIIENLKYFLLKKDPSQPFYLGHTIKSGDLEYVGMEGGIVLSVESMKRLNSLLNIPEKCPEQGGMIWKISEDKQLAVCLKYAGVFAENAEDADGKDVFNTKSVGLSIKEAMTYHPNQVVEGCCSDMAVTFNGLTPNQMHVMMYGVYRLRAFG\n\n### Answer:", "output": "Increases capacity to promote Tn synthase activity. The somatic mutation in the Cosmc gene enhances the proper folding and activity of T-synthase. Novel activating mutation in the coding sequence of C1GALT1C1.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01963", "category": "Mol", "instruction": "Answer the protein-ligand question using the provided molecular evidence.", "input": "Target: ALK\nLigand SMILES: COc1ccc(CN(CCN(C)C)c2ccccn2)cc1\nEvidence: No specific activity evidence is provided; the question is a general relationship inference.", "output": "Without supporting bioactivity evidence, the relationship between this ligand and anaplastic lymphoma kinase (ALK) is best described as an uncharacterized or putative association rather than a confirmed binding claim.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "protein_ligand_qa", "source": "BBBP", "canonical_smiles": "COc1ccc(CN(CCN(C)C)c2ccccn2)cc1", "compound_name": null, "pubchem_cid": null, "MW": 285.39, "LogP": 2.66, "TPSA": 28.6, "QED": 0.78, "functional_groups": ["aromatic ring", "amine", "ether", "pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_protein_ligand", "target": "ALK", "has_evidence": false}}
{"id": "eval_01964", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGAMQGEVSKAASADSTTEGTPADGFTVLSTKSLFLGQKLQVVQADIASIDSDAVVHPTNTDFYIGGEVGNTLEKKGGKEFVEAVLELRKKNGPLEVAGAAVSAGHGLPAKFVIHCNSPVWGADKCEELLEKTVKNCLALADDKKLKSIAFPSIGSGRNGFPKQTAAQLILKAISSYFVSTMSSSIKTVYFVLFDSESIGIYVQEMAKLDAN\n\n### Answer:", "output": "<SEQ_2D>CCCCCCCCCCCCCCCCCCCCCCCCCEEEEEEECTTSCEEEEEESCGGGCCCSEEEEECCTTCCCCHHHHHHHHHHHHHHHHHHHHHHHHHHCSCCTTCEEEEECTTSSSSEEEEECCCCTTSTTHHHHHHHHHHHHHHHHHHTTCSEEEEECCSSSTTCCCHHHHHHHHHHHHHHHHHHCTTCCCCEEEEECCSHHHHHHHHHHHHTCCCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01965", "category": "Mutation", "instruction": "Describe the effect of mutation P07143-K288L on protein Cytochrome c1, heme protein, mitochondrial.", "input": "### Wild-type:\nMFSNLSKRWAQRTLSKSFYSTATGAASKSGKLTQKLVTAGVAAAGITASTLLYADSLTAEAMTAAEHGLHAPAYAWSHNGPFETFDHASIRRGYQVYREVCAACHSLDRVAWRTLVGVSHTNEEVRNMAEEFEYDDEPDEQGNPKKRPGKLSDYIPGPYPNEQAARAANQGALPPDLSLIVKARHGGCDYIFSLLTGYPDEPPAGVALPPGSNYNPYFPGGSIAMARVLFDDMVEYEDGTPATTSQMAKDVTTFLNWCAEPEHDERKRLGLKTVIILSSLYLLSIWVKKFKWAGIKTRKF\n\n### Mutant:\nMFSNLSKRWAQRTLSKSFYSTATGAASKSGKLTQKLVTAGVAAAGITASTLLYADSLTAEAMTAAEHGLHAPAYAWSHNGPFETFDHASIRRGYQVYREVCAACHSLDRVAWRTLVGVSHTNEEVRNMAEEFEYDDEPDEQGNPKKRPGKLSDYIPGPYPNEQAARAANQGALPPDLSLIVKARHGGCDYIFSLLTGYPDEPPAGVALPPGSNYNPYFPGGSIAMARVLFDDMVEYEDGTPATTSQMAKDVTTFLNWCAEPEHDERKRLGLKTVIILSSLYLLSIWVLKFKWAGIKTRKF\n\n### Answer:", "output": "Loss of CYT1 and COB from the bc1 complex; when associated with L-289 and L-296.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01966", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMSRLIVVSNRVAAGEDTRPSAGGLAVGVMDALKQTGGVWFGWNGEIVGTPDAAPAVRRDGNVTYATLGLTRRDYDQYYRGFSNATLWPVFHYRSDLARFDRQEYAGYLRVNAMLARQLAALLRPDDLIWVHDYHLLPFAHCLRELGVKNPIGFFLHIPFPSPDMLRVVPPHDELVKFMCAYDIAGFQTDADKRAFSDYIERRGIGTAMLHAHGRVVKVAAYPIGVYPDAIAEAAVQYGGRKPVKMLRDALGGRKLVMSVDRLDYSKGLVERFQSFERMLAGAPDWQGRVSFVQIAPPTRSDVQTYQRIRETLEREAGRINGRFAQLDWTPIQYLNRKYERNLLMAFFRMSQVGYVTPLRDGMNLVAKEYVASQDPADPGVLVLSEFAGAAAELGGALLVNPYDHAQMADALARALAMPLAERQARHEENLAQLRNNDLSVWRDTFVADLRSVAAAA\n\n### Answer:", "output": "<SEQ_3Di>DAAAEEEEAADDDDPCLPCLNPPPNVLCVVRNQVHLYEYEHAHPAADCDWDPAWDWDDDDSYIYTHTHHHPLLCVQLPLNCVVQAVPCLLLVNLVRHDDDPSSVVSLVVVLLVSLVNVVVVDDLAYQYEYEADSRLQNLVSNVVVVRPAQYEYEHFWAADADVSCVSRVCSVVSLVSNLSHQEYEAQDPVRLVRSVCVCVVVVQFDPCTDGPNHRHHYYYQHAFADLVVLQVLLVVCLPPPVNVVVVVLLVQAAEEEEEEALDLQFLLLLLLVLLLVLCVVPVVSQPRYAYEYAYAYDSCVRVVSVVSVVVNVVSLVVSQVVRPDPSDRRYHYHNDDDRSSVSLNVLQRHQEYEGAGSAGRADHVQNSSLSSHDLQAGHAYEYEPRHSVVVQQVLHNYDHSVDSNRSSVSVVCRVPPDRVSRNVSSVRNVVSSVVNHSVSSNVVVVVVSVVVSVVD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01967", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nLPXC_IDILO\n\n### Answer:", "output": "Belongs to the LpxC family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01968", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nGSALKRINKELSDLARDPPAQCSAGPVGDDMFHWQATIMGPNDSPYQGGVFFLTIHFPTDYPFKPPKVAFTTRIYHPNINSNGSICLDILRSQWSPALTISKVLLSICSLLCDPNPDDPLVPEIARIYKTDRDKYNRISREWTQKYAM\n\n### Answer:", "output": "<SEQ_3Di>DCVVVQLVVQVVCCVVPPADQWHWDQPDPDSQKIKIKHAQDPLALLHGFIFIKIWGHDPPAPVAAIWIFTCFQAPAQQQDGRGGGDDCCNHPVVDRPDHPNVVVVVVNVCRSPPDNVPTDDNPLSCCCVPPVVVNSVVRNVCRVVGRD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01969", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nKVCEISGKRPIVANSIQRRGKAKREGGVGKKTTGISKRRQYPNLQKVRVRVAGQEITFRVAASHIPKVYELVERAKGLKLEGLSPKEIKKELL\n\n### Answer:", "output": "<SEQ_2D>CCBTTTBCCCEEETSSCCCCCCCCSSSSSSSSCCCCCEEECCCCCCEEECCSSSCEEECCCSTTHHHHHHHHHHHHTSCCCSSCHHHHHHHHC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01970", "category": "Mutation", "instruction": "Describe the effect of mutation Q9NY72-A130V on protein Sodium channel subunit beta-3.", "input": "### Wild-type:\nMPAFNRLFPLASLVLIYWVSVCFPVCVEVPSETEAVQGNPMKLRCISCMKREEVEATTVVEWFYRPEGGKDFLIYEYRNGHQEVESPFQGRLQWNGSKDLQDVSITVLNVTLNDSGLYTCNVSREFEFEAHRPFVKTTRLIPLRVTEEAGEDFTSVVSEIMMYILLVFLTLWLLIEMIYCYRKVSKAEEAAQENASDYLAIPSENKENSAVPVEE\n\n### Mutant:\nMPAFNRLFPLASLVLIYWVSVCFPVCVEVPSETEAVQGNPMKLRCISCMKREEVEATTVVEWFYRPEGGKDFLIYEYRNGHQEVESPFQGRLQWNGSKDLQDVSITVLNVTLNDSGLYTCNVSREFEFEVHRPFVKTTRLIPLRVTEEAGEDFTSVVSEIMMYILLVFLTLWLLIEMIYCYRKVSKAEEAAQENASDYLAIPSENKENSAVPVEE\n\n### Answer:", "output": "In ATFB16; results in decreased sodium current density. Mutation dramatically decreased the cardiac sodium current density when expressed in HEK293/Na(v)1. 5 stable cell line, but did not have significant effect on kinetics of activation, inactivation, and channel recovery from inactivation. When co-expressed with wild type SCN3B, the mutant SCN3B negated the function of wild type SCN3B, suggesting that acts by a dominant negative mechanism. Western blot analysis with biotinylated plasma membrane protein extracts revealed that did not affect cell surface expression of Na(v)1. 5 or SCN3B, suggesting that mutant SCN3B may not inhibit sodium channel trafficking, instead may affect conduction of sodium ions due to its malfunction as an integral component of the channel complex.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01971", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nNADA_ANOFW\n\n### Answer:", "output": "Quinolinate synthase.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01972", "category": "Cell", "instruction": "Infer the most likely tissue of origin from this cell profile.", "input": "CXCR2, S100A9, CSF3R", "output": "This profile is most consistent with neutrophil, which is commonly found in liver under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "neutrophil", "tissue": "liver", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["CXCR2", "S100A9", "CSF3R"]}}
{"id": "eval_01973", "category": "Cell", "instruction": "What cell type is represented by this marker gene list?", "input": "PPBP, GP1BA, PF4", "output": "This cell is most consistent with megakaryocyte, a megakaryocyte. Key markers such as GP1BA, PPBP, PF4 support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "megakaryocyte", "tissue": "lymph node", "disease": "colorectal carcinoma", "organism": "Homo sapiens", "assay": "10x multiome", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PPBP", "GP1BA", "PF4"]}}
{"id": "eval_01974", "category": "Structure", "instruction": "Predict the secondary-structure string for the amino-acid sequence below. Use DSSP-style symbols.", "input": "### Sequence 1D:\nGHMSELRILRAVDYPRMPWKNGAGSTEEIARDGGDGLDGFGWRLSIADVGESGGFSGFAGYQRIISVLEGGGMRLRVDGAESAPLRARQAFAFSGDSEVHCTLLDGAIRDFNLIYAPRRHRARLQWLRVEGELDWHGTASTLLLFAQQDGVAISLQGQPRGQLAAHDCLCAEGLQGLQHWRLTAHEPAWVCAVELDSLGS\n\n### Answer:", "output": "<SEQ_2D>CCCCCEEEECGGGCCCEECTTSSEEEEEEEESSCSGGGSCSEEEEEEEECSCEEECCCTTCEEEEEEEESSCEEEEETTEECCCBCBTCCEEEETTSCEEEEESSSCEEEEEEEECTTTEEEEEEEEEEEEEEEEEECCSEEEEEESSSCEEEEETTEEEEEECTTCEEEECSCCSCEEEEEEEEEEEEEEEEEEEECCC</SEQ_2D>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_secondary_structure", "source": "ss.txt"}}
{"id": "eval_01975", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nTLSNSIRMLGSQSPLIQAYGLVILQQPDIKVNAMSSLTNHQKFAKANVREWIDEYNPKLIDLNQEMMRYSIRFNSYYSKLYELAGNINKADFTNAYGKLQLQVQSIQENMEQDLLELNRFKTVLDKDSNNLSIKADEAIKTLQGDIVKLREDIKRIQGEIQAELTTILNRPQEIIKGSINIGKQVFTITNTKTIDFVSIGTLSNEIVNAADSQTREAALRIQQKQKELLPLIQKLSQTEAEATQITFVEDQVSSFTELIDRQITTLETLLTDWKVLNNNMIQIQKNVEETDSSLLQKHFNQIKKVSDEMNKQTNQFEDYVTNVEVH\n\n### Answer:", "output": "<SEQ_3Di>DLPQLLLLLQLLPQVLVQLLVLLVPFDQDADPLFRCPNVLSVVSNVLSVCCVPPQVVLSLVLLVLLLVLLVVCVVCVVVLLVCLQVVCPVPNLVVLVVSLVSLVVSLVSLVVSLVVLVVSLVVLVVSLVVLVVSLVVSQVVVDVLLVVLLVVVLVLVVLLLVLSSVLSPDDDPQQAWEWEKDWDWDDEDDDDIDIDIHTDAIGCCQCVPPDPSSVVSNVSSVVSVVVSPVSLNVSCNSPVVSSVSSRNSSSSVSSSVSSVVSSVSSVSSSVVSVVLSVLSVVLSCCRVPPNSVVVSVSSVVNSVVSVVSNVSSVSSVSSVPRHYYD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01976", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "CN(C)C(=C1C(Br)=C(Br)C(Br)=C1Br)N(C)C", "output": "This molecule is medium-sized and relatively nonpolar, with lipophilic character. It contains amine, halide. Its descriptor profile: MW=479.84, LogP=4.34, TPSA=6.48, HBD=0, HBA=2, QED=0.58.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "HIV", "canonical_smiles": "CN(C)C(=C1C(Br)=C(Br)C(Br)=C1Br)N(C)C", "compound_name": null, "pubchem_cid": null, "MW": 479.84, "LogP": 4.34, "TPSA": 6.48, "QED": 0.58, "functional_groups": ["amine", "halide"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01977", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nEVQLVESGAEVKKPGASVKVSCKASGYTFTNYAIHWVRQAPGQRLEWMGWINAGRGNTKYSQKFQGRVTITRDTSASTAYMELSSLRSEDAAVYFCARDESTGDYYYYMDVWGKGTTVTVSSASTKGPSVFPLAPSAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEP\n\n### Answer:", "output": "<SEQ_3Di>DWAKAKDAEAEEEAQAKDKIKIQTDDDQLQWWKKWKWWDAVPGDIGTAKIARSNPGDIDGDCVLVPQWDWGDDRVRSMIMIMGGRDDQVPFTWMKMFIFPTYCPDPPDTDGIHPTYTYGYHPDDFDAWDWDKQAPPVKIKIKTDFGDDDDKDKDKPVPPDDPQKDWDDWDQDPVRGIITMIMGDDDPVCAPPDWIWMWMADPVVRDIDIDTHHD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01978", "category": "Mol", "instruction": "Describe key molecular descriptors for the given SMILES.", "input": "O=C(Cl)c1ccccc1", "output": "This molecule is small and relatively nonpolar, with moderately lipophilic character. It contains aromatic ring, halide. Its descriptor profile: MW=140.57, LogP=2.07, TPSA=17.07, HBD=0, HBA=1, QED=0.55.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Tox21", "canonical_smiles": "O=C(Cl)c1ccccc1", "compound_name": null, "pubchem_cid": null, "MW": 140.57, "LogP": 2.07, "TPSA": 17.07, "QED": 0.55, "functional_groups": ["aromatic ring", "halide"], "moleculenet_labels": {"NR-AR": 0.0, "NR-AR-LBD": 0.0, "NR-AhR": 0.0, "NR-Aromatase": 0.0, "NR-ER": 0.0, "NR-ER-LBD": 0.0, "NR-PPAR-gamma": 0.0, "SR-ARE": 0.0, "SR-ATAD5": 0.0, "SR-HSE": 0.0, "SR-MMP": 0.0, "SR-p53": 0.0}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01979", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMNINFTLISQAMAFAIFIWFTVRFVWPPLMRAIENRQKTIAEGLAAGERGKRELELASQRSGDVVREAKQRASDIIAQAEKRAAEIVDEAKVAAREEGDRILVGAKAEVEQEVFRAKEVLRQQVAGLALAGAAKILRREVDEKAHAELLASLKAEL\n\n### Sequence 2:\nMNINFTLISQAIAFSLFILFTARFVWPYLLRAIEERQQKIADGLAAGERGKKELELASQRSSEVLKEAKQRASEIVIQAEKRASDIIEEAKQNARIEGEKIIAGAKAEIQHETFSARESLRQQVAGLAVQGASKILRREVNAKVHADLLASIEAEL\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01980", "category": "Mutation", "instruction": "Describe the effect of mutation P09391-H254C on protein Rhomboid protease GlpG.", "input": "### Wild-type:\nMLMITSFANPRVAQAFVDYMATQGVILTIQQHNQSDVWLADESQAERVRAELARFLENPADPRYLAASWQAGHTGSGLHYRRYPFFAALRERAGPVTWVMMIACVVVFIAMQILGDQEVMLWLAWPFDPTLKFEFWRYFTHALMHFSLMHILFNLLWWWYLGGAVEKRLGSGKLIVITLISALLSGYVQQKFSGPWFGGLSGVVYALMGYVWLRGERDPQSGIYLQRGLIIFALIWIVAGWFDLFGMSMANGAHIAGLAVGLAMAFVDSLNARKRK\n\n### Mutant:\nMLMITSFANPRVAQAFVDYMATQGVILTIQQHNQSDVWLADESQAERVRAELARFLENPADPRYLAASWQAGHTGSGLHYRRYPFFAALRERAGPVTWVMMIACVVVFIAMQILGDQEVMLWLAWPFDPTLKFEFWRYFTHALMHFSLMHILFNLLWWWYLGGAVEKRLGSGKLIVITLISALLSGYVQQKFSGPWFGGLSGVVYALMGYVWLRGERDPQSGIYLQRGLIIFALIWIVAGWFDLFGMSMANGACIAGLAVGLAMAFVDSLNARKRK\n\n### Answer:", "output": "Loss of catalytic activity. The mutation in GlpG was found to be essential for the proteolytic activities. The mutation in the AarA protein results in a nonfunctional protein.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01981", "category": "Cell", "instruction": "Which tissue is this cell most likely from?", "input": "MS4A7, CDKN1C, FCGR3A, LST1", "output": "This profile is most consistent with CD16+ monocyte, which is commonly found in lymph node under typical physiological conditions.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "profile_to_tissue", "source": "panglao_cellmarker_panels", "cell_type": "CD16+ monocyte", "tissue": "lymph node", "disease": "inflammatory bowel disease", "organism": "Homo sapiens", "assay": "Drop-seq", "split": "train", "generation_method": "template_tissue_inference", "top_genes": ["MS4A7", "CDKN1C", "FCGR3A", "LST1"]}}
{"id": "eval_01982", "category": "Cell", "instruction": "Based on the markers below, determine the cell type.", "input": "PDGFRB, ACTA2, S100A4, COL1A1, FAP", "output": "This cell is most consistent with cancer-associated fibroblast, a cancer-associated fibroblast (CAF). Key markers such as FAP, ACTA2, PDGFRB support this identity.", "metadata": {"source_file": "cell_sft_master.jsonl", "task": "marker_to_type", "source": "panglao_cellmarker_panels", "cell_type": "cancer-associated fibroblast", "tissue": "tumor", "disease": "rheumatoid arthritis", "organism": "Homo sapiens", "assay": "10x 5' scRNA-seq", "split": "train", "generation_method": "template_from_marker_genes", "top_genes": ["PDGFRB", "ACTA2", "S100A4", "COL1A1", "FAP"]}}
{"id": "eval_01983", "category": "Mutation", "instruction": "Describe the effect of mutation Q96P20-R157E on protein NACHT, LRR and PYD domains-containing protein 3.", "input": "### Wild-type:\nMKMASTRCKLARYLEDLEDVDLKKFKMHLEDYPPQKGCIPLPRGQTEKADHVDLATLMIDFNGEEKAWAMAVWIFAAINRRDLYEKAKRDEPKWGSDNARVSNPTVICQEDSIEEEWMGLLEYLSRISICKMKKDYRKKYRKYVRSRFQCIEDRNARLGESVSLNKRYTRLRLIKEHRSQQEREQELLAIGKTKTCESPVSPIKMELLFDPDDEHSEPVHTVVFQGAAGIGKTILARKMMLDWASGTLYQDRFDYLFYIHCREVSLVTQRSLGDLIMSCCPDPNPPIHKIVRKPSRILFL\n\n### Mutant:\nMKMASTRCKLARYLEDLEDVDLKKFKMHLEDYPPQKGCIPLPRGQTEKADHVDLATLMIDFNGEEKAWAMAVWIFAAINRRDLYEKAKRDEPKWGSDNARVSNPTVICQEDSIEEEWMGLLEYLSRISICKMKKDYRKKYRKYVRSRFQCIEDRNAELGESVSLNKRYTRLRLIKEHRSQQEREQELLAIGKTKTCESPVSPIKMELLFDPDDEHSEPVHTVVFQGAAGIGKTILARKMMLDWASGTLYQDRFDYLFYIHCREVSLVTQRSLGDLIMSCCPDPNPPIHKIVRKPSRILFL\n\n### Answer:", "output": "Impaired ability to activate the NLRP3 inflammasome.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01984", "category": "Mutation", "instruction": "Describe the effect of mutation Q9BU02-E81A on protein Thiamine-triphosphatase.", "input": "### Wild-type:\nMAQGLIEVERKFLPGPGTEERLQELGGTLEYRVTFRDTYYDTPELSLMQADHWLRRREDSGWELKCPGAAGVLGPHTEYKELTAEPTIVAQLCKVLRADGLGAGDVAAVLGPLGLQEVASFVTKRSAWKLVLLGADEEEPQLRVDLDTADFGYAVGEVEALVHEEAEVPTALEKIHRLSSMLGVPAQETAPAKLIVYLQRFRPQDYQRLLEVNSSRERPQETEDPDHCLG\n\n### Mutant:\nMAQGLIEVERKFLPGPGTEERLQELGGTLEYRVTFRDTYYDTPELSLMQADHWLRRREDSGWELKCPGAAGVLGPHTEYKALTAEPTIVAQLCKVLRADGLGAGDVAAVLGPLGLQEVASFVTKRSAWKLVLLGADEEEPQLRVDLDTADFGYAVGEVEALVHEEAEVPTALEKIHRLSSMLGVPAQETAPAKLIVYLQRFRPQDYQRLLEVNSSRERPQETEDPDHCLG\n\n### Answer:", "output": "Mildly decreases enzyme activity.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01985", "category": "Literature", "instruction": "What is the protein family that this protein belongs to?", "input": "### Protein Sequence:\nVP1_POVS1\n\n### Answer:", "output": "Belongs to the polyomaviruses coat protein VP1 family.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01986", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nSHVKDILGLINAFNEVKKITVDGTTPITVAHVAALARRHDVKVALEACRARVETCSSWVQRKAEDGADIYGVTTGFGACSSRRTNRLSELQESLIRCLLAGVFTELPATATRSAMLLRLNSFTYGCSGIRWEVMEALEKLLNSNVSPKVPLRGSVSDLIPLAYIAGLLIGKPSVIARIGDDVEVPAPEALSRVGLRPFKLQAKEGLALVNGTSFATAVASTVMYDANVLLLLVETLCGMFCEVIFGREEFAHPLIHKVKPHPGQIESAELLEWLLRSSPFQELSREYYSIDKLKKPKQDRYALRSSPQWLAPLVQTIRDATTTVETEVNSANDNPIIDHANDRALHGANFQGSAVGFYMDYVRIAVAGLGKLLFAQFTELMIEYYSNGLPGNLSLGPDLSVDYGLKGLDIAMAAYSSELQYLANPVTTHVHSAEQHNQDINSLALISARKTEEALDILKLMIASHLTAMCQAVDLRQLEEALVKVVENVVSTLADECGLPNDTKARLLYVAKAVPVYTYLESPCDPTLPLLLGLKQSCFDTILALTLVDRLAEFEKRLSDRLENEMTAVRVLYEVRIQGSKFLPFYRFVREELDTGVMSARREQTPQEDVQKVFDAIADGRITVPLLHCLQGFL\n\n### Answer:", "output": "<SEQ_3Di>DLVVVLVVLLVLLVPAQEAEQALPDFDALSNLLSLLFPLSRAYDYPCLVVLQVLLLVLQQVCLQVVDQFFLQQAADFPRRVHGHSPRLVSLVVLLVVLLDADPCFDDLSLQLQLLLLLLNLSSNNFQSDHVVVSVVSRLCSNQSKGWTAHQDDDLQHQNRLSLVLCLQQQPQPIWMDRPDPDIHGSQVVCVVSVHHHDRDGGRHSQSSRQANSSLLSLLSNLLSLLSVLLLLLLLVLLLLCLLLLADLCLLPQVVCVVVPLPQLNVSSLNNCVQPVPAPSNVVSVVQCVQPVVVDDPHADCLNSVVSVLCRVLSVLSVVLSVLSSVRSNHNRGPQGDPVVVSDTHGDGSSFPASLLVSLLVNLVSLLSSLVSLLVSVVLCQDPVRRSQADHLNAPDLPVVPTNHNVVVNVVSVVLSVVSVVLSDGPQVPQDQDDVSSCNHGRSSSVSSVSSVVSSLSSLLSSLSSLVSSLLSQLVVALQVVLLVLLLVLLQVLCVVLVADPQLSVQLSSLSNNDRVSQCLSPNLPCPTRNVVSNVVSLVVVVVVHVNVVSSVVSSVSCNVPRNVVSVVSVVVSLHSLVVGLLVLSSCCLCPVVPDGNDHPVDPDDPCSSSVSSSVCSVSVVVVVSVCSSCPPPD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01987", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMQTKEKLASLSYLAVKSLLYELNLTPKPGLVDCHNNGAHNDMDFYTFLDSILSLSPFFKKYIEVGWLYHNESPQYLFNQLRKLGIEAEAAMFSATERVNTHKGINFSFALLLGATGSYLAKHIELIQEKRRFMPQDSLTICHLAGEMSMHLIQNDLSHVETKRNLTYGEKLFLQYGLKGLRGEASQGYPSLTQKALPFFRNELLKKQDIQISQLKLLLYLMTFIEDSNIIHRGGIKSWKKVQQEAQTLLEKDLPPYQLKEQLNSYNQILTDRHLSPGGAADLLSLTLYFSFLEQLI\n\n### Sequence 2:\nMEENDHSNRDVERQDSGDESNRAILPLLQAPGNVLPHRITNFYIDNILRPDFGRRKEGSRRDEINIVERENRCPSAPGSGQVAPVSGEGTSSPRAVNASKKTDISTDESLKSRAETGDQCLSSDSDCSQRCAAQAKQPMLWPAWVYCTRYSDRPSSGPRSRKPKKKTPTKEDKRPRTAFTAEQLQRLKNEFQNNRYLTEQRRQALAQELGLNESQIKIWFQNKRAKIKKATGNKNTLAVHLMAQGLYNHATVTKDDKSDSD\n\n### Answer:", "output": "Non-Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01988", "category": "Structure", "instruction": "Predict the Foldseek 3Di structural alphabet sequence for the amino-acid sequence below.", "input": "### Sequence 1D:\nMKLTGKQTWEFENPLFVNSSGTAVGPKEKEGPLGHLFDKSYDEMHCNQKNWEMAERKLMEDAVQSALSKQNLKKEDIDIFLAGDLLNQNVTANYVARHLKIPFLCLFGACSTSMESIAISSALIDGGFAKRALAATSSHNATAERQFTVTGSGAVVLSQQPGGIKITSATVGRVIDLGITDSQDMGSAMAPAAADTIKQHLEDLGRTPDDYDLILTGDLSGVGSPILKDLLKEEGINVGTKHNDCGLMIYTPGCACSAVVTFAHIFKEIEAGRLNRVLVVATGALLSPTIIQQKESIPCIAHGVVFERA\n\n### Answer:", "output": "<SEQ_3Di>DQDFFDFKGFQDAAKFFAFKAKAFECQLLPFPLNVLHPHYYNHQQCPPDDNLVSFLVQFVRGQVVQCVSVVHALVLWAEEEEAEQADGRSNVVVVCVVSVHHYDYDYHWLQQQQVQVQVQSCCQSVPNTAKYKYKYADGPVCNPQFLATHMMMMTMMHSDDHFKTWGMKGWFDKDFDWDLPPVVVLVGFLVQLLVRVVSVCVRVVHALVVFQAEEEEQNPDSNPVVNQVSNVVVVHHNDVRYDYQNHRRGDVSHSDQRVCCNRPVSVCQSVVVTFKYKYKGKTWPGGSVCNVVSGTIIMHIMIIMMGGD</SEQ_3Di>", "metadata": {"source_file": "structure_task_sft.jsonl", "type": "aa_to_3di", "source": "pdb_aa+pdb_3di"}}
{"id": "eval_01989", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMAFVVGLTGGIASGKTTIANMFAELGATIIDADIVARQVVTKGSPLFLKIVQHFGQQVLTTQGELNRAQLRQLIFANQVEKNWLNNLLHPAIRREMLIQLRQARGPYVLFVVPLLIENKLIEFCQRVLVIDVYPEVQLARALTRDRSNIATIRGIMASQVNRLTRLSYANDIIENNLPLAESLERLQMQVQQLHQYYLTLATQQE\n\n### Sequence 2:\nMAFVVGLTGGIASGKTTIANMFAELGATIIDADIVARQVVTKGSPLFLKIVQHFGQQVLTTQGELNRAQLRQLIFANQVEKNWLNNLLHPAIRREMLIQLRQARGPYVLFVVPLLIENKLIEFCQRVLVIDVYPEVQLARALTRDRSNIATIRGIMASQVNRLTRLSYANDIIENNLPLAESLERLQMQVQQLHQYYLTLATQQE\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01990", "category": "Protein", "instruction": "Classify the pair below as homologous or non-homologous.", "input": "### Sequence 1:\nMADESKFDQFKGNVKETVGNVTDNKELEKEGQQDKVIGKAKEVVENAKNKITDAIDKLKK\n\n### Sequence 2:\nMADESKFDQFKGNVKETVGNVTDNKELEKEGQQDKATGKAKEVVENAKNKITDAIDKLKK\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01991", "category": "Mol", "instruction": "Summarize the physicochemical properties of this molecule.", "input": "Cc1cc(Nc2nc(N[C@@H](C)c3ncc(F)cn3)ncc2F)n[nH]1", "output": "This molecule is medium-sized and polar, with moderately lipophilic character. It contains halide. Its descriptor profile: MW=332.32, LogP=2.49, TPSA=104.3, HBD=3, HBA=7, QED=0.66.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "property_summary", "source": "Lipophilicity", "canonical_smiles": "Cc1cc(Nc2nc(N[C@@H](C)c3ncc(F)cn3)ncc2F)n[nH]1", "compound_name": null, "pubchem_cid": null, "MW": 332.32, "LogP": 2.49, "TPSA": 104.3, "QED": 0.66, "functional_groups": ["halide"], "moleculenet_labels": {"label": 1.79}, "split": "train", "generation_method": "template_from_rdkit_descriptors"}}
{"id": "eval_01992", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nSELO_YERPP\n\n### Answer:", "output": "Catalyzes the transfer of adenosine 5'-monophosphate (AMP) to Ser, Thr or Tyr residues of target proteins (AMPylation).", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01993", "category": "Mutation", "instruction": "Describe the effect of mutation P13501-A31T on protein C-C motif chemokine 5.", "input": "### Wild-type:\nMKVSAAALAVILIATALCAPASASPYSSDTAPCCFAYIARPLPRAHIKEYFYTSGKCSNPAVVFVTRKNRQVCANPEKKWVREYINSLEMS\n\n### Mutant:\nMKVSAAALAVILIATALCAPASASPYSSDTTPCCFAYIARPLPRAHIKEYFYTSGKCSNPAVVFVTRKNRQVCANPEKKWVREYINSLEMS\n\n### Answer:", "output": "Enhanced inhibition of activity by tick evasin-4.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01994", "category": "Protein", "instruction": "Determine whether these two protein sequences are homologous.", "input": "### Sequence 1:\nMLFPTPLISGRLERRYKRFLADVTLDDGRFITASVPNTGSMLGLTAPGSRVWLSFSDAPHRKYAHTLQIVEADNTLVGVNTGLPNRIAEEAILKGLIPDLDGYATLKREQKYGRNSRIDLLLDDGPRPRAYVEVKNVHFIRTPGLAEFPDTVTARGAKHLDELVDVVAAGHRGIMLFITQRADCSRFGISGDLDPFYARAFERAIASGVEAWAVRCHITENGIDATELVPIEDMRRIE\n\n### Sequence 2:\nMLFTPPLIPATLISRYKRFLFDAVLEDGTAITGSCPNTGSMRGLTTPGSRIWLSEHDSPTRKYRHMFEMVEADGTVVGINTGMPNRLAEEAILNGRIPELAGYSTIRREQKYGRNSRIDFLLSEPGRPDAYVEVKNVHFMREKGLAEFPDTATKRGAKHLEELGDAAEAGYRSVMLYLIQRDDCERMRICADLDPIYALAFQRAMARGVEAYAVKCAVSPTQISVSGTVKMDEWR\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_no_length_limit_but_diff_limited.csv"}}
{"id": "eval_01995", "category": "Mutation", "instruction": "Describe the effect of mutation P04509-L70A on protein Intermediate capsid protein VP6.", "input": "### Wild-type:\nMDVLYSLSKTLKDARDKIVEGTLYSNVSDLIQQFNQMIITMNGNEFQTGGIGNLPIRNWNFDFGLLGTTLLNLDANYVETARNTIDYFVDFVDNVCMDEMVRESQRNGIAPQSDSLIKLSGIKFKRINFDNSSEYIENWNLQNRRQRTGFTFHKPNIFPYSASFTLNRSQPAHDNLMGTMWLNAGSEIQVAGFDYSCAINAPANTQQFEHIVQLRRVLTTATITLLPDAERFSFPRVITSADGATTWYFNPVILRPNNVEIEFLLNGQIINTYQARFGTIIARNFDTIRLSFQLMRPPNM\n\n### Mutant:\nMDVLYSLSKTLKDARDKIVEGTLYSNVSDLIQQFNQMIITMNGNEFQTGGIGNLPIRNWNFDFGLLGTTALNLDANYVETARNTIDYFVDFVDNVCMDEMVRESQRNGIAPQSDSLIKLSGIKFKRINFDNSSEYIENWNLQNRRQRTGFTFHKPNIFPYSASFTLNRSQPAHDNLMGTMWLNAGSEIQVAGFDYSCAINAPANTQQFEHIVQLRRVLTTATITLLPDAERFSFPRVITSADGATTWYFNPVILRPNNVEIEFLLNGQIINTYQARFGTIIARNFDTIRLSFQLMRPPNM\n\n### Answer:", "output": "Loss of in vitro DLP transcriptase activity, no effect on particle assembly; when associated with D-65. The mutation in VP6 partially assembles with VP2 and does not rescue the transcriptase activity of reconstituted DLP.", "metadata": {"source_file": "mutadescribe_sft.jsonl", "type": "mutation_effect_description", "source": "MutaDescribe"}}
{"id": "eval_01996", "category": "Literature", "instruction": "What are the official names of this protein?", "input": "### Protein Sequence:\nELBB_SHIFL\n\n### Answer:", "output": "Glyoxalase ElbB.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01997", "category": "Literature", "instruction": "What is the function of this protein?", "input": "### Protein Sequence:\nARGB_LISMC\n\n### Answer:", "output": "Catalyzes the ATP-dependent phosphorylation of N-acetyl-L-glutamate.", "metadata": {"source_file": "uniprotqa_sft.jsonl", "type": "protein_knowledge_qa", "source": "UniProtQA"}}
{"id": "eval_01998", "category": "Protein", "instruction": "Are these two protein sequences structurally related like paraphrases?", "input": "### Sequence 1:\nMTFTASSSSCAITESPVVVALDYHERDKALAFVDKIDPRDCRLKVGKEMFTLFGPQLVRDLQQRGFDVFLDLKFHDIPNTTARAVAAAADLGVWMVNVHASGGARMMAAARDALAPFSKDAPLLIAVTVLTSMETSDLHDLGVTLSPAEHAERLARLTQQCGLDGVVCSAQEAVRFKQAFGAAFKLVTPGIRPAGSEAGDQRRIMTPEQALSAGVDYMVIGRPVTQSVDPAQTLKDINASLKREA\n\n### Sequence 2:\nMTFTASSSSCAITESPVVVALDYHERDKALAFVDKIDPRDCRLKVGKEMFTLFGPQLVRDLQQRGFDVFLDLKFHDIPNTTARAVAAAADLGVWMVNVHASGGARMMAAARDALAPFSKDAPLLIAVTVLTSMGTSDLHDLGVTLSPAEHAERLARLTQQCGLDGVVCSAQEAVRFKQAFGAAFKLVTPGIRPAGSEAGDQRRIMTPEQALSAGVDYMVIGRPVTQSVDPAQTLKDINASLKREA\n\n### Answer:", "output": "Homologous", "metadata": {"source_file": "homology_task_sft.jsonl", "type": "protein_homology_classification", "source": "protein_pair_20k_length_restricted_balanced.csv"}}
{"id": "eval_01999", "category": "Mol", "instruction": "Assess the drug-likeness of the given molecule.", "input": "SMILES: CCC[CH2][Sn]([CH2]CCC)([O]C(=O)c1cccnc1SC)[O]C(=O)c1cccnc1SC\nDescriptors: MW=569.34, LogP=5.98, TPSA=78.38, HBD=0, HBA=8", "output": "Descriptors: MW=569.34, LogP=5.98, TPSA=78.38, HBD=0, HBA=8, QED=0.23, Lipinski violations=2. Overall the molecule is somewhat outside the typical oral drug-like ranges, and a lower QED score suggests limited drug-likeness.", "metadata": {"source_file": "mol_sft_master.jsonl", "task": "drug_likeness", "source": "HIV", "canonical_smiles": "CCC[CH2][Sn]([CH2]CCC)([O]C(=O)c1cccnc1SC)[O]C(=O)c1cccnc1SC", "compound_name": null, "pubchem_cid": null, "MW": 569.34, "LogP": 5.98, "TPSA": 78.38, "QED": 0.23, "functional_groups": ["pyridine"], "moleculenet_labels": {"label": 0}, "split": "train", "generation_method": "template_drug_likeness"}}