update

scDifformer_data_process/loom_batch_100.py

@@ -18,32 +18,32 @@ def set_log() -> None:
 set_log()
 
 
-# 输入文件夹路径和输出文件夹路径
+# Input and output folder paths
 input_folder = '/mnt/nfs/data/geneformer/output/v2/looms'
 output_folder = '/mnt/nfs/data/geneformer/output/v2/looms_sub'
 per_group_num = 100
 
 
 if __name__ == '__main__':
-    # 创建输出文件夹（如果不存在）
+    # Create output folder if it does not exist
     os.makedirs(output_folder, exist_ok=True)
 
-    # 遍历输入文件夹中的文件
+    # Iterate through files in the input folder
     files = os.listdir(input_folder)
     group_count = 0
 
     for i, file in enumerate(files, start=1):
-        # 构造输出文件夹路径
+        # Build output subfolder path
         group_folder = os.path.join(output_folder, f'group_{group_count + 1}')
 
-        # 创建输出文件夹（如果不存在）
+        # Create output subfolder if it does not exist
         os.makedirs(group_folder, exist_ok=True)
 
-        # 构造输入文件路径和输出文件路径
+        # Build input and output file paths
         input_file = os.path.join(input_folder, file)
         output_file = os.path.join(group_folder, file)
 
-        # 移动文件到输出文件夹
+        # Copy file into the output subfolder
         shutil.copy(input_file, output_file)
 
         if i % per_group_num == 0:

scDifformer_data_process/preflight/convert_data.py

@@ -25,7 +25,7 @@ gene_info_table_path = '/mnt/nfs/data/geneformer/input/material/gene_info_table.
 
 gene_info = pd.read_csv(gene_info_table_path)
 
-# 转换为字典
+# Convert to dictionary
 gene_name_id_combine_dict = gene_info.set_index('gene_name')['ensembl_id'].to_dict()
 
 input_dir = '/mnt/nfs/hs/disco_1115_all'

scDifformer_data_process/preprocess/preprocess.py

@@ -32,7 +32,7 @@ home_dir_list = ['/scDifformer/data/240412_test/raw_h5ad']
 
 gene_info = pd.read_csv(gene_info_table_path)
 
-# 转换为字典
+# Convert to dictionaries
 gene_name_id_combine_dict = gene_info.set_index('gene_name')['ensembl_id'].to_dict()
 gene_name_type_dict = gene_info.set_index('gene_name')['gene_type'].to_dict()
 gene_id_type_dict = gene_info.set_index('ensembl_id')['gene_type'].to_dict()
@@ -47,34 +47,34 @@ def preprocess(file_path: str) -> None:
         adata = sc.read_h5ad(file_path)
         adata.raw = None
         adata.X = np.round(adata.X, decimals=1)
-        # 1. 去除线粒体和总基因数within 3 s.d. of the mean
+        # 1. Filter mitochondrial ratio and total gene counts within 3 standard deviations of the mean
         adata.var['mt'] = adata.var_names.str.startswith('MT-')  # annotate the group of mitochondrial genes as 'mt'
         sc.pp.calculate_qc_metrics(adata, qc_vars=['mt'], percent_top=None, log1p=False, inplace=True)
 
-        # 计算线粒体比例的均值和标准差
+        # Compute mean and standard deviation of mitochondrial ratio
         mean_ratio_mt = np.mean(adata.obs["pct_counts_mt"])
         std_ratio_mt = np.std(adata.obs["pct_counts_mt"])
 
-        # 计算在均值三个标准差范围内的细胞
+        # Keep cells within three standard deviations of the mean
         min_ratio_mt = mean_ratio_mt - 3 * std_ratio_mt
         max_ratio_mt = mean_ratio_mt + 3 * std_ratio_mt
         within_range_mt = (adata.obs["pct_counts_mt"] >= min_ratio_mt) & (adata.obs["pct_counts_mt"] <= max_ratio_mt)
-        # 从adata中删除不在范围内的细胞
+        # Remove cells outside the valid range
         adata = adata[within_range_mt, :]
 
-        # 计算线基因总表达量的均值和标准差
+        # Compute mean and standard deviation of total gene counts
         mean_ratio = np.mean(adata.obs["total_counts"])
         std_ratio = np.std(adata.obs["total_counts"])
 
-        # 计算在均值三个标准差范围内的细胞
+        # Keep cells within three standard deviations of the mean
         min_ratio = mean_ratio - 3 * std_ratio
         max_ratio = mean_ratio + 3 * std_ratio
         within_range = (adata.obs["total_counts"] >= min_ratio) & (adata.obs["total_counts"] <= max_ratio)
 
-        # 从adata中删除不在范围内的细胞
+        # Remove cells outside the valid range
         adata = adata[within_range]
 
-        # 2. 创建DoubletFinder对象并运行双峰细胞检测
+        # 2. Create a DoubletFinder-like object and run doublet detection
         scrub = scrublet.Scrublet(adata.X)
         # Run doublet detection
         doublet_scores, predicted_doublets = scrub.scrub_doublets()
@@ -141,6 +141,6 @@ if __name__ == '__main__':
         home_path = home_dir
         files = [os.path.join(home_path, file_name) for file_name in os.listdir(home_path) if
                  str(file_name).endswith(".h5ad") and file_name not in no_need_handle_set]
-        logger.info(f'本次处理的文件个数: {len(files)}')
+        logger.info(f'Number of files processed in this run: {len(files)}')
         with mp.Pool(processes=32) as pool:
             pool.map(preprocess, files)

scDifformer_data_process/print_data_info.py

@@ -12,22 +12,22 @@
 #
 # file_path = '/scDifformer/data/240412_test/looms/zheng68k_fold0_test.loom'
 #
-# # 使用loompy连接loom文件
+# # Use loompy to connect to the loom file
 # with loompy.connect(file_path) as ds:
 #
-#     # 打印所有的列属性（column attributes, 即细胞元数据）
+#     # Print all column attributes (cell metadata)
 #     print("\nColumn:")
 #     for key, val in ds.ca.items():
 #         print(f"{key}: {val}")
 #
-#     # 打印所有的行属性（row attributes, 即基因元数据）
+#     # Print all row attributes (gene metadata)
 #     print("\nRow:")
 #     for key, val in ds.ra.items():
 #         print(f"{key}: {val}")
 #
-#     # 打印矩阵的一部分（例如，前5个基因和前5个细胞的表达量）
+#     # Print part of the matrix (e.g., expression of first 5 genes and first 5 cells)
 #     print("\nData matrix:")
-#     matrix_slice = ds[:, :5]  # 获取所有基因在前5个细胞中的表达量
+#     matrix_slice = ds[:, :5]  # Expression of all genes in the first 5 cells
 #     print(matrix_slice)
 
 # import pickle

scDifformer_data_process/tokenizer.py

@@ -127,7 +127,7 @@ class TranscriptomeTokenizer:
         i = 1
         # loops through directories to tokenize .loom files
         for loom_file_path in loom_data_directory.glob("*.loom"):
-            logger.info(f"已经处理的文件个数 {i}")
+            logger.info(f"Number of files processed: {i}")
             i += 1
             print(f"Tokenizing {loom_file_path}")
             file_tokenized_cells, file_cell_metadata = self.tokenize_file(