Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: TypeError
Message: Couldn't cast array of type
struct<ome: struct<omero: struct<channels: list<item: struct<label: string>>>, version: string, multiscales: list<item: struct<datasets: list<item: struct<path: string, coordinateTransformations: list<item: struct<type: string, scale: list<item: double>>>>>, name: string, axes: list<item: struct<name: string, type: string>>, coordinateTransformations: list<item: struct<type: string, transformations: list<item: struct<type: string, scale: list<item: double>, input: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, output: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, translation: list<item: double>>>, input: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, output: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>>>>>>, spatialdata_attrs: struct<version: string>>
to
{'spatialdata_attrs': {'version': Value('string'), 'spatialdata_software_version': Value('string')}, 'atera': {'dataset': Value('string'), 'chemistry_version': Value('string'), 'frame': Value('string'), 'units': Value('string'), 'he_px_um': Value('float64'), 'coordinate_system': Value('string'), 'crop_window': {'frame': Value('string'), 'units': Value('string'), 'he_px_um': Value('float64'), 'size_um': Value('float64'), 'stride_um': Value('float64'), 'x0_um': Value('float64'), 'y0_um': Value('float64'), 'x1_um': Value('float64'), 'y1_um': Value('float64'), 'he_px': {'x0': Value('int64'), 'y0': Value('int64'), 'x1': Value('int64'), 'y1': Value('int64')}, 'n_cells': Value('int64'), 'entropy': Value('float64'), 'entropy_floor_used': Value('float64'), 'entropy_floor_config': Value('float64'), 'n_types': Value('int64'), 'composition': {'Tumour epithelial': Value('int64'), 'Proliferating tumour': Value('int64'), 'Myoepithelial': Value('int64'), 'Fibroblast': Value('int64'), 'Endothelial': Value('int64'), 'Perivascular': Value('int64'), 'T cell': Value('int64'), 'Plasma cell': Value('int64'), 'Macrophage': Value('int64'), 'Dendritic cell': Value('int64'), 'Mast': Value('int64'), 'Mixed (plasma+mast)': Value('int64')}, 'composition_frac': {'Tumour epithelial': Value('float64'), 'Proliferating tumour': Value('float64'), 'Myoepithelial': Value('float64'), 'Fibroblast': Value('float64'), 'Endothelial': Value('float64'), 'Perivascular': Value('float64'), 'T cell': Value('float64'), 'Plas
...
': Value('int64'), 'Mixed (plasma+mast)': Value('int64')}, 'n_cells_built': Value('int64')}, 'selection_rule': Value('string'), 'n_windows_scanned': Value('int64'), 'n_windows_passing_floor': Value('int64'), 'generated_by': Value('string'), 'generated_at': Value('timestamp[s]'), 'n_cells_built': Value('int64')}, 'he_overview_um_px': Value('float64'), 'he_zoom_um_px': Value('float64'), 'panel_genes': List(Value('string')), 'cell_type_markers': {'Tumour epithelial': List(Value('string')), 'Proliferating tumour': List(Value('string')), 'Myoepithelial': List(Value('string')), 'Fibroblast': List(Value('string')), 'Endothelial': List(Value('string')), 'Perivascular': List(Value('string')), 'T cell': List(Value('string')), 'NK': List(Value('string')), 'B cell': List(Value('string')), 'Plasma cell': List(Value('string')), 'Macrophage': List(Value('string')), 'Dendritic cell': List(Value('string')), 'Mast': List(Value('string'))}, 'cell_type_colors': {'Tumour epithelial': Value('string'), 'Proliferating tumour': Value('string'), 'Myoepithelial': Value('string'), 'Fibroblast': Value('string'), 'Endothelial': Value('string'), 'Perivascular': Value('string'), 'T cell': Value('string'), 'NK': Value('string'), 'B cell': Value('string'), 'Plasma cell': Value('string'), 'Macrophage': Value('string'), 'Dendritic cell': Value('string'), 'Mast': Value('string'), 'Unassigned': Value('string')}, 'read_instructions': Value('string'), 'built_by': Value('string'), 'built_at': Value('timestamp[s]')}}
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2312, in cast_table_to_schema
cast_array_to_feature(
~~~~~~~~~~~~~~~~~~~~~^
table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
feature,
^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1861, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2158, in cast_array_to_feature
raise TypeError(f"Couldn't cast array of type\n{_short_str(array.type)}\nto\n{_short_str(feature)}")
TypeError: Couldn't cast array of type
struct<ome: struct<omero: struct<channels: list<item: struct<label: string>>>, version: string, multiscales: list<item: struct<datasets: list<item: struct<path: string, coordinateTransformations: list<item: struct<type: string, scale: list<item: double>>>>>, name: string, axes: list<item: struct<name: string, type: string>>, coordinateTransformations: list<item: struct<type: string, transformations: list<item: struct<type: string, scale: list<item: double>, input: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, output: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, translation: list<item: double>>>, input: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>, output: struct<name: string, axes: list<item: struct<name: string, type: string, unit: string>>>>>>>>, spatialdata_attrs: struct<version: string>>
to
{'spatialdata_attrs': {'version': Value('string'), 'spatialdata_software_version': Value('string')}, 'atera': {'dataset': Value('string'), 'chemistry_version': Value('string'), 'frame': Value('string'), 'units': Value('string'), 'he_px_um': Value('float64'), 'coordinate_system': Value('string'), 'crop_window': {'frame': Value('string'), 'units': Value('string'), 'he_px_um': Value('float64'), 'size_um': Value('float64'), 'stride_um': Value('float64'), 'x0_um': Value('float64'), 'y0_um': Value('float64'), 'x1_um': Value('float64'), 'y1_um': Value('float64'), 'he_px': {'x0': Value('int64'), 'y0': Value('int64'), 'x1': Value('int64'), 'y1': Value('int64')}, 'n_cells': Value('int64'), 'entropy': Value('float64'), 'entropy_floor_used': Value('float64'), 'entropy_floor_config': Value('float64'), 'n_types': Value('int64'), 'composition': {'Tumour epithelial': Value('int64'), 'Proliferating tumour': Value('int64'), 'Myoepithelial': Value('int64'), 'Fibroblast': Value('int64'), 'Endothelial': Value('int64'), 'Perivascular': Value('int64'), 'T cell': Value('int64'), 'Plasma cell': Value('int64'), 'Macrophage': Value('int64'), 'Dendritic cell': Value('int64'), 'Mast': Value('int64'), 'Mixed (plasma+mast)': Value('int64')}, 'composition_frac': {'Tumour epithelial': Value('float64'), 'Proliferating tumour': Value('float64'), 'Myoepithelial': Value('float64'), 'Fibroblast': Value('float64'), 'Endothelial': Value('float64'), 'Perivascular': Value('float64'), 'T cell': Value('float64'), 'Plas
...
': Value('int64'), 'Mixed (plasma+mast)': Value('int64')}, 'n_cells_built': Value('int64')}, 'selection_rule': Value('string'), 'n_windows_scanned': Value('int64'), 'n_windows_passing_floor': Value('int64'), 'generated_by': Value('string'), 'generated_at': Value('timestamp[s]'), 'n_cells_built': Value('int64')}, 'he_overview_um_px': Value('float64'), 'he_zoom_um_px': Value('float64'), 'panel_genes': List(Value('string')), 'cell_type_markers': {'Tumour epithelial': List(Value('string')), 'Proliferating tumour': List(Value('string')), 'Myoepithelial': List(Value('string')), 'Fibroblast': List(Value('string')), 'Endothelial': List(Value('string')), 'Perivascular': List(Value('string')), 'T cell': List(Value('string')), 'NK': List(Value('string')), 'B cell': List(Value('string')), 'Plasma cell': List(Value('string')), 'Macrophage': List(Value('string')), 'Dendritic cell': List(Value('string')), 'Mast': List(Value('string'))}, 'cell_type_colors': {'Tumour epithelial': Value('string'), 'Proliferating tumour': Value('string'), 'Myoepithelial': Value('string'), 'Fibroblast': Value('string'), 'Endothelial': Value('string'), 'Perivascular': Value('string'), 'T cell': Value('string'), 'NK': Value('string'), 'B cell': Value('string'), 'Plasma cell': Value('string'), 'Macrophage': Value('string'), 'Dendritic cell': Value('string'), 'Mast': Value('string'), 'Unassigned': Value('string')}, 'read_instructions': Value('string'), 'built_by': Value('string'), 'built_at': Value('timestamp[s]')}}Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
ASI-FIMSA 2026 spatial omics workshop — staged data
Small, Colab-sized artifacts derived from the public 10x Genomics Atera whole-transcriptome Xenium preview dataset of FFPE human breast cancer, prepared for a two-hour hands-on workshop at the ASI-FIMSA meeting.
Notebooks and build scripts: https://github.com/xiao233333/ASI-FIMSA-workshop-2026
Attribution
Derived from Preview Data: Atera In Situ Gene Expression, FFPE Human Breast Cancer. Data © 10x Genomics, used under CC BY 4.0. These files are modified from the original — see What changed below.
Files
| file | size | contents |
|---|---|---|
atera_wholeslide_cells.h5ad |
19.9 MB | All 170,057 cells. Aligned centroids in obsm['spatial'], vendor graphclust and kmeans_10, obsm['X_umap'], obsm['X_pca'], our cell_type labels, and a 69-gene marker counts matrix. No images. |
atera_crop.zarr.zip |
18.0 MB | A SpatialData store for a 2,000 µm window: he (3, 1826, 1826), he_zoom (3, 1096, 1096) at native 0.2738 µm/px, 16,006 cell boundary polygons, and a (16006, 69) table. One "global" coordinate system, in micrometres. |
atera_crop_lr.h5ad |
5.4 MB | The same 16,006 Crop cells as atera_crop.zarr.zip, with the same cell_type labels and the same obsm['spatial'], but carrying a different panel: raw sparse counts for 1,673 connectomeDB2020 ligands and receptors that Atera measures. Built for ligand–receptor analysis, which the 69-gene panel cannot support. No images. |
cci_panel_genes.txt |
— | The 1,673 gene symbols in that panel, and how many complete ligand–receptor pairs each database contributes. |
atera_cluster_annotation.csv |
— | The 34 vendor clusters → 12 named cell types, with scores, margins and low_confidence flags. |
crop_window.json |
— | Exactly which window was chosen and why, including its cell-type composition. |
manifest.json |
— | Sizes, sha256 checksums and origins. |
Loading
from huggingface_hub import hf_hub_download
import anndata as ad, spatialdata as sd, zipfile, pathlib
REPO = "xiao233333/asi-fimsa-workshop-2026"
adata = ad.read_h5ad(hf_hub_download(REPO, "atera_wholeslide_cells.h5ad", repo_type="dataset"))
z = hf_hub_download(REPO, "atera_crop.zarr.zip", repo_type="dataset")
with zipfile.ZipFile(z) as f: # unzip first -- see the note below
f.extractall("atera_crop")
sdata = sd.read_zarr(next(pathlib.Path("atera_crop").glob("*.zarr")))
# The ligand-receptor view of the same cells.
lr = ad.read_h5ad(hf_hub_download(REPO, "atera_crop_lr.h5ad", repo_type="dataset"))
lr.obs["imagecol"] = lr.obsm["spatial"][:, 0] # stlearn reads positions from
lr.obs["imagerow"] = lr.obsm["spatial"][:, 1] # these, not from obsm["spatial"]
spatialdata.read_zarr() cannot open a .zarr.zip in place (its store resolver handles
LocalStore/FsspecStore and then wants store.root, which ZipStore lacks). Extract first.
Why there are two views of the same 16,006 cells
atera_crop.zarr.zip carries a 69-gene panel chosen to name cell types. Intersected with
connectomeDB2020 that leaves four complete literature-supported ligand–receptor pairs, so
no interaction analysis is possible on it. Atera is whole-transcriptome, so the genes were
measured all along — atera_crop_lr.h5ad carries 1,673 of them and supports 2,168 pairs.
The two files describe identical cells in identical coordinates and can be joined on
obs['cell_id'].
What changed from the original
The original bundle is ~65 GB. From it we produced ~38 MB by: subsetting to one 2,000 µm spatial window for the imaging artifact; restricting the expression matrix to a 69-gene marker panel; downsampling the H&E overview; and adding cell-type labels. Coordinates are in the H&E-aligned frame (0.27377667386671845 µm/px), not the Xenium instrument frame — the two differ in scale and by roughly a 90° rotation, and are not interconvertible.
Caveats — please read before reusing
- This is pre-release data. 18,028-target "Human WTA (pre-release)" panel,
Atera v1chemistry, Gen2 prototype instrument. 10x notes preview data may be superseded. - The cell-type labels are ours, not 10x's, and they are teaching material rather than
a reference annotation. They were derived automatically from the vendor's per-cluster
differential expression and then reviewed. 14 of 34 clusters carry
low_confidence; one isMixed (plasma+mast)and three areUnassignedbecause the evidence did not support a call. - B cells are not resolved as their own population. 9.75% of cells express at least one
of
MS4A1/CD79A/CD79B/BANK1, but they sit inside the large lymphocyte cluster labelledT cellrather than forming their own. All four genes are in the panel, so the question can be re-asked. - The 69-gene panel is a marker panel chosen for teaching. For anything quantitative, go back to the full 18,028-gene matrix in the original dataset.
- Downloads last month
- 235