Make tutorial for how to import data from H5ad/loom(/Seurat)

[pascaltimshel]

Make a tutorial that demonstrates how to import data from commonly used single cell data formats into CELLEX:

• H5ad
• loom
• (Seurat)

[mohamednhassan]

adata = sc.read_h5ad('adata.h5ad')
adata.obs['cellID'] = adata.obs_names

adata.write_loom(filename='data.loom')

import loompy
import pandas as pd

with loompy.connect('data.loom') as ds:
# Use 'var_names' for gene names (row identifiers)
rows = ds.row_attrs["var_names"]

# Use 'obs_names' for cell IDs (column identifiers)
cols = ds.col_attrs["obs_names"]

# Expression matrix
data = pd.DataFrame(ds[:, :], index=rows, columns=cols)

# Use appropriate annotation columns — here are some options you might want:
metadata = pd.DataFrame({
    "cell_type": ds.col_attrs["celltype"]  # or try 'annotation_heatmap' if more relevant
}, index=cols)

metadata_class = pd.DataFrame({
    "cell_class": ds.col_attrs["celltype2"]  # you can swap this with a more suitable one if needed
}, index=cols)

eso = cellex.ESObject(data=data, annotation=metadata, verbose=True)
Then you can continue using the tutorial provided with CELLEX