cellxgene-census

Query CELLxGENE Census using open_soma to access millions of cells across datasets. Filter by tissue, disease, and cell type; retrieve expression matrices and embeddings; then load into scanpy or PyTorch for analysis.

Yes. CELLxGENE Census supports filtering by cell type, tissue, and disease through metadata APIs. Explore available datasets and their characteristics before retrieving expression data for your analysis.

Load CELLxGENE Census expression data as AnnData objects using get_anndata, then integrate directly into scanpy pipelines or convert for PyTorch machine-learning workflows. Supports cross-dataset analysis and out-of-core processing.

CELLxGENE Census provides get_anndata, axis_query, and metadata APIs for flexible data retrieval. Context-managed opening and versioning enable reproducible analyses across small-to-large queries without loading entire datasets into memory.

Yes. CELLxGENE Census supports out-of-core analyses for large-scale population studies. Use axis_query and metadata APIs to retrieve subsets by cell type, tissue, or disease, enabling analyses across millions of cells without memory constraints.

Yes. CELLxGENE Census integrates seamlessly with AnnData and scanpy. Retrieve data as AnnData objects via get_anndata, then apply standard single-cell analysis, visualization, and machine-learning workflows directly.