cellxgene-census

You can query single-cell genomics data from the CZ CELLxGENE Census programmatically using standardized metadata filters to retrieve specific cell types and expression matrices without requiring local storage.

Yes, cellxgene-census supports integration with scanpy for advanced bioinformatics analysis and PyTorch for machine learning applications on single-cell genomics datasets.

Yes, you can analyze population-scale single-cell data without high-performance computing clusters by using out-of-core processing capabilities to handle multi-terabyte datasets efficiently.

You can retrieve an AnnData object containing specific cell types from human lung tissue by applying metadata-driven filters to query the human and mouse datasets available in the Census.

The best way to access large-scale single-cell expression matrices for cross-study analysis is by using programmatic Census queries to pull filtered subsets across thousands of datasets instantly.

Yes, tiledbsoma and pyarrow are required dependencies alongside cellxgene-census, scanpy, and torch to enable programmatic access and out-of-core processing of single-cell data.