Overview
What It Is
Chan Zuckerberg CELLxGENE Discover is a tool to find, download, and visually explore curated and standardized single-cell biology datasets. It is built and maintained by the Chan Zuckerberg Initiative (CZI) in collaboration with the Wellcome Sanger Institute and Stanford University.
The platform is open-source and free to use. Contributions of data are welcome from the scientific community and are not confined to a single consortium or funder. It serves as a centralized hub that promotes collaboration across researchers, labs, and consortia. CELLxGENE is differentiated from other single-cell data portals by its enforcement of a standardized schema for gene, cell, assay, and donor metadata that evolves to address contributor requirements.
Scale
As of October 1, 2024, the platform hosts over 1,550 datasets and 169.3 million cells (93.6 million unique cells).
CZ CELLxGENE has been adopted as a primary data-sharing platform for small research labs and large consortia like the BRAIN Initiative, the Human Tumor Atlas Network (HTAN), and the Human Cell Atlas (HCA).
Key Features
1. CELLxGENE Explorer — an interactive no-code visual interface for exploring datasets, filtering by metadata, and identifying differentially expressed genes between user-defined cell groups.
2. CELLxGENE Census — an API and data object hosted by CZ CELLxGENE that enables efficient access and custom slices of the corpus for programmatic access and computational use cases, supporting corpus-wide views, summaries, and meta-analyses of tens of millions of cells across studies.
The Census provides fast, efficient access to the largest corpus of standardized single-cell data via TileDB-SOMA, enabling researchers to access slices of data from more than 500 single-cell datasets spanning about 33 million unique cells across more than 60,000 genes from human or mice, with standardized cell and gene metadata and harmonized labels. Data can be easily loaded into Scanpy or Seurat.
3. Foundation Models — models trained on CZ CELLxGENE Census data are available and regularly retrained as the data corpus grows.
4. Spatial Data — the Census now includes data from Slide-seq and Visium spatial assays, integrating spatial transcriptomics alongside single-cell RNA-seq.
Notes
Filter by 'Ovarian Carcinoma' or 'Endometrial Carcinoma' disease tags. Regner 2021 and Deng 2022 HGSOC datasets cataloged here.