Public Datasets

Public single-cell and spatial datasets you can pull from for practice, reanalysis, and integration with your own work — grouped by the generation technology so you can find data that matches (or deliberately differs from) the 10x Chromium data the workshop uses.

Atlases, portals & searchable repositories

Start here — these aggregate thousands of datasets across technologies and tissues.

• 10x Genomics datasets — the canonical source for Chromium (scRNA), scATAC, Multiome, Visium, and Xenium demo data, including the 20k NSCLC DTC set used in the Full Talapas run from raw FASTQs.
• CELLxGENE Discover and the CZ CELLxGENE Census — curated, queryable, standardized atlases with built-in viewers and a programmatic API.
• Human Cell Atlas Data Portal — the HCA’s cross-tissue human reference data.
• Tabula Sapiens (human) and Tabula Muris (mouse) — multi-organ atlases with both droplet and plate-based assays of the same tissues (useful for technology comparisons).
• GEO, ArrayExpress / BioStudies, and the Single Cell Portal (Broad) — searchable primary-data repositories.

Droplet-based scRNA-seq (3′ / 5′)

High-throughput, shallow-per-cell — the family the workshop’s ifnb (10x 3′ v1) belongs to.

• 10x Chromium datasets — PBMC 3k/10k, the standard teaching sets, plus 5′ and CITE-seq examples.
• Drop-seq — Macosko et al. 2015 mouse retina, the original droplet method: GEO GSE63472.
• inDrops — Klein et al. 2015 mouse embryonic stem cells: GEO GSE65525.

Plate-based / full-length scRNA-seq (Smart-seq)

Lower throughput, deep per-cell coverage with full-length transcripts — good for isoforms and rare-cell detail.

• Smart-seq2 — the plate-based half of Tabula Muris (run alongside its 10x droplet data on the same organs).
• Smart-seq3 — Hagemann-Jensen et al. 2020, full-length + UMI: ArrayExpress E-MTAB-8735.

Combinatorial-barcoding scRNA-seq (split-pool)

Instrument-free, very high throughput via successive rounds of barcoding (Parse Evercode, SPLiT-seq, sci-RNA-seq).

• SPLiT-seq — Rosenberg et al. 2018 mouse brain & spinal cord: GEO GSE110823.
• sci-RNA-seq3 — the Mouse Organogenesis Cell Atlas (MOCA), Cao et al. 2019 (~2 M cells).
• Parse Biosciences Evercode datasets — public split-pool whole-transcriptome data, incl. very large PBMC sets.

scATAC-seq & multiome (chromatin accessibility)

The assay behind Modules 12 & 15 — open-chromatin (peaks/fragments) instead of transcripts.

• 10x scATAC & Multiome datasets — incl. the PBMC 10k scATAC set used in Module 15, and joint RNA + ATAC Multiome examples.
• sci-ATAC-seq — the Mouse sci-ATAC-seq Atlas, Cusanovich et al. 2018 (~100k cells, 13 tissues).
• scATAC of human hematopoiesis — Buenrostro et al. 2018: GEO GSE96772.

Spatial transcriptomics

Expression with tissue coordinates — the family of Module 16’s stxBrain (10x Visium).

• 10x Visium & Xenium datasets — array-based (Visium) and high-plex imaging (Xenium) demo tissues.
• Slide-seqV2 — near-cellular-resolution beads, Stickels et al. 2021 (on the Single Cell Portal).
• MERFISH — imaging-based, single-molecule: the Vizgen mouse brain MERFISH datasets.
• Stereo-seq — the Mouse Organogenesis Spatiotemporal Transcriptomic Atlas (MOSTA), Chen et al. 2022.