Cancer3.AI › AI in Oncology › AI Models › scGPT
Foundation model Open weights Not applicable genomics

scGPT

Generative pretrained transformer for single-cell multi-omics, trained on over 33 million cells, supporting cell-type annotation, batch integration, perturbation response prediction and gene-network inference.

GitHub Paper DOI { } JSON

At a glance

DeveloperBo Wang Lab, University of Toronto
Released2024-02-08
LicenceMIT
AvailabilityOpen weights
KindFoundation model
Regulatory statusNot applicable

What it does

Not yet documented on this card.

Tasks, data types and cancers

TaskSingle-cell analysisGene expressionFeature extraction (embeddings)
Data typeSingle-cellTranscriptomics (RNA)
Clinical settingBasic research
CancerPan-cancer
Inputsingle-cell expression matrix (h5ad)
Outputcell/gene embeddings; predicted expression

Architecture

FamilyGPT-style transformer with gene and expression-value tokens
Pre-traininggenerative masked prediction of expression values

Training data

33 million cells from CELLxGENE and other public collections; organ- and tissue-specific checkpoints (including a pan-cancer checkpoint) are released.

Training set size33M cells

Evaluation

Not yet documented on this card.

How to run

LibraryPyTorch (scgpt package)

Checkpoints (whole-human, pan-cancer, per-organ) are linked from the repo README; expects scanpy AnnData input.

Regulatory status and intended use

Regulatory statusNot applicable
Intended useResearch tool.

Source →

Regulatory status is quoted from the source linked above and can change. Research-use-only models must not be used for clinical decisions.

Limitations and bias

  • Perturbation predictions are validated mainly on Perturb-seq lines, not patient tumours.
  • GPU memory scales with gene vocabulary; subsample genes for large atlases.

Sources

  1. Cui H et al. scGPT: toward building a foundation model for single-cell multi-omics using generative AI. Nat Methods 2024
  2. GitHub — bowang-lab/scGPT

This page is educational — it is not medical advice and does not replace consultation with an oncologist. Diagnostic and treatment decisions are made solely by specialist physicians.