scPerturb

Harmonized single-cell perturbation datasets (genetic + chemical) with standardized metadata.

Composite

88.9

Experimental validation

Retrospective

Stages

Virtual CellTarget ID

Modalities

cross-modality

Task types

perturbation-prediction

Size

cells: 1,700,000
datasets: 44

License

MIT

First release

2023-05

Last updated

2024-11

Official site

→ project page

Leaderboard

→ leaderboard

Dataset

→ dataset

Code / GitHub

→ repository

HuggingFace

→ HF

Paper

scPerturb: Harmonized single-cell perturbation data · Peidli S, Green TD, Shen C, Theis FJ, et al. · 2024 · paper · doi:10.1038/s41592-023-02144-y · 280 citations

Flags

none

Experts

Chris Sander, Fabian Theis

Groups

Sander Lab (Harvard/DFCI), Helmholtz Munich (Theis Lab)

Hosted by

—

Related benchmarks

CZ Virtual Cell Challenge, Open Problems: Perturbation Prediction, LINCS L1000 / CMap

Rubric (7-criterion)

rigor

coverage

maintenance

adoption

quality

accessibility

industry_relevance

Notes

Canonical harmonized resource. Strong Perturb-seq coverage; weaker for chemical perturbations.

← Back to all benchmarks

Compare:

Open comparison →