Perturb-Seq Data Alliance

Perturb-seq datasets, harmonized once, used by every member, forever

Every pharma company and academic center is independently sourcing, cleaning, and re-processing the same public perturb-seq studies. The Rancho Biosciences Perturb-Seq Alliance is a pre-competitive, shared-cost community that does it once. Members get better data, at lower cost, sooner.

Become a member

Why Perturbation Data?

Atlases show correlation. Perturb-Seq shows cause and effect.

Genetic or chemical perturbations are applied to cells, then single-cell RNA sequencing reads out exactly how each cell responded. That link between a specific perturbation and a specific transcriptional response is what makes the data useful.

Causal Relationships

Knockdowns and knockouts reveal cause-and-effect, not just correlation from observational atlases.

Functional Annotation

Every perturbation is a controlled experiment that teaches gene function and regulatory network structure.

Counterfactual Learning

Models learn "what happens if" scenarios — the basis for prediction and intervention design.

Richer Training Signal

Perturbation identity plus cellular response is more structured supervision than unlabeled atlas data alone.

What Membership Includes

Harmonized once, delivered in the formats your team already uses.

Uniform processing, a consistent model, and harmonized metadata turn scattered public studies into a single, analysis-ready resource

Perturb-seq data catalog

An updated, living list of available datasets with key metadata categories.

Extended data model

Organizes perturbation concepts with consistent, agreed-upon terminology.

Up to 20 datasets you designate

Plus every other dataset selected by fellow members — the list only grows.

Perturb-Seq Visualization App

Unlocked once the Alliance reaches three members.