Tool to evaluate drug induced molecular phenotypes in 100M single-cell transcriptomics
The researchers developed Tahoe-100M, a giga-scale single-cell perturbation atlas capturing 100 million transcriptomes from 50 diverse cancer cell lines treated with ~1,100 drug-dose conditions.
By pairing transcriptional outputs with systemic cellular behaviors, Tahoe-100M offers an open-source sandbox for training AI frameworks to model drug-response network dynamics.
The tool helps to analyze shared pathway signatures, and can differentiate a drug's primary therapeutic mechanism from its unintended collateral pathways.
When combined with functional genetic maps (such as CROP-seq databases), Tahoe-100M allows scientists to compare chemical modifications head-to-head against clean genetic knockouts, accelerating the identification of precise phenocopying leads.
