Parse and Tahoe to build 300M‑cell perturbation atlas to train biology AI

Parse Biosciences and Tahoe Therapeutics announced a collaboration to generate a 300‑million‑cell single‑cell perturbation atlas using Parse’s GigaLab services and Tahoe’s Mosaic perturbation platform. The dataset will contain multi‑tissue, multi‑perturbation single‑cell transcriptomes intended to train foundational AI models that predict therapeutic outcomes and virtual cell behavior. Tahoe previously produced a 100M dataset; the new effort scales that groundwork and aims to accelerate AI‑driven target discovery and phenotypic prediction. The project underscores industry momentum to create large, standardized biological datasets to underpin robust machine‑learning models.