Illumina unveiled a large-scale genome‑wide perturbation dataset intended to accelerate AI-driven drug discovery and target validation. The company said the "Billion Cell Atlas" will be built across multiple years to generate controlled genetic perturbation data at scale and to help pharma train foundation models. AstraZeneca, Merck and Eli Lilly are participating as founding partners. Illumina presented early data and commercial plans at JPM, stressing that the atlas will enable mechanistic interrogation of genetic signals in hundreds of cell lines and disease contexts. The resource is pitched to validate targets, train virtual‑cell models and reduce reliance on lower-resolution or literature‑based training data. The project shifts Illumina from pure sequencing infrastructure toward providing curated, perturbation‑heavy datasets as an input to AI models — a move that could reshape early discovery workflows by coupling large experimental atlases with model training and commercial collaborations.