Tahoe Therapeutics, the Arc Institute and the Biohub committed multimillion‑dollar funding to build a public, perturbation‑heavy single‑cell dataset intended to accelerate virtual cell model development. Tahoe will produce more than 120 million single‑cell data points across roughly 225,000 drug–patient perturbations; partners said the new resource will expand diversity of perturbations and patient contexts beyond existing datasets. The dataset will be shared openly after an exclusive period and is intended to provide the training data gap hindering predictive cellular models. The collaboration aims to standardize perturbation experiments and provide the scale needed for AI models to predict transcriptomic responses to drugs and genetic perturbations.