Two collaborations signaled rapid scaling of data and AI infrastructure in discovery research. Parse Biosciences and Tahoe Therapeutics agreed to generate a 300 million cell perturbation atlas to train next‑generation biological AI models — a project built to enable predictive virtual cell models and accelerate target identification. Tahoe’s plan follows its prior 100M dataset and broader ambitions to reach billion‑cell scale. Separately, contract research organization Inotiv partnered with Lithuanian multiomics AI firm Vugene to integrate AI‑assisted bioinformatics into early drug discovery workflows. The Inotiv–Vugene tie reflects an industry push to unite wet‑lab CRO capacity with AI platforms to improve predictive decision‑making on efficacy and safety. Collectively, these moves point to a data‑scale arms race: large, standardized perturbation datasets plus integrated AI analysis are becoming essential levers for modern translational pipelines and for de‑risking early candidates.