Parse Biosciences and Graph Therapeutics launched a strategic partnership to build a large, functional immune perturbation atlas by combining Graph’s lab‑in‑the‑loop disease models with Parse’s GigLab single‑cell technology. The program aims to profile hundreds of millions of patient‑derived immune cells under systematic perturbations to map context‑specific immune behaviors and accelerate AI‑first drug discovery. Graph will select conditions via active learning and patient‑derived assays; Parse will generate massive single‑cell datasets using Evercode™ technology. Gregory Vladimer, PhD, CEO of Graph, described the collaboration as closing the gap between prediction and clinical reality by investing in biological context to derisk target selection and prevent late‑stage failures.