Parse Biosciences and Tahoe Therapeutics announced a collaboration to generate a 300‑million‑cell perturbation atlas—expanding on Tahoe’s prior Tahoe‑100M dataset—to train AI models for predicting therapeutic responses. Parse will run single‑cell GigaLab sequencing on perturbation panels spanning tissues and disease states, feeding foundational datasets for virtual cell models. Concurrently, major sequencing firms reiterated multiomics roadmaps amid funding headwinds, highlighting platforms for spatial, proteomic and integrated genomics to support translational studies. The pair of moves signals industry momentum behind large, high‑quality biological datasets to power AI drug discovery and clinical decision tools, even as customers weigh costs and regulatory paths.