insitro vs Valo Health (2026)

insitro and Valo Health make the same opening bet: drug discovery should start from causal human biology rather than from chemistry. Both mine human data for disease drivers before committing to a molecule, both run their own programs alongside large pharmaceutical partnerships, and neither has yet put a platform derived candidate into a clinical study. The difference is the evidence each uses to back a target. insitro pairs population scale genetics with more than 20 petabytes of its own cellular experiments, built from induced pluripotent stem cells, genome editing and high content phenotyping, so a target found by its Virtual Human engine is tested in cell models of the disease. Its TherML engine then designs across modalities, and the lead program, CTRO-1013, is an siRNA against IRS1 for liver fat expected to reach the clinic in early 2027. Bristol Myers Squibb paid $50 million upfront for an ALS collaboration with a potential value over $2.1 billion. Valo works from patient records rather than cell lines: histories mapped into subtypes with AI and knowledge graphs and ranked with causal inference and statistical genetics. Its chemistry is small molecules only, made in a closed loop where models are refit after each laboratory round, and Novo Nordisk's cardiometabolic collaboration spans up to twenty programs. A partner that wants targets checked in engineered human cell models, or needs a modality beyond small molecules, fits insitro. A partner whose disease is best read in long patient histories, and who wants small molecule chemistry in the same loop, fits Valo.