insitro is a South San Francisco drug discovery company that builds machine learning models over data it generates itself, pairing high-throughput cell biology and automated microscopy with human genetic and clinical cohorts. Its stated aim is to find causal human biology — targets and patient segments supported by genetics rather than by association alone — and to carry the resulting programs into small-molecule and biologic therapeutics.
The company's small-molecule work is organized under its TherML platform, which spans in silico screening, affinity models trained on adaptive DNA-encoded libraries, active-learning medicinal chemistry design, and property prediction. insitro also works through data partnerships with large pharmaceutical companies, including a collaboration with Eli Lilly and Company announced in September 2025 that trained in vivo pharmacokinetics models on Lilly's decades-long preclinical corpus.
Predicts in vivo pharmacokinetics from structure alone: plasma clearance, steady-state volume, and residence time in four preclinical species.