Valence Labs is Recursion's AI research engine, developing machine learning and computational methods to predict, explain, and decode biology for therapeutic discovery. The group operates from offices in Montreal and London, leveraging Recursion's platform, data, and computing infrastructure alongside deep ties to Mila, the deep-learning research institute.
The team builds three categories of tools. Virtual Cells are mechanistic models of cellular function designed to guide discovery of novel therapeutics. Lab-in-the-loop discovery engines integrate laboratory experiments with computational models to generate novel therapeutic hypotheses. Machine learning systems convert biological and chemical data into predictive, mechanistic, and generative models of living systems, enabling prediction of cellular responses to perturbations and explanation of molecular interactions.
Valence Labs publishes research in top-tier journals and conferences, reflecting its orientation toward advancing real-world drug discovery outcomes for complex diseases. The combination of rigorous computational methods, experimental integration, and access to large-scale biological data positions the group at the intersection of academic research and pharmaceutical application.