About
Dramatic recent progress in both artificial intelligence and structural biology has created tremendous opportunities for using machine learning methods not only to predict three-dimensional structures of drug targets but also to design safer, more effective drugs. Since 2015, multiple research groups at Stanford have been developing machine learning methods to leverage structure for the design of both biologics and small-molecule therapeutics. Current research directions include prediction of ligand binding poses, affinities, functional effects, and off-target properties; virtual screening; generative models for drug candidates; methods to achieve selectivity; and design of antibodies to optimize their developability.
The Artificial Intelligence for Structure-Based Drug Discovery program provides opportunities for exchange of ideas between Stanford researchers developing groundbreaking machine learning methods that leverage molecular structure and industry scientists who wish to apply such methods to bring better drugs to the market efficiently. In order to maximize the real-world impact of their research, Stanford researchers welcome input from industry partners—for example, on which problems to tackle or which software features to add. Industry partners also benefit through exposure to cutting-edge research, a forum to ask questions about deployment of algorithms and software, and opportunities to network with both Stanford researchers and other industry partners.
The Team
AISBDD includes multiple Stanford professors and their research groups. Founding faculty are:

Prof. Ron Dror (Computer Science and, by courtesy, Molecular and Cellular Physiology and Structural Biology)

Prof. Russ Altman (Bioengineering, Genetics, Biomedical Data Science, Medicine and, by courtesy, Computer Science)

Prof. Possu Huang (Bioengineering)
All three have extensive experience in machine learning for structural biology and drug discovery. Their work in this area includes structure prediction for proteins, RNA, and target-ligand complexes; protein design; virtual screening; atomic-level simulation to guide drug discovery; toxicity prediction; and effects of genetic variation on drug response. They collaborate with a wide variety of experimentalists and companies.