Invited Talk
in
Workshop: AI for Drug Discovery: Bridging the Translation Gap

OpenADMET - Raising the bar on ADMET prediction with open science

Pat Walters

Abstract

Drug discovery programs routinely stall at the "Avoid-ome," the complex array of off-target interactions and ADMET liabilities that conventional heuristic screens fail to resolve. To address this bottleneck, OpenADMET establishes an integrated framework that combines high-throughput experimentation, structural biology, and machine learning. Rather than relying on retrospective evaluations of biased datasets, the initiative prospectively generates standardized, high-quality experimental data to anchor community-wide blind benchmarks. These challenges evaluate machine learning models against authentic multiparameter optimization and medicinal chemistry tasks in real time. By uniting academic and industrial researchers through open data, transparent benchmarks, and scientific forums, OpenADMET provides an open-science foundation for rigorously assessing predictive tools, derisking therapeutic pipelines, and establishing reproducible standards across the field.

Speaker

Pat Walters

Pat Walters is the Chief Scientist at OpenADMET, an open-science initiative integrating high-throughput experimentation, structural biology, and machine learning to advance ADMET prediction. Alongside his work at OpenADMET, he serves as an Adjunct Professor in the Department of Bioengineering & Therapeutic Sciences at UCSF. Pat brings over thirty years of industry leadership to his roles, having previously held senior positions at Relay Therapeutics and Vertex Pharmaceuticals. A dedicated educator and communicator, he is the author of the widely read blog “Practical Cheminformatics,” where he shares technical insights on computational drug discovery. His contributions to the field were recognized by the American Chemical Society with the 2023 Herman Skolnik Award for Chemical Information Science. Pat is a co-author of Deep Learning for the Life Sciences (O’Reilly, 2019) and serves on the editorial advisory boards for the Journal of Chemical Information and Modeling and Artificial Intelligence in the Life Sciences. He earned his Ph.D. in Organic Chemistry from the University of Arizona and his B.S. in Chemistry from UC Santa Barbara.

Chat is not available.