Accelerated Discovery of Carbamate Cbl-b Inhibitors Using Generative AI Models and Structure-Based Drug Design
Taylor R. Quinn, Kathryn A. Giblin, Clare Thomson, Jeffrey A. Boerth, Gayathri Bommakanti, Erin Braybrooke, Christina Chan, Alex J. Chinn, Erin Code, Caifeng Cui, Yukai Fan, Neil P. Grimster, Keishi Kohara, Michelle L. Lamb, Lina Ma, Adelphe M. Mfuh, Graeme R. Robb, Kevin J. Robbins, Marianne Schimpl, Haoran Tang, Jamie Ware, Gail L. Wrigley, Lin Xue, Yun Zhang, Huimin Zhu, Samantha Jane Hughes
Abstract
Casitas B-lymphoma proto-oncogene-b (Cbl-b) is a RING finger E3 ligase that has an important role in effector T cell function, acting as a negative regulator of T cell, natural killer (NK) cell, and B cell activation. A discovery effort toward Cbl-b inhibitors was pursued in which a generative AI design engine, REINVENT, was combined with a medicinal chemistry structure-based design to discover novel inhibitors of Cbl-b. Key to the success of this effort was the evolution of the "Design" phase of the Design-Make-Test-Analyze cycle to involve iterative rounds of an in silico structure-based drug design, strongly guided by physics-based affinity prediction and machine learning DMPK predictive models, prior to selection for synthesis. This led to the accelerated discovery of a potent series of carbamate Cbl-b inhibitors.