Expanding the immunotherapy toolkit with protein design AI

Speaker: Possu Huang, Ph.D.
Elected Co-Chair of the Gordon Research Conference (GRC) protein engineering
Assistant Professor, Department of Bioengineering at Stanford University
Co-Chair of the Keystone Symposia, Machine Learning Applied to Macromolecular Structure and Function

Abstract:
We have developed unique capabilities to design proteins that can target a multitude of protein-based markers of diseased cells: surface proteins, intracellular proteins, and processed exogenous antigens. The overall strategy we use is to develop generalizable and physics-grounded methods, including artificial intelligence/machine learning (AI/ML) tools to address these challenges. Our generative design workflow builds towards a complete Build-Measure-Learn loop, allowing the models to refine their world view through experimental feedback. In this seminar, I will describe our progress in building these methods, their application to developing “zero shot” design of cyclic peptides against cancer markers, as well as our TRACeR system that enables targeting of diseased cells broadly through their immunopeptidome
 

Speaker Bio:

Dr. Possu Huang received his PhD from Caltech with the first demonstration of a computationally designed novel protein-protein interface. He subsequently conducted postdoctoral research at the University of Washington before starting his group at Stanford. His research focuses on advancing the understanding of proteins for the engineering of novel therapeutics and other protein-based nanotechnology. His group uses machine learning, computational modeling, structural biology, and experimental library optimization to continue the expansion of protein-based molecular platforms.

Host: Qian Cong, Ph.D.