Philip Kim: Machine Learning Methods for Protein, Peptide and Zinc Finger Design

Join us for an insightful talk on the rapid advancements in AI-driven biologics discovery. Our speaker discusses the transition from groundbreaking achievements in 2020 to the current state where these innovations have become almost trivial. He delves into the work at Fable, a company pushing the boundaries of antibody therapeutics through cutting-edge diffusion models. Learn how these models outperform traditional benchmarks in structure and sequence design, leading to high success rates in lab-tested antibody candidates.
The talk also explores the future of computational and AI-driven biologics, emphasizing the shift towards dynamic modeling and conformational energy landscape prediction. Discover how machine learning, particularly using GANs and energy landscapes, is revolutionizing the field by predicting entire energy landscapes in one go.
Moreover, the speaker introduces PepFlow, a novel approach leveraging hyper-networks for precise confirmation predictions, significantly outperforming traditional methods like molecular dynamics (MD). Finally, the session touches on the exciting potential of zinc finger proteins in gene regulation, highlighting their advantages over CRISPR for in vivo therapeutic applications.
Don't miss this deep dive into the intersection of AI and biotechnology, showcasing how innovative approaches are shaping the future of drug discovery and gene editing.