Unsupervised Viral Antibody Escape Prediction for Future-Proof Vaccines - Lecture
Explore a comprehensive lecture on viral antibody escape prediction and protein language models for fitness prediction and design. Delve into EVEscape, a novel model for quantifying viral escape potential of mutations at scale, applicable before surveillance sequencing or experimental data are available. Learn how this approach can enhance pandemic preparedness by predicting immune-evasive viral mutations and designing future-proof vaccines. Discover hybrid strategies that combine alignment-based and large language models to improve protein sequence fitness landscape modeling. Gain insights into the challenges and advancements in predicting viral evolution, evaluating vaccines proactively, and applying these techniques to various viruses, including SARS-CoV-2.