You are here

MIA Talks

Learning to read and write protein evolution

February 1, 2023
Stanford University School of Medicine

Evolution is the powerful force driving the real-time emergence of pathogen resistance to drugs and immunity, as well as the diversity of natural forms and functions that have emerged over longer timescales. Modern evolutionary models, especially those that leverage advances in machine learning, can improve both our ability to understand evolution as it happened in the past and our ability to artificially design new proteins in the laboratory. First, this talk will cover how models of sequence evolution called protein language models can learn evolutionary rules that predict the fitness effects of mutations and can also predict the directionality of evolution in phylogenetic landscapes spanning both decades and geologic eons. Next, this talk will cover how protein language models, as well as models that learn the evolutionary rules of protein structure, can guide artificial evolution, including the affinity maturation of antibodies against diverse viral antigens and the in-silico evolution of modular de novo proteins with structures not found in nature.