Deep-learning methods

…His laboratory develops deep-learning methods that generate proteins with new functions, synthesizes genes encoding the designs, and experimentally characterizes the resulting molecules. The lecture explains how this design-build-test approach extends beyond modifying existing natural proteins and presents applications of computational protein design to contemporary problems. Online via the Royal S

Develop deep-learning methods for generative protein design at the interface of machine learning, molecular simulation, structural biology and experimental protein engineering. Projects can investigate conformational ensembles, integrate protein language models with physics, or design proteins for challenging membrane and sugar interfaces. The researcher can develop an independent direction while collaborating with computational and experimental colleagues. Applicants submit a CV, research-interest statement and three referee contacts to the laboratory contact listed on its recruitment page.