Title: Computational structural biology in the era of deep learning

Speaker: Dr. John Moult, Institute for Bioscience and Biotechnology Research

Host: Dr. Frank Robb

Abstract: The problem of computing protein structure from amino-acid sequence was an unsolved computational grand challenge for more than half a century. Progress in the field has been aided by a series of community experiments aimed at maximizing rigor, transparency, and communication. These events saw advances, but until recently methods were only partially successful at best. All that changed in the last three years, with atomic level agreement with experiment achieved in many instances. These dramatic changes result from the introduction of recently developed deep learning methods, a branch of artificial intelligence.

In this talk, I’ll describe the intellectual journey to the solution, the nature of the AI methods used, and the latest applications to other problems in structural biology, particularly protein complexes. I’ll also discuss the role of alternative approaches to how science is done in this and other areas.

Contact imetdirectorsoffice@umces.edu for meeting information.