BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Learning the shape of the protein universe
DTSTART:20221121T160000
DTEND:20221121T163000
DTSTAMP:20260506T084533Z
UID:18fd7da3673ec89c7f9e11d2763e82d236e42afdff450bf42fcd0a12
CATEGORIES:Conferences - Seminars
DESCRIPTION:Armita Nourmohammad is an Assistant Professor of Physics and A
 pplied Mathematics at the University of Washington (Seattle)\, and an Affi
 liate Investigator at the Fred Hutchinson Cancer Research Center. Dr. Nour
 mohammad works at the interface of statistical physics and biology and dev
 elops theoretical and data driven approaches to study evolutionary process
 es across scales. Dr. Nourmohammad obtained her Ph.D in 2012 from the Univ
 ersity of Cologne\, and then joined Princeton University as a James S. McD
 onnell postdoctoral fellow\, where she started working on immunological pr
 oblems from a biophysical and an evolutionary perspective. In 2017 Dr. Nou
 rmohammad joined the Max Planck Society as Max Planck Research Group Leade
 r and the University of Washington as an Assistant Professor of Physics. D
 r. Nourmohammad is a recipient of an NSF CAREER award\, NIH MIRA award\, 
 and the APS-DBIO Early Career Award.\n\nProteins are the machinery of lif
 e facilitating the key processes that drive living organisms. The physical
  arrangement of amino acids dictates how proteins fold and interact with t
 heir environment. Recent advances have increased the number of experimenta
 lly resolved or computationally predicted tertiary structures\, however we
  still lack a practical understanding of how 3D structure determines the f
 unction of a protein. While machine learning has been at the forefront of 
 protein science\, the inferred models are often hard to interpret physical
 ly.  In this talk I will introduce physically motivated machine learning 
 approaches to  learn interpretable models of protein micro-environments\,
  reflecting the underlying biophysics. With these models we infer amino ac
 id preferences given a surrounding atomic neighborhood\, and predict the i
 mpact of evolutionary substitutions in proteins. Our computational approac
 h establishes an interpretable model for how biological function emerges f
 rom protein micro-environments. The flexibility and efficiency of this app
 roach also show promise for building generative models to design novel pro
 tein structures with desired function.\n 
LOCATION:BSP 231 https://plan.epfl.ch/?room==BSP%20231
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR