Zoom ID: 943 3556 5836
Passcode: 754381
- Join Via Zoom -
Abstract
Protein conformational changes play an important role in numerous biological processes. Markov State Model (MSM) built from molecular dynamics (MD) simulations provides a useful approach to study these complex dynamic processes, but it is challenging to build truly Markovian models due to the limited length of lag time (bound by the length of relatively short MD simulations). In this talk, I will introduce our recent work on developing Generalized Master Equation (GME) methods based on the projection operator scheme that encodes the non-Markovian dynamics in a generally time-dependent memory kernel, whose characteristic decay time scale corresponds to the kernel lifetime. We show that GME methods can greatly improves upon Markovian models by accurately predicting long timescale dynamics using much shorter MD trajectories on complex conformational changes including clamp opening of RNA polymerase. Based on the projection operator scheme, I will also introduce our recent development of the Encoder-neural-network based “RPnet” method for coarse-graining protein dynamics. The key insight of our RPnet method is that we designed a reverse projection scheme that allows us to quantify the difference (defined as the loss function) of transition modes between original dynamics and coarse-grained dynamics. RPnet provides a new way to define the loss function based on transition modes, which is different from other methods where the loss function is defined based on the variational principle of conformational dynamics. RPnet could yield comparable or better results than competing methods in terms of state partitioning and reproduction of slow dynamics in various systems. Finally, I will also introduce our recent work on the development of another Encoder-neural-network: Memory-Net that can efficiently identify the slow collective variables of protein conformational changes by minimizing the memory integrals (defined as the loss function). We expect that these GME-based methods hold promise to be widely applied to study functional dynamics of proteins.
Biography
Xuhui Huang obtained his Ph.D. from Columbia University in 2006 with Prof. Bruce Berne. He did his postdoc research at Stanford University with Profs. Michael Levitt and Vijay Pande. He was as an Assistant, Associate and Full Professor of the Hong Kong University of Science and Technology (HKUST) between 2010 and Summer 2021. Since Fall 2021, he took up the position of the Hirschfelder Endowed Chair Professor in Theoretical Chemistry, and Director of Theoretical Chemistry Institute at University of Wisconsin-Madison. He has received numerous awards, including Biophysical Society Theory & Computation Award for Mid-Career Scientists (2023), Pople Medal from the Asia-Pacific Association of Theoretical and Computational Chemists (2021), 16th China Youth Science and Technology Award (2020), American Chemical Society OpenEye Outstanding Junior Faculty Award (2014), and Hong Kong Research Grant Council Early Career Award (2013). He is a founding member of Young Academy of Sciences of Hong Kong (YASHK) and Fellow of Royal Society of Chemistry (FRSC). His group pioneered in elucidating the dynamics of protein conformational changes by developing new methods based on statistical mechanics that can bridge the gap between experiments and atomistic MD simulations.
Seminar Series by the NYU-ECNU Center for Computational Chemistry at NYU Shanghai