Skip to main content
Loading Events

« All Events

BMOL Seminar Series: Dr. Matthias Heyden

Wednesday, October 23 @ 3:00 pm - 4:15 pm MDT

Matthias Heyden Headshot

Seminar Details

Speaker: Dr. Matthias Heyden, Associate Professor of Molecular Sciences at Arizona State University

Host: Dr. Konrad Meister, Chemistry & Biochemistry

Title: From Anharmonic Vibrations to Protein Dynamics

Abstract: The relationship between conformational dynamics and protein function is a key challenge for computational enzyme design, where strategies based on structure alone result in limited catalytic activities. The same applies to the computational design of small-molecule drugs with the aim to modulate the function of a target protein with a highly specific allosteric mechanism that avoids side effects. The key to exploring structure-dynamics-activity relationships in computer simulations is the ability to efficiently sample quantitative free energy surfaces. For practical applications, converged results should be obtainable in 24 hours or less. Here, we explore new strategies that aim to enable such studies with standard computer hardware. Our approach centers on the identification of optimal collective variables (CVs) that describe conformational transitions in proteins with minimal free energy barriers. Identifying such CVs reliably and without prior knowledge maximizes the achievable acceleration in so-called “enhanced sampling simulations”. We recently introduced the FREquency-SElective Anharmonic (FRESEAN) mode analysis, which identifies collective degrees of freedom contributing to the vibrational spectrum at any given frequency. Notably, this formalism is free of harmonic approximations and applicable in the diffusive regime, i.e., at zero frequency. Using this approach, we identify optimal CVs for to dramatically improve the sampling of protein conformational transitions. As a proof of concept, we present data for multiple proteins with known conformational ensembles. Comparisons to the literature demonstrate that our approach reproduces and improves on existing data while minimizing required computational costs and need for a priori knowledge.

Publication Related to Talk: Exploring Conformational Landscapes Along Anharmonic Low-Frequency Vibrations