Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

CPP: Fachverband Chemische Physik und Polymerphysik

CPP 22: Data-driven Methods in Molecular Simulations of Soft-Matter Systems

CPP 22.1: Vortrag

Dienstag, 13. März 2018, 09:30–09:45, C 230

Consistent interpretation of protein simulation kinetics using Markov state models biased with external information — •Joseph Rudzinski, Kurt Kremer, and Tristan Bereau — Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz

Molecular simulation models are often required to provide a microscopically-detailed interpretation of observations from, e.g., single-molecule spectroscopy experiments of proteins. In general, model errors may lead to inconsistencies between simulated and experimentally-measured observables. For static properties, well-established reweighting methods exist for adjusting the simulated ensemble with minimum bias in order to attain consistency for observables of interest. This work presents an analogous reweighting framework that adjusts the ensemble of dynamical paths sampled in a molecular simulation in order to ensure consistency with kinetic observables. The proposed methodology leverages Markov state modeling techniques to efficiently treat the simulated dynamical paths, while employing a Bayesian scheme to reweight these paths with minimum bias according to external reference data. Using small peptide systems as a proof of concept, we demonstrate how biasing a Markov state model significantly improves the kinetic description of the system, while refining the static equilibrium properties. Our conclusions highlight the potential of the methodology for providing consistent interpretations of kinetic protein experiments.