Berlin 2014 – wissenschaftliches Programm

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

MO: Fachverband Molekülphysik

MO 9: Theory 2: Molecular Dynamics & Quantum Chemistry

MO 9.3: Vortrag

Dienstag, 18. März 2014, 14:30–14:45, BEBEL SR144

Protein long timescale simulations using kinetic Monte Carlo — •Mateusz Marianski, Carsten Baldauf, and Matthias Scheffler — Fritz-Haber-Institut der MPG, Berlin, Germany

The time scales that govern the structure formation and dynamics of biopolymers are hardly accessible with established molecular dynamics (MD) simulations. Too short simulation times may even lead to misinterpretation of the data, as previously shown for the dynamics of the amyloid beta peptide (Lin et al. Biophys. J. 2012, 102, 315).

This project aims at the development of an universal kinetic Monte Carlo (kMC) protocol based on a separation of short- (bonded) and long-range (non-bonded) contributions as an alternative sampling technique applicable to peptides. Such kMC model of the structural dynamics of a molecule requires two components: (1) a library of accessible conformations, and (2) the transition states between these conformers. For the development, we turn to classical force fields, especially since we can compare to results from brute-force parallel-tempering MD sampling.

In this contribution, we will present results for polyalanine peptides of different lengths. The conformational library is constructed using structure-search algorithms and refined with clustering techniques. The resulting conformational basins are connected by transition states which later are used, by applying harmonic transition state theory, to estimate rates that are the input to the kMC simulation. Here we compare these results to brute-force sampling based on MD simulations to test the potential of such a protocol.