Biased Metropolis Sampling for Rugged Free Energy Landscapes
Bernd A. Berg
Abstract
Metropolis simulations of all-atom models of peptides (i.e. small proteins) are considered. Inspired by the funnel picture of Bryngelson and Wolyness, a transformation of the updating probabilities of the dihedral angles is defined, which uses probability densities from a higher temperature to improve the algorithmic performance at a lower temperature. The method is suitable for canonical as well as for generalized ensemble simulations. A simple approximation to the full transformation is tested at room temperature for Met-Enkephalin in vacuum. Integrated autocorrelation times are found to be reduced by factors close to two and a similar improvement due to generalized ensemble methods enters multiplicatively.
Create a lesson
Related papers
Real-Space renormalization for the drawups and drawdowns of stochastic trajectories: universality classes for diffusion processes and random walks via the scale function
Cecile Monthus
Stochastic synthesis and degradation of arbitrary diffusive processes
Denis Boyer, Gabriel Mercado-Vásquez
Critical amplitudes of the Baxter-Wu model from MCPA and Wolff simulations: free-exponent and fixed-exponent fits
Vyacheslav Mozolenko, Ian Pile, Lev Shchur
Capillary Coarsening Drives Re-entrant Statistical Transitions in Early Coffee-Ring Growth
Arundas Chandradasan, Dileep Mampallil
Maximum entropy explanation for the ubiquity of power-law mechanical responses
Sverre Holm
Thermodynamic Bethe Ansatz for the Toda chain from inverse scattering transform
Alberto Brollo