Rugged Metropolis Sampling with Simultaneous Updating of Two Dynamical Variables

TL;DR

The paper introduces an extension of the Rugged Metropolis algorithm to update two variables simultaneously, significantly improving sampling efficiency in rugged energy landscapes, demonstrated with Met-Enkephalin.

Contribution

It presents the RM$_2$ algorithm for simultaneous two-variable updates and evaluates its performance improvements over traditional methods.

Findings

RM$_2$ enhances sampling efficiency by a factor of four for Met-Enkephalin.

Correlations among multiple dihedral angles limit further improvements at low temperatures.

A multi-hit scheme allocates more CPU time to variables with high autocorrelation.

Abstract

The Rugged Metropolis (RM) algorithm is a biased updating scheme, which aims at directly hitting the most likely configurations in a rugged free energy landscape. Details of the one-variable (RM$_1$) implementation of this algorithm are presented. This is followed by an extension to simultaneous updating of two dynamical variables (RM$_2$). In a test with Met-Enkephalin in vacuum RM$_2$ improves conventional Metropolis simulations by a factor of about four. Correlations between three or more dihedral angles appear to prevent larger improvements at low temperatures. We also investigate a multi-hit Metropolis scheme, which spends more CPU time on variables with large autocorrelation times.