Structuring and sampling complex conformation space: Weighted ensemble dynamics simulations

TL;DR

This paper introduces a weighted ensemble dynamics method that uses spectral analysis of trajectory data to explore complex conformational spaces, predict ergodicity, and classify states without prior knowledge.

Contribution

The method systematically analyzes simulation trajectories to identify metastable states and transition pathways in complex conformational spaces, enhancing understanding of molecular dynamics.

Findings

Successfully applied to glassy potential system

Effectively characterized alanine dipeptide conformations

Predicted transition states without prior system knowledge

Abstract

Based on multiple simulation trajectories, which started from dispersively selected initial conformations, the weighted ensemble dynamics method is designed to robustly and systematically explore the hierarchical structure of complex conformational space through the spectral analysis of the variance-covariance matrix of trajectory-mapped vectors. Non-degenerate ground state of the matrix directly predicts the ergodicity of simulation data. The ground state could be adopted as statistical weights of trajectories to correctly reconstruct the equilibrium properties, even though each trajectory only explores part of the conformational space. Otherwise, the degree of degeneracy simply gives the number of meta-stable states of the system under the time scale of individual trajectory. Manipulation on the eigenvectors leads to the classification of trajectories into non-transition ones within the states and transition ones between them. The transition states may also be predicted without a priori knowledge of the system. We demonstrate the application of the general method both to the system with a one-dimensional glassy potential and with the one of alanine dipeptide in explicit solvent.