Temperature-Shuffled Structural Dissimilarity Sampling Based on a Root-Mean-Square Deviation.

Structural dissimilarity sampling (SDS) has been proposed as an enhanced conformational sampling method for finding neighboring metastable states of a given reactant or generating transition pathways starting from the reactant. SDS repeats a cycle of two steps: (1) selections of initial structures based on structural dissimilarities by referring to a measure and (2) conformational resampling by restarting short-time molecular dynamics (MD) simulations from the initial structures. In the present study, the measure was defined as the root-mean-square deviation (RMSD) among the resampled snapshots to characterize their structural dissimilarities. Additionally, the temperatures in restarting the short-time MD simulations were randomly shuffled at the beginning of each cycle to further promote the conformational transitions. We call this approach temperature-shuffled SDS (TSF-SDS). As a demonstration, TSF-SDS was applied to promote the open-closed transition of T4 lysozyme (T4L) in explicit water. TSF-SDS successfully reproduced the relevant domain motion with nanosecond-order simulation time, whereas conventional SDS without shuffling of the temperatures failed to promote the transition of T4L, indicating the high conformational sampling efficiency of TSF-SDS for promoting essential conformational transitions of proteins. Furthermore, as a wide-range application, TSF-SDS efficiently identified the native state of trp-cage and a dissociation process of ubiquitin dimer in explicit water.