The influence of multi-dimensionality and off-diagonal non-Markovian friction coupling on coarse-grained dynamics

Abstract: Coarse-graining complex molecular systems to lower-dimensional reaction coordinates is a powerful approach for capturing their effective dynamics. The generalized Langevin equation (GLE) provides an exact framework for modeling coarse-grained dynamics, and is particularly useful when non-Markovian effects are significant. While one-dimensional GLE models are commonly used, many systems require multi-dimensional reaction coordinates to account for coupled dynamics. Here, we study the GLE formalism for multi-dimensional reaction coordinates, incorporating a memory matrix to quantify non-Markovian frictional coupling between coordinates, and a multi-dimensional potential. Using the GLE model, in conjunction with a multi-dimensional Markovian embedding scheme, we investigate different systems that are characterized by two-dimensional reaction coordinates, namely the dihedral dynamics of pentane and alanine dipeptide, obtained from molecular dynamics simulations in explicit water. We identify significant off-diagonal friction couplings arising from intramolecular and hydrodynamic interactions. Unlike previous studies, our results highlight the critical role of different terms in the multi-dimensional GLE in accurately capturing key dynamical properties, including mean first-passage times and mean-squared displacements, particularly in systems with coupled non-Markovian coordinates.