Machine Learning in Molecular Dynamics Simulations of Biomolecular Systems

Abstract

Machine learning (ML) has emerged as a pervasive tool in science, engineering, and beyond. Its success has also led to several synergies with molecular dynamics (MD) simulations, which we use to identify and characterize the major metastable states of molecular systems. Typically, we aim to determine the relative stabilities of these states and how rapidly they interchange. This information allows mechanistic descriptions of molecular mechanisms, enables a quantitative comparison with experiments, and facilitates their rational design. ML impacts all aspects of MD simulations -from analyzing the data and accelerating sampling to defining more efficient or more accurate simulation models. This chapter focuses on three fundamental problems in MD simulations: accurately parameterizing coarse-grained force fields, sampling thermodynamically stable states, and analyzing the exchange kinetics between those states. In addition, we outline several state-of-the-art neural network architectures and show how they are combined with physics-motivated learning objectives to solve MD-specific problems. Finally, we highlight open questions and challenges in the field and give some perspective on future developments.

Neural potentials of proteins extrapolate beyond training data

“…[56,57] Four main steps are involved in building a MSM -1) featurization 2) dimensionality reduction 3) clustering and 4) estimation of the transition matrix. [58] They are extensively discussed in literature on approximating observables from MD simulations. [59][60][61][62][63] We selected the 𝛼-Carbon pairwise distances to featurize mapped trajectories for the dimensionality reduction.…”

Neural potentials of proteins extrapolate beyond training data

“…[48,49] The main steps involved in building a MSM are, 1) featurization 2) dimensionality reduction 3) clustering and 4) estimation of the transition matrix. [50] They are extensively discussed in literature on approximating observables from MD simulations. [51][52][53][54][55] We selected the 𝛼-Carbon pairwise distances to featurize mapped trajectories for the dimensionality reduction.…”