Chapter 12: Machine Learning Algorithms for the Analysis of Molecular Dynamics Trajectories

Molecular dynamics (MD) simulations have been a pivotal tool to understand many processes, including drug binding, protein folding and many others in the field of materials science. In the past decades, machine learning (ML) has become a valuable tool in the field of MD simulations. The synergies between ML algorithms and MD simulations, including both classical and quantum mechanical simulations, can substantially transform the way we predict and solve computational structural biology, drug discovery and MD simulation problems. In this chapter, we describe how ML advances the understanding and interpretation of MD trajectories. This chapter also concentrates on the implementation of ML algorithms in MD simulations that can be programmed so that they can be used as input to train ML models for the quantitative comprehension of molecular systems.