Uncertainty Quantification in Classical Molecular Dynamics

Abstract

Molecular dynamics simulation is now a widespread approach for understanding complex systems on the atomistic scale. It finds applications from physics and chemistry to engineering, life and medical science. In the last decade, the approach has begun to advance from being a computer-based means of rationalising experimental observations, to producing apparently credible predictions for a number of real-world applications within industrial sectors such as advanced materials and drug discovery. However, key aspects concerning the reproducibility of the method have not kept pace with the speed of its uptake in the scientific community. Here, we present a discussion of uncertainty quantification for molecular dynamics simulation designed to endow the method with better error estimates that will enable the method to be used to report actionable results. The approach adopted is a standard one in the field of uncertainty quantification, namely using ensemble methods, in which a sufficiently large number of replicas are run concurrently, from which reliable statistics can be extracted. Indeed, because molecular dynamics is intrinsically chaotic, the need to use ensemble methods is fundamental and holds regardless of the duration of the simulations performed. We discuss the approach and illustrate it in a range of applications from materials science to ligand-protein binding free energy estimation.

33 pages, 3 figures

References in corpus (4)

• A Gaussian Treatment for the Friction Issue of Lennard-Jones Potential in Layered Materials: Application to Friction between Graphene, MoS2 and Black Phosphorus
• When we can trust computers (and when we can't)
• Probabilistic performance estimators for computational chemistry methods: Systematic Improvement Probability and Ranking Probability Matrix. I. Theory
• Functional uncertainty quantification for isobaric molecular dynamics simulations and defect formation energies

Cited by in corpus (7)

• Best practices for constructing, preparing, and evaluating protein-ligand binding affinity benchmarks
• Universal QM/MM Approaches for General Nanoscale Applications
• The long road to calibrated prediction uncertainty in computational chemistry
• Thirty years of molecular dynamics simulations on posttranslational modifications of proteins
• Bayesian, frequentist, and information geometric approaches to parametric uncertainty quantification of classical empirical interatomic potentials
• Towards data-driven mass spectrometry in atmospheric science
• The effect of non-ionizing excitations on the diffusion of ion species and inter-track correlations in FLASH ultra-high dose rate radiotherapy