Global ranking of the sensitivity of interaction potential contributions within classical molecular dynamics force fields

Abstract

Uncertainty quantification (UQ) is rapidly becoming a sine qua non for all forms of computational science out of which actionable outcomes are anticipated. Much of the microscopic world of atoms and molecules has remained immune to these developments but due to the fundamental problems of reproducibility and reliability, it is essential that practitioners pay attention to the issues concerned. Here a UQ study is undertaken of classical molecular dynamics with a particular focus on uncertainties in the high-dimensional force-field parameters, which affect key quantities of interest, including material properties and binding free energy predictions in drug discovery and personalized medicine. Using scalable UQ methods based on active subspaces that invoke machine learning and Gaussian processes, the sensitivity of the input parameters is ranked. Our analyses reveal that the prediction uncertainty is dominated by a small number of the hundreds of interaction potential parameters within the force fields employed. This ranking highlights what forms of interaction control the prediction uncertainty and enables systematic improvements to be made in future optimizations of such parameters.

Artificial Intelligence Must Be Made More Scientific

“…We have been able to use new, scalable forms of uncertainty quantification which allow us to show that typically only between ten and 20 of these force field parameters have a significant influence on the properties of interest. 10 In other words, we acquire real insights and understanding about what parameters are important.…”

“…In fact, we do understand the science of molecular interactions, so one can also use a physics-based interaction potential (with terms which are scientifically meaningful) with perhaps hundreds to a few thousand parameters. We have been able to use new, scalable forms of uncertainty quantification which allow us to show that typically only between ten and 20 of these force field parameters have a significant influence on the properties of interest . In other words, we acquire real insights and understanding about what parameters are important.…”