Determining Lennard-Jones Parameters Using Multiscale Target Data through Presampling-Enhanced, Surrogate-Assisted Global Optimization

Abstract

Force field-based models are a Newtonian mechanics approximation of reality and are inherently noisy. Coupling models from different molecular scale domains (including single, gas-phase molecules up to multimolecule, condensed phase ensembles) is difficult, which is also the case for finding solutions that transfer well between the scales. In this contribution, we introduce a surrogate-assisted algorithm to optimize Lennard-Jones parameters for target data from different scale domains to overcome the difficulties named above. Specifically, our approach combines a surrogate-assisted global evolutionary optimization method with a presampling phase that takes advantage of one scale domain being less computationally expensive to evaluate. The algorithm's components were evaluated individually, elucidating their individual merits. Our findings show that the process of parametrizing force fields can significantly benefit from both the presampling method, which alleviates the need to have a good initial guess for the parameters, and the surrogate model, which improves efficiency.

Open-Source Machine Learning in Computational Chemistry

“…Molecule-specific (i.e., ammonium perchlorate, pentafluoroethane, difluoromethane) examples of optimizing Lennard-Jones parameters through multiobjective surrogate-assisted Gaussian process regression and support vector machine workflows can be found in ref and its two cited GitHub repositories. Similarly, PREMSO uses a presampling-enhanced, surrogate-assisted global evolutionary optimization strategy that allows the use of features at different scales (e.g., single-molecule and bulk-phase observables) . Also released recently, Thürlemann et al.…”

“…Similarly, PREMSO uses a presampling-enhanced, surrogate-assisted global evolutionary optimization strategy that allows the use of features at different scales (e.g., single-molecule and bulk-phase observables). 147 Also released recently, Thürlemann et al. developed a GNN to predict nonbonded parameters based on QM target data, 148 which includes atom typing prediction.…”