Exploring the Conformers of an Organic Molecule on a Metal Cluster with Bayesian Optimization

Abstract

Finding low-energy conformers of organic molecules is a complex problem due to the flexibilities of the molecules and the high dimensionality of the search space. When such molecules are on nanoclusters, the search complexity is exacerbated by constraints imposed by the presence of the cluster and other surrounding molecules. To address this challenge, we modified our previously developed active learning molecular conformer search method based on Bayesian optimization and density functional theory. Especially, we have developed and tested strategies to avoid steric clashes between a molecule and a cluster. In this work, we chose a cysteine molecule on a well-studied gold−thiolate cluster as a model system to test and demonstrate our method. We found that cysteine conformers in a cluster inherit the hydrogen bond types from isolated conformers. However, the energy rankings and spacings between the conformers are reordered.

Accelerated Bayesian Inference for Molecular Simulations using Local Gaussian Process Surrogate Models

“…Bayesian methods are the gold standard for these aims, with examples spanning from neutrino and dark matter detection, materials discovery and characterization, − quantum dynamics, , to molecular simulation. − The Bayesian probabilistic framework is a rigorous, systematic approach to quantify probability distribution functions on model parameters and credibility intervals on model predictions, enabling robust and reliable parameter optimization and model selection. , Interest in Bayesian methods and uncertainty quantification for molecular simulation has surged − due to its flexible and reliable estimation of uncertainty, ability to identify weaknesses or missing physics in molecular models, and systematically quantify the credibility of simulation predictions. Additionally, standard inverse methods including relative entropy minimization, iterative Boltzmann inversion, and force matching have been shown to be approximations to a more general Bayesian field theory…”

A Bayesian method for selecting data points for thermodynamic modeling of off-stoichiometric metal oxides

“…In this work, we aim to ll this methodological gap through a novel algorithm which systematically selects the most important experimental data points to collect without prior knowledge of the system behavior. This method uses a combination of our recently developed cross-t experimental and computational CEF model 42 (CrossFit CEF) and Bayesian inference [43][44][45][46] to select the data points to be experimentally examined and guide the tting of robust thermodynamic models. In our algorithm, we leverage ab initio methods to estimate material enthalpies while the complex temperature and entropy interactions arise from TGA data.…”

MolOpt: Autonomous Molecular Geometry Optimization Using Multiagent Reinforcement Learning

“…Recently, much emphasis has been on developing efficient ML-based search algorithms to explore chemical space, − but the same is not the case for conformational space. There are very few attempts to develop an efficient ML-based search algorithm that can explore the conformational space, i.e., probe the potential energy surface (PES). − These ML-based search algorithms have applications in 3D structure generation , and molecular geometry optimization (MGO). MGO aims to find the nearest/local molecular conformation with minimum potential energy on PES, starting from a given initial 3D conformation.…”