Machine learning has become an integral tool in the theoretical and computational molecular sciences. Uses of machine learning in this area include prediction of molecule and materials properties from large databases of descriptors, design of new molecules with desired characteristics, design of chemical reactions and processes, representation of high-dimensional potential energy and free energy surfaces, creation of new enhanced sampling strategies, and bypassing of costly quantum chemical calculations, to highlight just a few. This talk will focus on the use of machine learning in the analysis and creation of free-energy based enhanced sampling approaches. We will first briefly review collective-variable based enhanced sampling techniques and then compare different classes of machine learning models, including kernel methods, neural networks, decision-tree approaches, and nearest-neighbor schemes for representing and deploying high-dimensional free energy surfaces generated from these enhanced-sampling algorithms based on limited sampled data. We will then show how machine learning can be used to define collective variables capable of generating kinetic pathways between conformational basins and elucidating mechanisms. Finally, we will explore strategies for discovering reaction coordinates from large, redundant sets of collective variables based on machine learning the committor distribution.

Synergizing enhanced sampling and machine learning for the representation and deployment of high-dimensional free-energy surfaces and discovery of reaction coordinates for rare events

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.