Current state-of-the art in computational modeling of biochemical processes is to use hybrid quantum mechanics-molecular mechanics (QM/MM) methods that provide a balance between computational accuracy and efficiency. Furthermore, polarizable model potentials and polarizable QM/MM schemes become increasingly more important as they provide a more rigorous description of the classical environment. However, despite enormous computational speed-ups attained through describing the majority of the system classically, remaining bottlenecks of the QM/MM models are due to the necessity of computing long-range electrostatic interactions in an extended system. Here we aim to eliminate these bottlenecks and enable the users to perform simulations of biological systems in a more efficient and robust way, using either classical point-charge or polarizable QM/MM models. Specifically, we develop new algorithms and computer codes based on the continuous fast multipole method (CFMM) that will dramatically decrease the computational complexity of large-scale modeling. New computer codes are implemented within the Q-Chem quantum chemistry package and benchmarked through the modeling of realistic systems such as redox conversions in biochemical systems, protein-ligand binding energies, and lattice energies of molecular crystals.

Effective fragment potentials made faster

2019 Fall National Meeting

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