Theoretical Models for Potential Energy Landscapes of Challenging Chemical Systems

Georgia Tech Research Corporation, Atlanta GA

Investigators

Abstract

David Sherrill of Georgia Tech University is supported by an award from the Chemical Theory, Models and Computational Methods program to develop theoretical and computational approaches to study weak, non-bonded interactions between molecules. Such interactions determine how medicines bind to their targets in the body, how molecules pack into solids, and how liquids behave. Hence, interactions between molecules are of fundamental importance for science and technology, and understanding these interactions better could aid in the design of better materials or more effective medicines. Sherrill and his research group focus on new theoretical techniques based on quantum mechanics. They implement their methods in user-friendly, open-source, widely-available and free computer software. This allows other researchers to model and analyze interactions between molecules more accurately, in more situations, and with a higher level of detail. The project is contributing to the training of postdocs, graduate students, and undergraduates in quantum chemistry methods and software development. The project is cofounded by the Computational and Data-enabled Science and Engineering Program in the Division of Advanced Cyberinfrastructure. Symmetry-adapted perturbation theory (SAPT) will be extended to compute interactions between different molecular fragments, even when those fragments are in the same molecule (this is not possible with current versions of SAPT). The intramolecular SAPT allows investigations of how linker groups in molecular torsion balances affect the interactions they have been made to study. These techniques are also being extended to molecules with unpaired electrons, allowing an investigation of how methionine-aromatic contacts contribute to protein stabilization, and how these contacts change their strength upon ionization of the methionine group. The project is also investigating various many-body expansion approximations (including density embedding) for use in high-accuracy computations of extended systems. The methods developed as part of this project are being implemented as open-source software, to be made available to the entire scientific community. The enhanced ability to model interactions between molecules should benefit researchers in chemistry, materials science, biochemistry, and drug design. At the same time, associated computational laboratory exercises are being developed that teach fundamental concepts of intermolecular interactions.