Multi scale QM-MM modeling of biomolecules chemical reactivity

D. A. ESTRIN

sao Paulo

Conferencia; Seminario Universidad Sao Paulo; 2021

Universidad de Sao Paulo

Computational techniques for modeling chemical reactivity in complex environments have emerged during the last decades as an important tool to complement experimental information, providing atomic resolution insight into the reaction mechanisms, as well as thermodynamic and kinetic information. An elegant way to deal with this issue consists in employing multi level quantum classical schemes (QM-MM). We will present in this talk an overview of our group QM-MM implementation, which combines a density functional theory (DFT) description of the QM subsystem using Gaussian basis sets with a classical description of the environment using the Amber force field. The implementation, called LIO, has been effficiently adapted for using graphical process units (GPU). A key issue in the study of reaction mechanisms is the definition of a suitable reaction coordinate. In this context, we will show an implementation of a nudge elastic band scheme which allow us to judge the quality of a given reaction coordinate. We will illustrate this approach by analyzing the mechanism of the reduction of ferric heme proteins by nitric oxide. In addition, we will show results of a study of the reactions of thiols with peroxides, which are very relevant in redox biochemistry, both in aqueous and protein environments.