UC San Diego

An extended QM/MM NWChem/AMBER interface has been developed and implemented to offer additional features in computations within the QM/MM framework. This includes the interface for the QM/MM multiregion optimization, nudged elastic band (NEB), and free energy perturbation (FEP). With these functionalities, it is feasible to apply ab-initio or density functional (DFT) QM/MM methods to study various problems, for example, reaction mechanisms of enzymes, in which many degrees of freedom are involved and considered. The QM/MM multiregion optimization provides an efficient approach that reduces computational costs for optimization problems, e.g. geometry optimization. The minimum energy pathway of enzyme catalysis can be determined using the NEB method. Together with the FEP method, free energy profile of the reaction can be established.

A brief introduction of QM/MM methodology has been discussed in chapter 1. Chapter 2 covers the implementation of the extended QM/MM NWChem/AMBER interface with the related theoretical backgrounds also discussed. Chapter 3 includes the validation of the implemented interface using small test cases. Water dimer was chosen to validate the QM/MM multiregion optimization algorithm. The QM/MM NEB and FEP methods were applied to determine the reaction mechanism of a SN2 reaction of CH3Br + OH- in aqueous solution. Chapter 4 illustrates the use of the QM/MM multiregion optimization in combination with a high-level QM methodology (CR-EOMCCSD(T)) to investigate photophysical properties of a green fluorescent chromophore (GFP) analogue in aqueous solution. In chapter 5, the QM/MM method was applied to investigate the catalysis of cellulose degradation catalyzed by copper-dependent polysaccharide monooxygenases (PMOs). The computations were achieved through the utilization of the developed NWChem-AMBER interface. As a primary step, the hydrogen abstraction (HAT) process was examined. The corresponding proposed reaction mechanisms were investigated and activation free energies were estimated. This work also illustrates the use of the full explicit enzyme-substrate model complex in the study of reaction mechanism of PMOs.