Next-generation DFT-based quantum models for simulations of biocatalysis

Darrin M. York, york@chem.umn.edu, Department of Chemistry, University of
Minnesota, 207 Pleasant Street S.E, Minneapolis, MN 55455

The theoretical study of biocatalytic processes has been in the spotlight of
computational chemistry over the last several years. The use of combined quantum
mechanical/molecular mechanical (QM/MM) simulations and linear-scaling
electronic structure methods have been instrumental in providing insight into
the mechanisms of biocatalysis. In many instances the spatial extent of the
quantum region and/or the configurational space that must be sampled on the
quantum surface precludes the use of high-level electronic structure methods. In
these instances, empirical quantum models that are several orders of magnitude
more efficient have emerged as invaluable tools. The current quantum models such
as NDDO-based semiempirical, SCC-DFTB and related methods have limitations, some
of which are in common and others of which are more specific to a particular
class of methods. Some of these limitations include the accuracy of barriers in
chemical reactions, charge-dependent response properties, coupling of exchange
and polarization, completeness of the electrostatic description, modeling of
dispersion interactions and relative conformational energies. In this talk, a
strategy for the design of a next-generation DFT-based quantum model for
biocatalysis is outlined from functional form derived from density-functional
theory to parameterization and validation based on high-level quantum chemical
calculations and experiment. The integration of this model with many-body (e.g.,
polarizable) force fields will be outlined, and applications to highly charged
phosphoryl transfer reactions in solution will be presented.