Towards Biomolecular Simulations with Explicit Inclusion of Polarizability: Development of a CHARMM Polarizable Force Field based on the Classical Drude Oscillator Model

Abstract

Molecular dynamics simulations, based on molecular mechanics force fields, have been instrumental in increasing our understanding of a range of biomolecular systems. Most of the force fields commonly used for the simulation of biomolecules, however, represent electrostatic properties by a set of fixed partial atomic charges and are referred to as additive force fields. This approach is problematic because it does not explicitly include polarizability, an important component of the electrostatic interaction arising from the response of the molecular dipoles to an external electric field.To overcome this problem, multiple efforts are currently underway to develop force fields including an explicit representation of polarizability: one such effort is the CHARMM Drude polarizable force field. In this chapter, we begin by discussing the development and implementation of the theory associated with the Drude model, as well as a robust scheme for parameter optimization. This is followed by discussion of the way in which these tools are being used to optimize a polarizable force field for the simulation of biomolecules. The primary focus has been the development of parameters for small molecule analogues of functional groups present within biomolecules, but is shifting to the optimization of covalent connections between these small molecules, and testing of the resulting parameters in fully hydrated biomolecules. For the nucleic acids a case study is presented, illustrating that the polarizable force field already reveals atomic-level details not observed with an additive force field.