UC Irvine

Molecular dynamics simulations of biomolecules have been widely adopted in biomedical studies. Efficient simulations require efficient functional forms, i.e. force fields, to model intra- and inter-molecular interactions. As molecular interactions can be separated into two main types: ‘long-range’ and ‘short-range’, a reliable force field is often designed to contains different terms model each of these interactions. This dissertation concerns the modeling of long-range interactions, particularly electrostatic and induction interactions. Another long-range interaction, dispersion interaction, would follow widely accepted formulation. Our goal here is to establish an accurate and efficient multi-scaled modeling scheme for biomolecular systems, which will include atomic, coarse-grained and implicit-solvent level of models. Once completed, our multi-scaled models can be combined in different manner for different biomolecular applications, serving for various accuracy and efficiency requirements in molecular dynamics simulations.In the atomic level, as classical point-charge models continue to be used in routine biomolecular applications, there have been growing demands on developing polarizable force fields for handling more complicated biomolecular processes. Here, we focus on a recently proposed polarizable Gaussian Multipole (pGM) model for biomolecular simulations. We present an efficient formulation for the pGM model, and its implementation under various simulation conditions. It is hoped that the reformulated pGM model will facilitate the development of future force fields. Since most particles in explicit molecular model are water molecules that solvate the target biomolecules, treating these water molecules implicitly would allow higher computational efficiency without losing any atomic-level resolution of the biomolecules. Poisson–Boltzmann equation (PBE)-based implicit solvent model has been one such attempt and been widely used in biomolecular applications. Thus, in the implicit-solvent level, we developed several new implementations of Poisson–Boltzmann (PB) models. We also present here the GPU implementations of the linear solver needed in those new PB models. Our analysis shows that the new strategies improve the convergence and stability of PB models, and also improves the efficiency of the method.