New solvent models, sampling methods and maintenance of Amber software

State University New York Stony Brook, Stony Brook NY

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): This proposal responds to PAR-11-028, Continued Development and Maintenance of Software: The goal of this program is to support the continued development, maintenance, testing and evaluation of existing software. Amber is a popular software package, licensed to over 800 academic and industry institutions, for simulating the structural, thermodynamic and kinetic properties of molecular systems. There are 885 citations to the popular Amber ff99SB force field, developed at Stony Brook by PI Carlos Simmerling. The Simmerling group is one of the six academic groups responsible for Amber maintenance and development. Computational Molecular Dynamics simulations using Amber and other software packages have become essential counterparts to experimental research for understanding the mechanisms of biomolecules, and for discovering drugs to inhibit them. The popular virtual screening program called DOCK interfaces directly with Amber. One goal of the software improvements proposed here is to reduce the time it takes to develop new drugs. We propose here new developments for Amber, addressing the most pressing needs of the field. Aim 1, Solvation: We will incorporate improved solvation models: (a) the SEA semi-explicit water model and (b) a new Generalized Born model. Aim 2, Sampling: We will add fast and targeted sampling methods: (a) variants of the general tools Hamiltonian exchange and thermal Replica Exchange Molecular Dynamics, (b) the new Modeling with Limited Data method, which samples conformations subject to sparse and noisy data; (c) the very fast Kinetic-Loop-Closure and Constrained-Loop-Closure methods for sampling loop conformations, and (d) the Confine and Transition method, which computes free- energy differences between two different conformations of a biomolecule. We will incorporate our recently developed algorithms into the Amber production code for distribution, and port the codes to the recently developed GPU version of Amber using CUDA.