Molecular and Coarse-Grained Simulations of Biomolecular Processes at the Petascale

Gregory Voth, University of Chicago

Usage Details

Gregory Voth, Glen Hocky, Zachary Jarin, Aleksander Durumeric, Alexander Pak, Jesper Madsen, Dudu Tong, Arpa Hudait, Alvin Yu, Paul Calio

Computer simulations of biological systems can offer insights that are difficult or impossible to access with conventional experimental techniques, providing significant benefits for basic scientific research. However, the large range of characteristic time and length scales observed in biological processes makes the use of any single computational technique difficult. While atomic-resolution simulations can furnish a scientist with exquisite levels of detail, the sheer computational expense of these simulations sometimes presents a significant barrier for their application to large-scale biological problems. This project proposes to us coarse-grained (CG) molecular models to expand the reach of computer simulations to cellular scales. The project propose to integrate atomic-resolution and CG simulations to study a range of biologically relevant systems, in close collaboration with an international cohort of scientists from various experimental fields. This work will involve not only elucidating and explaining biomolecular processes, but also the development and dissemination of cutting-edge simulation software to the wider scientific community.

The project aims to combine experimental data with cutting-edge computer simulations to investigate a number of important biomolecular systems. The systems of interest can be grouped into two main categories: critical stages of the viral lifecycles of HIV-1 and influenza, and studies of the actin filaments and microtubules of the cellular cytoskeleton. The project will develop an integrated pipeline which allows scientists to convert experimental data into computer models capable of investigating biomolecular processes at scales that are inaccessible to other approaches. While the results of atomic-scale simulations will clearly be important in and of themselves, they will also, in combination with experimental data, form the basis for generating and parameterizing rigorous UCG models. Results and predictions made by these models will be validated by close collaboration with experimental scientists, and used to suggest new directions in both the theoretical and experimental field. In addition to the development and deployment of advanced biomolecular simulation techniques, this proposal will also assist in the dissemination of the advances to the wider research community by the integration of the UCG model generation and simulation algorithms with the popular LAMMPS software.