CDS&E: Tools to facilitate deeper data analysis, exploration, management, and sharing of ensembles of molecular dynamics trajectory data

University Of Utah, Salt Lake City UT

Investigators

Abstract

Thomas Cheatham, of the University of Utah is supported by an award from the Chemical Theory, Models and Computational Methods in the Chemistry Division to develop a software framework and set of tools to facilitate data analysis and management across scales and cyberinfrastructure resources. Cheatham and other researchers are now able to run large ensembles of Molecular Dynamics simulations including those exchanging information for enhanced sampling with multidimensional replica-exchange and with data saved at different levels of resolution. An important goal of this project is to provide an easy means to deeply inspect the data generated from these simulations, visualize the results, and ultimately guide the simulation workflow. The Computational and Data-Enabled Science and Engineering (CDS&E) programs in the Chemistry Division and in the Division of Advanced Cyberinfrastructure cofund this award. The project involves the development of two complementary software environments, specifically iBIOMES and the MD trajectory analysis code cpptraj. iBIOMES provides a framework that allows automatic annotation, analysis, searching, and display of the simulation results The ultimate goal is to extend and generalize these environments to provide the means to allow anyone to search, explore, and analyze the raw data, derived data, and/or subsets of the data. The program cpptraj will be further developed by providing new ways of thinking about the analysis of MD trajectories and the management of the data with a focus on deeper analysis of the data, including more "interactive" analysis capabilities. Biomolecular simulations provide detailed insight into biomolecule structure, dynamics, interactions and function. With access to tremendously powerful computational resources, the community is being flooded with raw data from these simulations. Not only is managing and processing this data difficult, understanding the meaning of the data is challenging. Through further software developments we aim to make it easier to manage and analyze the data, and ultimately annotate, search and disseminate the data to the wider community.