Computing Paradigms Invited Talk

Thomas Cheatham (University of Utah)

Molecular dynamics simulation of nucleic acids: Convergence, reproducibility, assessment/validation, and data dissemination enabled by GPUs on XSEDE and Blue Waters

Advances in simulation methods, continued optimizations of atomistic force fields, and increases in available computational power have coupled together over the past few decades to transform our understanding of biomolecular structure, dynamics, interactions and function. Such methods can aid in the design of new drugs, monitor how RNA and protein molecules fold, and aid in the interpretation of experimental data. Development of advanced “ensemble-based” sampling methodologies in the AMBER suite of programs, such as multidimensional replica-exchange molecular dynamics methods, running on large CPU and GPU resources has pushed the capabilities further. We are now able to reproducibly converge (in independent runs) the conformational ensembles of RNA tetranucleotides [1-2] and even RNA tetraloops. This provides a means to assess and validate the force fields and to test improvements or alterations to the force fields in a systematic manner. Using large ensembles of independent simulations poses significant data management and analysis issues which we continue address with further developments in the CPPTRAJ program optimization and parallelization [3]. To make better use of the data we are also exploring means to automatically mine and display the data, and also better means to search and disseminate the data (iBIOMES) [4]. All of these issues, developments and current barriers will be discussed in the context of large-scale use of XSEDE and Blue Waters computational resources.

1. Bergonzo, C., et al., Multidimensional REMD Yields a Converged Ensemble of an RNA Tetranucleotide. JCTC, 2014. 10: 492-499.
2. Roe, D.R., C. Bergonzo, and T.E. Cheatham III, Evaluation of Enhanced Sampling Provided By Accelerated Molecular Dynamics with Hamiltonian Replica Exchange Methods. J Phys Chem B, 2014. [in press]
3. Roe, D.R. and T.E. Cheatham III, PTRAJ and CPPTRAJ: Software for processing and analysis of molecular dynamics trajectory data. JCTC, 2013. 9: 3084-3095.
4. Thibault, J.C., J.C. Facelli, and T.E. Cheatham, 3rd, iBIOMES: managing and sharing biomolecular simulation data in a distributed environment. JCIM, 2013. 53: 726-36.