The atomic-scale interactions of solute and solvent molecules determine the fundamental properties of materials as well as the pathways of important biological functions. Detailed molecular dynamics simulations can be used to determine these fundamental properties and pathways. Such simulations require accurate force field models to enable molecular structure prediction and binding regimes, and accurate geometry- and biochemistry-based trajectory analyses. However, detailed molecular dynamics simulations are often too time-consuming to become the basis of computational measurements that can be integrated into the research workflows used in developing new materials (e.g., carbon nanotubes) or in determining disease pathologies (e.g., Alzheimer’s disease). We want to explore solutions that combine parallel architectures (i.e., GPUs or accelerator boards, clusters) and numerical algorithms suited to such architectures with the goal of improving the speed of convergence and the stability of the numerical solutions. The project is expected to produce a set of computational measurements, models, and simulation protocols that will lead to the accurate molecule and microstructure interaction and dynamics prediction in various environments.

 

key words

Molecular dynamics simulations; Parallel algorithms; Spatial data structures; Bioinformatics;

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral applicants