Analysis Methods for Biomolecular Simulations (Biophysics)
Parallel computing has revolutionized computational biology and, more specifically, molecular dynamics simulation by allowing systems of increasing size and biomedical relevance to be simulated using clusters and supercomputers.
Simulations of biological molecules on parallel computers record the time-varying motions of the atoms in "trajectory" files, frequently terabytes in size. The relationships between atoms and their motions are the key to the discovery of how these biological molecules function. The large size typical of molecular dynamics trajectories and the computational complexity of common analysis algorithms can hinder scientific progress when performed with sequential software. This project involves the parallelization of existing sequential molecular dynamics trajectory analysis methods. Ideally, these researchers would even be able to manipulate simulations interactively that they previously could only observe the end result.
This has huge and far ranging benefits as all scientists who utilize and depend on these tools would receive a huge boost in productivity from the speedup hopefully in turn answering important biological questions.