DescriptionThe molecules of life are large, complex machines that drive the operations of the cell. Modeling the atomic structure and underlying dynamics of these molecules is critical for understanding disease and developing therapeutics. Recent advances in experimental instrumentation and scientific software have have made high resolution biological structures more accessible than ever, but large data volumes and complex calculation often require extensive computation. Cryo-electron microscopy (Cryo-EM), for example, can now reveal structures from heterogeneous biological samples to atomic resolution - better than 3 angstroms. These structures require terabytes of experimental data and upwards of 20,000 hours of compute time for accurate determinations to be made. X-ray crystallography, the workhorse of structural biology, can now combine complex computational modeling algorithms like Rosetta with experimental data to arrive at a complete structure determination, which may require more than 1000 hours of compute time. Drug discovery efforts have embraced computational “virtual screening’ to filter the most likely targets from vast drug fragment libraries by combining computational chemistry and experimentally-determined molecular structures. In these screens, the only limit to the number of drug candidates screened is computational time. Finally, molecular dynamics simulations give insight into the motions biological molecules adopt as they perform their jobs, but also require high-performance computing resources for meaningful results. As a Campus Champion at Harvard Medical School and SBGrid, I support the research computing needs of a diverse structural biology community and XSEDE is an essential resource in driving this critical research.
OrganizationHarvard Medical School
DepartmentBCMP / SBGrid
Sponsor Campus GridOSG-XSEDE
Principal Investigator
Jason Key
Field Of ScienceMolecular and Structural Biosciences