Protein Folding Landscapes
he research group of Charles L. Brooks, III at The Scripps Research Institute is attempting to better understand how proteins self-assemble into their biologically functional forms by exploring the free-energy landscape for protein folding. Brooks and postdoctoral colleague Badry Bursulaya have explored the folding landscape (center) and sampled conformations of a small, designed all beta-sheet protein, Betanova. Such calculations can require months of processor time using the CHARMM simulation package on conventional high-end parallel platforms such as NPACI's CRAY T3E at SDSC. Information from simulations such as these may eventually provide input into the design of protein sequences with specific biological functions or help understand the origins of degenerative diseases such as Alzheimer's disease. Working together, NPACI partners Brooks, Michael Crowley of TSRI, Andrew Grimshaw from the University of Virginia, and their colleagues are developing and testing tools to combine the tightly coupled, parallel scalability of the CHARMM code with the Legion metasystem architecture. By doing so, emerging teraflops computer systems will be harnessed to efficiently perform free-energy landscape calculations for protein folding. This work was supported by the NIH, NSF, and PACI computing resources allocations. See the story on Legion and its applications.