Timezone: »

 
Spotlight
Resampling Methods for Protein Structure Prediction with Rosetta
Ben Blum · David Baker · Michael Jordan · Philip Bradley · Rhiju Das · David Kim

Mon Dec 03 08:10 PM -- 08:25 PM (PST) @ None

Rosetta is one of the leading algorithms for protein structure prediction today. It is a Monte Carlo energy minimization method requiring many random restarts to find structures with low energy. In this paper we present a resampling technique for structure prediction of small alpha/beta proteins using Rosetta. From an initial round of Rosetta sampling, we learn properties of the energy landscape that guide a subsequent round of sampling toward lower-energy structures. Rather than attempt to fit the full energy landscape, we use feature selection methods--L1-regularized linear regression--to identify structural features that give rise to low energy. We then enrich these structural features in the second sampling round. Results are presented across a benchmark set of nine small alpha/beta proteins demonstrating that our methods seldom impair, and frequently improve, Rosetta's performance.

Author Information

Ben Blum (UC Berkeley)
David Baker (University of Washington)

Dr. Baker is a Professor of Biochemistry at the University of Washington, Seattle. He received his Ph.D. degree in biochemistry from the University of California, Berkeley, where he worked with Randy Schekman. His postdoctoral work in biochemistry and biophysics was done with David Agard at the University of California, San Francisco. Dr. Baker has received young investigator awards from the Packard Foundation, the National Science Foundation, and the Beckman Foundation; the Irving Sigal Young Investigator Award from the Protein Society; and the Overton Award from the International Society of Computational Biology. He is a recipient of the Feynman Prize from the Foresight Institute and the AAAS Newcomb Cleveland Prize, and was recently elected a member of the National Academy of Sciences. David Baker uses a combination of experimental and computational approaches to understand the basic principles underlying protein folding and protein-protein interactions. He is applying this knowledge to the prediction and design of macromolecular structures and interactions.

Michael Jordan (UC Berkeley)
Philip Bradley
Rhiju Das (University of Washington)
David Kim

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors