Large-Scale Optimization Methods With a Focus on Chemistry Problems
Abstract
The objective of this research is to develop large scale optimization methods for optimization problems that arise in molecular chemistry. The main applications that are being targeted are problems whose solution is of direct and immediate interest to the Air Force, such as finding the structure of proteins and polymers. The primary optimization problem being considered is the large scale global optimization problem, via which protein structure can be determined. During this research period we have made major advances in the applicability of our methodology. The culmination of these advances has been our participation over the summer in the fourth CASP (Critical Assessment of Techniques for Protein Structure Prediction) competition, where we have attempted blind prediction of eight proteins of arbitrary complexity including mixed alpha helices and beta sheets, and of size ranging up to 242 amino acids. Enabling us to reach this stage have been advances in the past year in three key areas. The foremost of these is the ability to handle beta sheets in our algorithm, including the development of new biasing techniques. Second is improvements to the main portion of our global optimization method to enable it to better select the portions of the protein to work on in the small scale global optimizations. Third, in conjunction with our collaborators at Berkeley, is continued improvement of the ability of the energy function to accurately distinguish correct folds from misfolds, through the treatment of hydration.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 19, 2001
- Accession Number
- ADA387837
Entities
People
- Richard H. Byrd
- Robert B. Schnabel
Organizations
- University of Colorado Boulder