| College of Sciences Newsletter | Edition 21 | June 10, 2004 |
| Back to Research News Math prof's research on molecular structure modeling gets boost from 2003 Summer Faculty Fellowship award
Glenn Williams, assistant professor of mathematics and statistics, was one of nine faculty awarded an ODU summer fellowship last year to further his research on the "Efficient Use of Secondary Structure Constraints in Molecular Structure Modeling." The funding, though small by granting-agency standards, was timely and helped him take his research to the next level. "I am very encouraged with the results from my research this summer. The fellowship enabled me to focus my efforts and make good progress," said Williams, who joined Old Dominion in 2002 from Stanford University Medical Center where he was a postdoctoral research fellow. "The results from my research will not only form the basis of several manuscripts, but will also provide a foundation from which I can foster research collaborations and effectively compete for future funding." As research on the human genome project produces large amounts of information about the atomic makeup of individual molecules, methods for determining how the atoms within molecules arrange themselves to form three-dimensional structures become critical. "The determination of molecular structure is critical for many pursuits in biology and medicine, including the study of biomolecular function and the design of drugs to remove, modify or enhance this function," Williams said. Williams' novel computer algorithm for estimating three-dimensional molecular structure, called GNOMAD, has proven accurate for determining separation distances between pairs of atoms in a molecule. Over the summer, Williams focused initially on integrating secondary structure elements of molecules into the GNOMAD code. He developed a method for accurately and efficiently constraining the dihedral angle of the group of atoms that determines secondary structure in protein molecules. "Adding constraints to a nonlinear optimization algorithm typically hinders global convergence, so it was encouraging to see that the new method actually serves to aid in global convergence," said Williams. An article on his project is under review by the Journal of Computational Biology and a manuscript on reduced atomic representation for proteins is being considered for the International Conference for Mathematics in Biology and Medicine in Ann Arbor, Mich., this year. Williams is also preparing a proposal for a National Institutes of Health program, "Innovations in Biomedical Computational Science and Technology," which could result in a $100,000 grant. This would allow him to continue his work in protein structure prediction and apply his methods to RNAstructure prediction. Researchers at Eastern Virginia Medical School have expressed interest in his RNA modeling and he hopes to develop a collaboration with them. - Michelle Nery, Media Relations Other stories in Research News Section..
Phyllis
Brown, Editor |