Michael Colvin
Biology and Biotechnology Research Program
Lawrence Livermore Labs
Livermore, CA

Monday, March 16, 1998
3:00 pm (CT)
3269 Beckman Institute

Abstract

In the conditions prevalent on earth, chemical reactions mediate much of the physical change, most of the energy-transduction processes, and all known biology. Although the quantum mechanical equations needed to predict the properties of these ubiquitous chemical reactions have been known for more than fifty years, their computational intractability prohibited the widespread application of ab initio (from first principles) chemical modeling to real-world problems. With the advent of supercomputers, ab initio quantum chemistry became feasible for molecules containing up to many dozens of atoms and has since had a dramatic impact on chemistry, biochemistry, and materials science. Nevertheless, before quantum chemistry can fulfill its ultimate promise as a true "virtual chemistry laboratory", many advances must be made. In order to use the vast range of modern computing hardware, from workstations to massively parallel computers, we have written an ab initio quantum chemistry program that runs efficiently on both serial processors and shared- and distributed-memory parallel computers. We have used this software on massively parallel computers at Lawrence Livermore and Sandia National Laboratories to study several real-world projects in biochemistry, including joint theoretical/experimental collaborations on anticancer drugs, environmental mutagens, and modified DNA oligomers for antisense therapies. These projects involved performing some of the largest ab initio quantum chemical calculations ever done. In this talk I will describe the role our computer modeling has played in these biochemical studies and discuss the capabilities of state-of-the art quantum chemical methods.

Tea and coffee will be served in R3151 Beckman Institute at 2:15pm.