SEMINAR

Parallel Computing in Chemistry

Computational Center for Molecular Structure and Interactions
Department of Chemistry
Jackson State University

ABSTRACT

While computational chemistry has made spectacular progress in the last couple of decades, the current state of the art nevertheless leaves both room and need for improvement. Many important problems in surface science, pharmacology, and materials science require a quantitative description of extended molecular systems. Since experimental information is scarce for many of these chemical systems, a computational approach would be particularly desirable. However, the ultimate breakthrough for computational methods in the study of such systems has been limited by the availability of inexpensive computing power.

This situation is now rapidly changing as the emergence of parallel computing appears to offer a remedy for these concerns. Over the past few years, the evolving computer hardware situation has allowed new areas of chemistry to be explored with computational methods. Many parallel architectures allow for very high nominal performance, measured in MFLOPS, at a very affordable cost. In general, however, the computational methods and algorithms traditionally used do not lend themselves well to efficient implementation on parallel hardware. Extensive development of algorithms and computer codes is therefore mandatory. Indeed, one of the most significant challenges facing contemporary computational quantum chemists involves the restructuring of application software to fully utilize current computer hardware. Considering the wide variety of available computer architectures and the ephemeral nature of cutting-edge technology upon which they are based, this is not a one-time task but rather an ongoing development project. As a result of such advances, molecular modeling has become an inseparable part of the research activities devoted toward achieving an understanding of the molecular bases of environmental, biochemical, and biological processes. Such techniques could also be used as tools to direct experimental work. Below are a few examples of the application of high performance computational chemistry methods to problems of practical importance.

Nitroaromatic compounds (NACs) are widely used as pesticides, herbicides, solvents, explosives, and intermediates in the synthesis of dyes. Many of these compounds and their transformation products are ubiquitous environmental pollutants of significant toxicological concern. Our recent studies contribute to a better understanding of the interactions of NACs with soil. Different realistic models of clusters representing clays were developed. Details of their interactions with explosives were studied by rigorous ab initio calculations. In addition, the process of reduction of nitroaromatic compounds on the surface of Fe was investigated using semiempirical approaches.

The current advances in computational techniques allow for the application of accurate ab initio calculations to problems directly related to biomedical research. They range from understanding the action of drugs to designing efficient antitoxins.

This talk will also present an overview of the vital areas where computational chemistry reveals new details of DNA structure and functionality. A theoretical investigation explains the role of cisplatin--the oldest anticancer drug--in destabilizing the guanine-cytosine (GC) hydrogen bonded base pair. This study concludes with a description of the unique role of the Pt cation in this process and attributes it to the strong modification of the electrostatic potential around guanine bonded to the Pt ion.

WHERE: TEC 251

WHEN(day): Friday, November 12th, 1999

WHEN(time): 2:00

EVERYBODY IS INVITED