Integration of Quantum Chemistry and Graph Theory for Structure Activity Relations and Predictive Toxicology. K. Balasubramanian, University of California-Davis, Lawrence Livermore National Laboratory, CA
The objective is to outline both graph-theoretically derived indices and quantum chemically derived indices for QSAR. It is demonstrated that there is an intertwining relation between the two techniques and cross-fertilization can be highly beneficial for more accurate descriptors that are particularly sensitive to the electronic and stereochemical features of the molecules. Graph-theoretical indices such as the Wiener index, topological indices, connectivity index, and molecular ID indices and stereo-chemically based and geometrical indices will be considered. Structural indices derived from quantum chemical computations such as the charge densities, Mulliken overlaps, populations, frontier orbital indices, hardness, free valence density matrices, quantum spectral indices etc., will be discussed. The use of electron static potentials and charge densities for prediction of reactivity will be discussed. These computations will be demonstrated with theoretical studies of structures of neutral molecules and their anions as well as dissociate electron attachment properties for halomethanes of general formula CXnYm; X=H,F; Y=Cl,Br,I; n=0,4; m=4-n. These results demonstrate that the dissociative electron attachment is the primary process resulting in toxicity of these species.
Go to . . .
2001 SRA Annual Meeting Table of Contents
2001 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page