The Quantitative Structure-Mutagenicity Relationship of Polycylic Aromatic Hydrocarbon Metabolites

Quantitative structure-activity relationships (QSARs) for benz[a]anthracene (BA)mutagens using 73 descriptors were searched. The mutagenicity data was obtained fromAmes assays for Mycobacterium vanbaalenii, Mycobacterium gilvum and Mycobacteriumflavescens strains. These data were fitted using a mutagenicity-cytotoxicity competitionmodel which defines the mutagenic potencies of BA metabolites, and include oxides,phenols, quinones, and dihydrodiols. The QSAR equations were derived using the moleculardescriptor set (charged partial surface area, spatial, thermodynamic and electronicdescriptors) and semi-empirical energetic and charge descriptors. Genetic functionapproximation was used to reduce and fit independent variables, including linear- andquadratic-based functions. Multiple QSAR equations were generated and a separate QSARequation was chosen and evaluated for each strain using conventional r2, F-test, and cross-validated r2. Each strain exhibited its own characteristic descriptors.