|
1F - QSAR (TU13/6) Stepwise Discrimination between Toxic Mechanisms of Phenols in the Tetrahymena Pyriformis Assay. Aptula, Aynur1, Kühne, Ralph1, Ebert, Ralf-Uwe1, Schüürmann, Gerrit1, 1 Department of Chemical Ecotoxicology, UFZ Centre for Environmental Research, Permoserstrasse 15, Leipzig, Germany ABSTRACT- For a set of 220 phenols with literature data on their toxicity towards the ciliate Tetrahymena pyriformis, a stepwise classification scheme was developed that allows the identification of four mechanisms of action (MOAs) based on only few molecular descriptors. The initial MOA assignment with respect to polar narcosis, oxidative uncoupling, soft electrophilicity and pro-electrophilicity was done a priori using published structural rules, and both binary logistic regression (BLR) and linear discriminant analysis (LDA) were used as statistical tools to derive the classification model. For the model development, hydrophobicity and AM1 parameters of the geometric and electronic structure were employed as molecular descriptors. Taking the LUMO energy as only parameter, an initial separation of polar narcotics and pro-electrophiles from oxidative uncouplers and soft electrophiles yields a 90% correct classification. For the subsequent discrimination between polar narcotics and pro-electrophiles as well as between oxidative uncouplers and soft electrophiles, 99% and 98% correct classification are achieved using two and three molecular descriptors, respectively. BLR and LDA performed equally well, despite deviations of the data from normal distributions. Model validation was performed by evaluating the simulated external prediction through classification models built from complementary subsets. This work was supported in part by the European Union IMAGETOX Research Training Network (HPRN-CT-1999-00015). Key words: toxic mechanism, phenols, linear discriminant analysis, binary logistic regression |
