QSAR-based toxicity classification and prediction for single and mixed aromatic compounds

Abstract

Quantitative structure-activity relationships (QSARs) based on the octanol/water partition coefficient were employed to predict acute toxicities of 36 substituted aromatic compounds and their mixtures. In this study, the model developed by Verhaar et al. was modified and used to calculate octanol/water partition coefficients of chemical mixtures. To validate the model, acute toxicities of these chemicals were measured to Vibrio fischeri in terms of EC⁵⁰. The results indicated that the obtained QSAR models could be used to predict toxicities of samples consisting of these substituted aromatic compounds, individually or in combinations. The obtained equations were proved to be robust enough by using the leave-one-out test method. By classifying these chemicals into two groups, polar and non-polar, the toxicities of chemical mixtures within each group can be predicted accurately from their calculated partition coefficients.

Keywords:

• QSAR
• Vibrio fischeri
• Mixture toxicity
• Aromatic compounds
• Octanol/water partition coefficient

Acknowledgements

We are grateful to James Wang (U.S. EPA) for his valuable contributions. This study was funded by the Major State Basic Research Development Program of China (G1999045711), National Natural Science Fund of China (20277025), and China Postdoctoral Science Foundation.