Structure- and property-activity relationship models for prediction of microbial toxicity of organic chemicals to activated sludge.

Two quantitative structure-activity relationship (QSAR) and two quantitative property-activity relationship (QPAR) models reported in the literature for predicting toxicity of synthetic organic chemicals to activated sludge microorganisms are summarized and compared. The QSAR models were developed using solvatochromic parameters and molecular connectivity indices; the QPAR models, using octanol-water partition coefficient and aqueous solubility. Experimental data on 16 chemicals not used in developing the above models are used to compare and evaluate the predictive ability of these QSAR/QPAR models. Based on the quality of the original models, their predictions, ease of application, and availability of model parameters, molecular connectivity indices and log P appear to be the most suitable in toxicity predictions.