PLS and N-PLS-based MIA-QSTR modeling of the acute toxicities of phenylsulfonyl carboxylates to Vibrio fischeri

Abstract

Descriptors based on multivariate image analysis have been used to derive predictive QSTR models of the acute toxicities of phenylsulphonyl carboxylates to Vibrio fischeri. Classical and multilinear partial least squares, PLS and N-PLS, respectively, were applied as regression methods, demonstrating similar predictive capability to each other. Model performance was improved in c. 10% after removing an outlier, and the results were in general agreement with the ones previously obtained through CoMFA and extended topochemical atom indices analysis. Overall, this study showed that a simple procedure is able to give highly predictive models, useful in ecotoxicology, independent of the regression method used for this class of compounds.