Predicting the Mechanism of Action from Hydrophobicity and Experimental Toxicity

Predicting the Mechanism of Action from Hydrophobicity and Experimental Toxicity 

In the first test we used SVM models to discriminate between nonpolar narcotic compounds (chemicals that have baseline toxicity) and other com-potmds having excess toxicity (representing the following MO As polar narcotics, ester narcotics, amine narcotics, weak acid respiratory uncouplers, electrophiles, proelectrophiles, and nucleophiles). From the total set of 337 compotmds, 126 represent the SVM class -Fl (nonpolar narcotic) and 211 represent the SVM class —1 (all other MOA classes). 

The best cross-validation results for each kernel type are presented in Table 9. The linear, polynomial, RBF, and anova kernels have similar results that are of reasonably quality, whereas the neural kernel has very bad statistics the slight classification improvement obtained for the RBF and anova kernels is not statistically significant. 

In Table 10, we show the best cross-validation results for each kernel type. The radial kernel has the best predictions, followed by the linear SVM model. The remaining kernels have worse predictions than does the linear model. 

---