SVM Regression QSAR for the Phenol Toxicity to Tetrahymena pyriformis

Aptula et al. used multiple linear regression to investigate the toxicity of 200 phenols to the ciliated protozoan Tetrahymena pyriformis Using their MLR model, they then predicted the toxicity of another 50 phenols. Here we present a comparative study for the entire set of 250 phenols, using multiple linear regression, artificial neural networks, and SVM regression methods. Before computing the SVM model, the input vectors were scaled to zero mean and unit variance. The prediction power of the QSAR models was tested with complete cross-validation leave-5%-out (L5%0), leave-10%-out (L10%O), leave-20%-out (L20%O), and leave-25%-out (L25%0). The capacity parameter C was optimized for each SVM model. 

The SVM regression results for the prediction of phenol toxicity to Tetrahymena pyriformis are presented in Tables 11 and 12. In calibration 

5%-out q RMSEl5%o leave-5 %-out root-mean-square error L, linear kernel P, polynomial kernel (parameter degree d) R, radial basis function kernel (parameter y) N, neural kernel (parameters a and b) and A, anova kernel (parameters y and d). 

In Table 13, we present the best regression predictions for each kernel. Despite the large number of SVMR experiments we carried out for this QSAR (34 total), the cross-validation statistics of the SVM models are well below those obtained with MLR. 

SVM Regression QSAR for the Toxicity of Aromatic Compounds to CMorella vulgaris 

---