Skip to main content

Table 6 Five-fold cross-validation results of the models trained with the combination of hybrid features obtaining best predictive performance in training datasets

From: Investigation and identification of protein carbonylation sites based on position-specific amino acid composition and physicochemical features

Residue Classifier Hybrid features Sn Sp Acc MCC
K SVM PWM + AAC + AAindex 0.796 0.767 0.771 0.432
R SVM PWM + AAindex + AAPC 0.782 0.798 0.795 0.472
T SVM PWM + AAindex 0.750 0.795 0.788 0.443
P RF PWM + AAC + AAindex 0.787 0.777 0.779 0.467