Skip to main content

Table 2 Five-fold cross validation results of the DT, RF and SVM models trained with multiple types of features

From: iDPGK: characterization and identification of lysine phosphoglycerylation sites based on sequence-based features

Training feature Classifier Sensitivity (%) Specificity (%) Accuracy (%) MCC
AAC + AAPC DT 53.9 50.6 51.7 0.04
RF 58.4 60.1 59.6 0.18
SVM 53.9 57.3 56.2 0.11
AAC + B62 DT 42.7 66.3 58.4 0.09
RF 59.6 59.0 59.2 0.18
SVM 68.5 34.8 46.1 0.03
AAC + PSSM DT 32.6 64.6 53.9 − 0.03
RF 59.6 59.0 59.2 0.18
SVM 39.3 59.6 52.8 − 0.01
AAPC + PSSM DT 31.5 69.7 56.9 0.01
RF 62.9 54.5 57.3 0.16
SVM 39.3 59.6 52.8 − 0.01
AAC + AAPC + B62 DT 40.4 63.5 55.8 0.04
RF 60.7 54.5 56.6 0.14
SVM 68.5 34.3 45.7 0.03
AAC + AAPC + PSSM DT 31.5 64.6 53.6 − 0.04
RF 62.9 62.9 62.9 0.24
SVM 69.7 39.3 49.4 0.09