Prediction of G Protein-Coupled Receptors with SVM-Prot Features and Random Forest

Table 5

Performance qualities measure for test dataset by using the models from the corresponding training dataset.

Test dataset	Sn	Sp	Acc	MCC	AUC
1st	0.5952	0.9812	0.7882	0.6248	0.930
2nd	0.5832	0.9807	0.7820	0.6146	0.909
3rd	0.6013	0.9620	0.7817	0.5763	0.879
4th	0.7675	0.9726	0.8700	0.7562	0.943
5th	0.9238	0.9654	0.9446	0.8900	0.980
Mean ± SD	0.6942 ± 0.1491	0.9724 ± 0.0087	0.8333 ± 0.0726	0.6924 ± 0.1296	0.928 ± 0.038

AUC, also called receiver operating characteristic (ROC) area, means the area under the receiver operating characteristic curve which is a measure of the accuracy of a classification model.