IGPred-HDnet: Prediction of Immunoglobulin Proteins Using Graphical Features and the Hierarchal Deep Learning-Based Approach

Table 2

Analysis of various classifiers using feature encoding schemes on training and testing datasets D_train and D_test.

Dataset	Predictor	ACC	SN	SP	MCC	Pre	NPV	F1	AUC
Training	CC_PSSM	0.960	—	—	0.921	0.961	—	—	0.994
	IGPred	0.969	0.963	0.975	—	—	—	—	0.994
	XGBoost	0.972	0.945	0.985	0.950	0.980	—	—	0.970
	IGPred-HDnet	0.980	0.945	1.000	0.958	1.000	1.000	0.971	0.998
Testing	CC_PSSM	0.883	—	—	0.847	0.884	—	—	0.914
	IGPred	0.891	0.886	0.897	—	—	—	—	0.914
	XGBoost	0.894	0.869	0.906	0.874	0.902	—	—	0.892
	IGPred-HDnet	0.991	1.000	0.986	0.980	0.9750	1.000	0.987	1.000