DTi2Vec: Drug–target interaction prediction using network embedding and ensemble learning

Table 4 Prediction performances for DTi2Vec and all comparison methods across all benchmark datasets

Dataset	Metric	Method
Dataset	Metric	NRLMF	DNILMF	DDR	TriModel	DTiGEMS+	DTi2Vec
Yamanishi_08 datasets	*AvgAUPR*	0.80	0.78	0.87	0.88	0.92	0.95
Yamanishi_08 datasets	AvgAUC	0.95	0.95	0.96	0.98	0.99	1.00
All datasets (Yamanishi_08 and FDA_DrugBank)	*AvgAUPR*	0.72	0.69	0.82	0.84	0.86	0.93
All datasets (Yamanishi_08 and FDA_DrugBank)	AvgAUC	0.94	0.95	0.96	0.98	0.99	0.99
FDA_DrugBank (Hold-out test set)	*AUPR*	0.34	0.31	0.63	0.66	0.62	0.82
FDA_DrugBank (Hold-out test set)	AUC	0.93	0.95	0.97	0.99	0.97	0.99

We rounded off all results to two decimal places. The bold underlined font indicates the best result in each category, while underlined values indicate the second-best outcome

ISSN: 1758-2946