BMC Bioinformatics

Table 4 Performance comparison over five datasets in the S^T task

From: DTI-HeNE: a novel method for drug-target interaction prediction based on heterogeneous network embedding

Methods	Performance indicators	Datasets
Methods	Performance indicators	E	IC	GPCR	NR	DrugBank
DTI-HeNE	PR-AUC	0.941	0.974	0.914	0.989	0.429
	ROC-AUC	0.997	0.997	0.976	0.996	0.891
NEDD	PR-AUC	0.929	0.901	0.876	0.853	0.421
	ROC-AUC	0.992	0.982	0.995	0.983	0.881
DDR	PR-AUC	0.924	0.947	0.862	0.818	0.486
	ROC-AUC	0.974	0.987	0.964	0.929	0.885
NRLMFβ	PR-AUC	0.797	0.791	0.527	0.541	0.268
	ROC-AUC	0.931	0.954	0.939	0.921	0.766
DTINet	PR-AUC	0.477	0.425	0.093	0.272	0.176
	ROC-AUC	0.895	0.860	0.681	0.676	0.841
CMF	PR-AUC	0.273	0.365	0.402	0.366	0.104
	ROC-AUC	0.765	0.754	0.809	0.533	0.702
BLM-NII	PR-AUC	0.650	0.738	0.352	0.418	0.158
	ROC-AUC	0.911	0.914	0.775	0.533	0.831
NetLapRLS	PR-AUC	0.651	0.708	0.302	0.348	0.140
	ROC-AUC	0.907	0.912	0.758	0.523	0.810
BiNE	PR-AUC	0.674	0.612	0.432	0.459	0.183
	ROC-AUC	0.936	0.919	0.831	0.651	0.841

Best performing methods under the current dataset and performance indicator are indicated in bold

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com