Drug-target interactions prediction using marginalized denoising model on heterogeneous networks

BMC Bioinformatics

Table 8 AUC and AUPR for DNILMF and DTIP_MDHN on new dataset1 under CVP setting

Dataset	Method	AUPR	AUC
Enzyme^a	DNILMF	0.9245	0.9950
Enzyme^a	DTIP_MDHN	0.9071	0.9911
Ion Channel (IC)	DNILMF	0.9921	0.9991
Ion Channel (IC)	DTIP_MDHN	0.9968	0.9998
GPCR	DNILMF	0.9239	0.9935
GPCR	DTIP_MDHN	0.9615	0.9963
Nuclear Receptor (NR)	DNILMF	0.9341	0.9897
Nuclear Receptor (NR)	DTIP_MDHN	0.9610	0.9910
protein kinase	DNILMF	0.8713	0.9875
protein kinase	DTIP_MDHN	0.9408	0.9959
transporter	DNILMF	0.8852	0.9907
transporter	DTIP_MDHN	0.9523	0.9978
cytokine and cytokine receptor	DNILMF	0.8166	0.9827
cytokine and cytokine receptor	DTIP_MDHN	0.8630	0.9842
cell surface molecule and ligand	DNILMF	0.8557	0.9817
cell surface molecule and ligand	DTIP_MDHN	0.9076	0.9887
ALL^a	DNILMF	0.7578	0.9813
ALL^a	DTIP_MDHN	0.9743	0.9978

^aoptimized parameters (numLatent = 90, c = 20, thisAlpha = 0.7, λ_u = 10, λ_v = 10, K = 2) were used in DNILMF

ISSN: 1471-2105