Predicting potential drug-drug interactions by integrating chemical, biological, phenotypic and network data

BMC Bioinformatics

Table 2 Performances of different models evaluated by 20 runs of 5-CV

Method	Similarity	Index	AUC	AUPR	Recall	Precision	Accuracy	F
Neighbor recommender Method	Substructure	1	0.936	0.759	0.765	0.617	0.950	0.683
	Target	2	0.820	0.365	0.338	0.548	0.867	0.418
	Transporter	3	0.714	0.329	0.290	0.389	0.862	0.331
	Enzyme	4	0.756	0.377	0.471	0.346	0.909	0.399
	Pathway	5	0.812	0.571	0.657	0.474	0.932	0.550
	Indication	6	0.912	0.599	0.555	0.591	0.923	0.572
	Label	7	0.936	0.754	0.750	0.618	0.949	0.678
	Off label	8	0.940	0.768	0.765	0.629	0.951	0.691
	CN	9	0.941	0.767	0.745	0.635	0.949	0.685
	AA	10	0.941	0.767	0.747	0.634	0.949	0.686
	RA	11	0.943	0.770	0.752	0.634	0.950	0.688
	Katz	12	0.937	0.735	0.707	0.608	0.944	0.653
	ACT	13	0.931	0.752	0.723	0.618	0.947	0.667
	RWR	14	0.941	0.766	0.746	0.634	0.949	0.685
Random walk Method	Substructure	15	0.936	0.758	0.763	0.616	0.950	0.681
	Target	16	0.852	0.559	0.596	0.501	0.927	0.544
	Transporter	17	0.713	0.363	0.297	0.381	0.864	0.329
	Enzyme	18	0.760	0.470	0.657	0.344	0.927	0.451
	Pathway	19	0.811	0.594	0.709	0.479	0.937	0.572
	Indication	20	0.941	0.777	0.768	0.641	0.952	0.699
	Label	21	0.936	0.760	0.764	0.621	0.950	0.685
	Off label	22	0.937	0.763	0.761	0.627	0.950	0.688
	CN	23	0.938	0.757	0.736	0.625	0.948	0.676
	AA	24	0.938	0.755	0.734	0.624	0.947	0.675
	RA	25	0.937	0.748	0.729	0.616	0.946	0.667
	Katz	26	0.937	0.750	0.730	0.619	0.946	0.669
	ACT	27	0.930	0.748	0.727	0.632	0.938	0.671
	RWR	28	0.939	0.764	0.742	0.635	0.949	0.684
Matrix perturbation method		29	0.948	0.782	0.755	0.666	0.952	0.707

ISSN: 1471-2105