DASPfind: new efficient method to predict drug–target interactions

Table 1 Comparison between methods over six different datasets based on LOOCV for each known DTI

Method	AUC (%)	‘Top 1’ (%)	‘Top 2’ (%)	‘Top 5’ (%)
Enzyme
NRWRH	92.89	1.06	8.07	12.82
HGBI	91.60	2.36	8.1	12.41
DT-Hybrid	89.80	0	7.55	10.95
DASPfind	92.91	52.08	55.29	62.74
Ion channels
NRWRH	91.56	1.69	2.91	10.16
HGBI	88.93	1.42	2.24	6.1
DT-Hybrid	92	0	1.42	14.3
DASPfind	90.68	32.72	35.09	46.54
GPCR
NRWRH	84.93	2.52	11.50	40.94
HGBI	91.29	5.83	12.28	31.5
DT-Hybrid	83.87	0	6.93	31.65
DASPfind	88.10	46.61	51.18	64.4
Nuclear receptors
NRWRH	73.9	7.78	32.22	52.22
HGBI	87.57	15.56	42.22	57.78
DT-Hybrid	69.95	0	14.44	22.22
DASPfind	85.27	53.3	65.5	77.7
HGBI_Dataset
NRWRH	86.19	0	5.9	20.47
HGBI	89.07	0	5.17	16.19
DT-Hybrid	86.75	0	5.74	21.04
DASPfind	89.61	28.30	33.32	42.51
DrugBank_Approved
NRWRH	89.5	1.04	5.65	18.63
HGBI	80.10	2.11	4.36	11.32
DT-Hybrid	84.44	0.34	5.88	22.69
DASPfind	88.84	27.82	32.89	48.56

ISSN: 1758-2946