Table 4.

Overall AUC and AUPR values of different methods on imbalanced dataset for four species.

Evaluation	Method	Enzyme	Ion Channel	GPCR	Nuclear Receptor
AUC	NetLapRLS	0.972 ± 0.002	0.969 ± 0.003	0.915 ± 0.006	0.850 ± 0.021
	BLM-NII	0.978 ± 0.002	0.981 ± 0.002	0.950 ± 0.006	0.905 ± 0.023
	WNN-GIP	0.964 ± 0.003	0.959 ± 0.003	0.944 ± 0.005	0.901 ± 0.017
	KBMF2K	0.905 ± 0.003	0.961 ± 0.003	0.926 ± 0.006	0.877 ± 0.023
	CMF	0.969 ± 0.002	0.981 ± 0.002	0.940 ± 0.007	0.864 ± 0.026
	NRLMF	0.987± 0.001	0.989 ± 0.001	0.969± 0.004	0.950± 0.011
	DAWN	0.981 ± 0.004	0.990± 0.014	0.952 ± 0.009	0.906 ± 0.067
AUPR	NetLapRLS	0.789 ± 0.005	0.837 ± 0.009	0.616 ± 0.015	0.465 ± 0.044
	BLM-NII	0.752 ± 0.011	0.821 ± 0.012	0.524 ± 0.024	0.659 ± 0.039
	WNN-GIP	0.706 ± 0.017	0.717 ± 0.020	0.520 ± 0.021	0.589 ± 0.034
	KBMF2K	0.654 ± 0.008	0.771 ± 0.009	0.578 ± 0.018	0.534 ± 0.050
	CMF	0.877 ± 0.005	0.923± 0.006	0.745 ± 0.013	0.584 ± 0.042
	NRLMF	0.892 ± 0.006	0.906 ± 0.008	0.749 ± 0.015	0.728± 0.041
	DAWN	0.895± 0.011	0.921 ± 0.036	0.786± 0.023	0.603 ± 0.087

Results excerpted from [12]. The best results in each column are in bold faces and the second best results are underlined. BLM-NII: improved BLM with neighbor-based interaction-profile inferring; CMF: collaborative matrix factorization; KBMF2K: kernelized Bayesian matrix factorization with twin kernels; NRLMF: neighborhood regularized logistic matrix factorization; WNN-GIP: weighted nearest neighbor with Gaussian interaction profile kernels.