Table 6.

A comparison of PseUI with iRNA-PseU and re-iRNA-PseU on three training datasets

Training datasets	Predictor	Sen (%)	Spe (%)	Acc (%)	MCC	AUC
H_990	iRNA-PseUa	61.01	59.80	60.40	0.21	0.64
	re-iRNA-PseUb	65.05	58.79	61.92	0.24	0.65
	PseUIc	64.85	63.64	64.24	0.28	0.68
S_628	iRNA-PseUa	64.65	64.33	64.49	0.29	0.81
	re-iRNA-PseUb	66.88	64.33	65.61	0.31	0.69
	PseUIc	62.10	71.02	66.56	0.33	0.69
M_944	iRNA-PseUa	73.31	64.83	69.07	0.38	0.75
	re-iRNA-PseUb	79.87	60.81	70.34	0.41	0.75
	PseUIc	74.58	66.31	70.44	0.41	0.77

^aThe predictor developed by Chen et al. [14]

^bThe predictor we re-implemented by the method proposed by Chen et al. [14]

^cThe predictor proposed in this paper