Table 6.

The prediction performance of DEEPred and the state-of-the-art protein function prediction methods on CAFA3 challenge benchmark dataset.

	F-max		Precision		Recall (at F-max)		Smin
	No-knowledge	All	No-knowledge	All	No-knowledge	All	No-knowledge	All
MF
Naive	0.35	0.29	0.49	0.41	0.27	0.23	6.87	6.43
Blast	0.40	0.39	0.42	0.36	0.38	0.44	6.99	6.48
FFPred3	0.32	0.31	0.34	0.30	0.30	0.32	7.35	6.66
GoFDR*	0.55	0.45	0.67	0.55	0.46	0.38	5.06	4.41
DeepGO	0.40	0.34	0.58	0.48	0.30	0.27	6.36	6.01
DEEPred_raw	0.32	0.33	1.00	1.00	0.19	0.19	6.63	6.16
DEEPred_hrchy	0.49	0.50	1.00	1.00	0.32	0.33	5.41	5.03
CC
Naive	0.55	0.54	0.56	0.58	0.55	0.50	7.61	7.65
Blast	0.46	0.45	0.39	0.39	0.56	0.53	9.74	9.94
FFPred3	0.54	0.52	0.54	0.54	0.53	0.50	8.61	8.44
GoFDR*	0.48	0.45	0.46	0.42	0.51	0.48	10.98	10.86
DeepGO	0.54	0.53	0.61	0.58	0.48	0.49	7.68	7.55
DEEPred_raw	0.30	0.29	0.19	0.18	0.69	0.69	10.68	10.41
DEEPred_hrchy	0.34	0.35	1.00	1.00	0.20	0.22	9.85	9.53
BP
Naive	0.26	0.30	0.25	0.39	0.26	0.24	24.27	20.85
Blast	0.28	0.32	0.22	0.27	0.37	0.38	25.11	21.35
FFPred3	0.26	0.34	0.23	0.29	0.30	0.40	24.74	21.48
GoFDR*	0.19	0.18	0.25	0.26	0.15	0.14	24.75	28.83
DeepGO	0.28	0.34	0.40	0.52	0.21	0.26	23.41	20.19
DEEPred_raw	0.16	0.16	1.00	1.00	0.09	0.09	24.65	22.05
DEEPred_hrchy	0.32	0.33	1.00	1.00	0.19	0.19	22.04	19.69

*The results of GoFDR is given based on the CAFA3 preliminary benchmark set since the results for the full benchmark dataset were not available for this method.