A comprehensive survey of scoring functions for protein docking models

Shirali, Azam; Stebliankin, Vitalii; Karki, Ukesh; Shi, Jimeng; Chapagain, Prem; Narasimhan, Giri

doi:10.1186/s12859-024-05991-4

BMC Bioinformatics

Table 2 Success rates of all twelve scoring functions on eight different datasets. The first four methods are DL-based, and the rest are classical methods (details in Methods) section

From: A comprehensive survey of scoring functions for protein docking models

Dataset	Method	Top1	Top10	Top25	Top100	Top200
CAPRI Score v2022 Difficult (39,130 models)	DeepRank-GNN	14	35	33	65	78
	GNN-DOVE	0	7	28	42	57
	dMaSIF	14	32	46	78	85
	PIsToN	17	28	46	71	78
	FireDock	0	0	3	28	46
	AP-PISA	0	24	17	46	57
	CP-PIE	7	25	39	60	85
	PyDock	0	7	7	35	46
	ZRANK2	0	14	17	46	53
	RosettaDock	3	10	25	46	60
	SIPPER	3	14	35	53	60
	HADDOCK	3	14	21	46	60
CAPRI Score v2022 Easy (41,191 models)	DeepRank-GNN	35	79	82	92	100
	GNN-DOVE	15	48	56	76	84
	dMaSIF	51	71	89	97	100
	PIsToN	41	82	87	94	97
	FireDock	5	23	35	53	69
	AP-PISA	15	33	51	79	87
	CP-PIE	35	64	74	97	97
	PyDock	5	17	25	58	76
	ZRANK2	20	41	61	87	97
	RosettaDock	23	43	64	87	97
	SIPPER	25	58	74	92	100
	HADDOCK	15	69	71	84	97
CAPRI Score Refined (16,666 models)	DeepRank-GNN	23	46	53	69	84
	GNN-DOVE	15	61	69	76	100
	dMaSIF	15	38	53	76	92
	PIsToN	38	69	76	76	100
	FireDock	0	0	7	23	46
	AP-PISA	0	0	7	46	53
	CP-PIE	23	46	69	69	84
	PyDock	0	0	0	7	30
	ZRANK2	7	38	53	61	61
	RosettaDock	7	38	53	69	76
	SIPPER	7	46	53	69	76
	HADDOCK	7	23	53	69	76
BM4 (7,600 models)	DeepRank-GNN	21	73	73	89	94
	GNN-DOVE	21	68	84	100	100
	dMaSIF	47	68	89	94	100
	PIsToN	54	89	89	94	100
	FireDock	26	84	89	89	94
	AP-PISA	26	78	89	94	94
	CP-PIE	52	78	84	100	100
	PyDock	15	63	84	94	94
	ZRANK2	42	78	89	94	94
	RosettaDock	42	73	89	94	94
	SIPPER	47	63	68	78	94
	HADDOCK	26	89	94	94	100
BM5 (7,500 models)	DeepRank-GNN	93	100	100	100	100
	GNN-DOVE	6	26	33	53	73
	dMaSIF	40	80	100	100	100
	PIsToN	66	93	100	100	100
	FireDock	0	0	0	6	13
	AP-PISA	0	6	6	13	26
	CP-PIE	93	100	100	100	100
	PyDock	0	0	0	6	6
	ZRANK2	0	0	6	6	20
	RosettaDock	13	46	53	66	80
	SIPPER	60	93	93	100	100
	HADDOCK	0	0	0	6	26
Dockground (6,725 models)	DeepRank-GNN	37	53	72	100	100
	GNN-DOVE	70	91	94	100	100
	dMaSIF	34	75	87	100	100
	PIsToN	15	55	81	100	100
	FireDock	1	18	48	98	100
	AP-PISA	1	6	17	98	100
	CP-PIE	22	56	75	100	100
	PyDock	0	8	20	93	100
	ZRANK2	5	25	53	100	100
	RosettaDock	3	31	56	96	100
	SIPPER	43	72	89	100	100
	HADDOCK	10	44	65	100	100
PDB 2023 (5,300 models)	DeepRank-GNN	36	48	67	100	100
	GNN-DOVE	3	19	34	100	100
	dMaSIF	51	84	94	100	100
	PIsToN	88	96	98	100	100
	FireDock	0	3	11	100	100
	AP-PISA	0	1	5	100	100
	CP-PIE	76	92	96	100	100
	PyDock	0	0	3	100	100
	ZRANK2	0	11	23	100	100
	RosettaDock	3	17	32	100	100
	SIPPER	50	75	84	100	100
	HADDOCK	0	1	1	100	100

Bold values indicate the best value for that column

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com