. 2010 Oct 12;11:504. doi: 10.1186/1471-2105-11-504

Table 3.

(i) Impact of core-attachment refinement on MCL; (ii) Role of affinity scoring in reducing the impact of natural noise on MCL and MCL-CAw

Benchmark	Method	PPI Network	#Predicted complexes	#Matched prediction	Precision	#Derivable benchmarks	#Derived benchmarks	Recall
Wodak	MCL	G+K	242	55	0.226	182	62	0.338
		ICD(G+K)	136	68	0.500	153	76	0.497
		FSW(G+K)	120	69	0.575	153	78	0.510
		Consol_3.19	116	70	0.603	145	79	0.545
		Boot_0.094	203	76	0.374	172	85	0.494

	MCL-CAw	G+K	310	77	0.248	182	77	0.423
		ICD(G+K)	129	80	0.620	153	80	0.523
		FSW(G+K)	117	72	0.615	153	83	0.542
		Consol_3.19	122	82	0.672	145	82	0.566
		Boot_0.094	199	79	0.397	172	88	0.512

MIPS	MCL	G+K	242	35	0.143	177	40	0.226
		ICD(G+K)	136	47	0.346	151	60	0.397
		FSW(G+K)	120	46	0.383	151	61	0.404
		Consol_3.19	116	48	0.414	157	63	0.401
		Boot_0.094	203	44	0.271	168	56	0.333

	MCL-CAw	G+K	310	53	0.171	177	53	0.300
		ICD(G+K)	129	63	0.488	151	63	0.417
		FSW(G+K)	117	48	0.410	151	66	0.437
		Consol_3.19	122	68	0.557	157	68	0.433
		Boot_0.094	199	47	0.236	168	59	0.351

Aloy	MCL	G+K	242	43	0.179	76	42	0.556
		ICD(G+K)	136	58	0.426	75	56	0.747
		FSW(G+K)	120	57	0.475	75	57	0.760
		Consol_3.19	116	54	0.466	76	55	0.724
		Boot_0.094	203	56	0.276	76	55	0.724

	MCL-CAw	G+K	310	52	0.168	76	52	0.684
		ICD(G+K)	129	59	0.457	75	59	0.787
		FSW(G+K)	117	60	0.513	75	60	0.800
		Consol_3.19	122	57	0.467	76	57	0.750
		Boot_0.094	199	57	0.286	76	58	0.763

Affinity scoring of PPI networks improved the performance of MCL and MCL-CAw. Affinity scoring followed by CA refinement had a compounded effect in improving the performance of MCL.