An ensemble framework for identifying essential proteins

BMC Bioinformatics

Table 1 The number of common predicted proteins (overlap) among the top 100 proteins ranked by different centrality measures

Datasets	Centrality measures	BC		EC		SC		CC		#overlaps among five centrality measures
Datasets	Centrality measures	#overlap	#EP	#overlap	#EP	#overlap	#EP	#overlap	#EP	#overlap	#EP
PIN76K	DC	66	22	57	26	92	38	51	20	29	7
	BC			36	12	63	21	49	16
	EC					65	30	41	12
	SC							51	19
PIN24K	DC	84	36	57	29	97	43	71	31	43	21
	BC			47	24	82	36	64	28
	EC					56	29	67	29
	SC							70	31

Proteins are ranked according to the scores calculated by each centrality measure. Then top 100 ranked proteins for each centrality measure are selected as the candidates of essential proteins, out of which the number of true essential proteins are determined. #overlap represents the number of common predicted proteins among the top 100 ranked proteins by different centrality measures and #EP the number of true essential proteins among the overlaps

ISSN: 1471-2105