MCL-CAw: a refinement of MCL for detecting yeast complexes from weighted PPI networks by incorporating core-attachment structure

Table 1 Properties of the PPI networks used for the evaluation of MCL-CAw

PPI Network	# Proteins	# Interactions	Avg node degree
Gavin	1430	7592	10.62
Krogan 'Core'	2708	7123	5.26
Gavin+Krogan	2964	13507	9.12
ICD(Gavin+Krogan)	1628	8707	10.69
FSW(Gavin+Krogan)	1628	8688	10.67
Consolidated_3.19	1622	9704	11.96
Bootstrap_0.094	2719	10290	7.56
Inferred	954	11892	24.93
Gavin+Krogan+Inferred	3418	25352	14.83
ICD(Gavin+Krogan+Inferred)	2034	12009	11.81
FSW(Gavin+Krogan+Inferred)	1892	11705	12.37

The Gavin+Krogan network was generated by combining the Gavin and Krogan-Core datasets, obtained from BioGrid [32]. The ICD(Gavin+Krogan) and FSW(Gavin+Krogan) networks were generated by scoring the Gavin+Krogan network using the Iterative-CD^k and FS Weight^k schemes [14, 15] (with k = 2 iterations). The Consolidated_3.19 network refers to the high-confidence subset (PE cut- off: 3.19) of the Consolidated network derived by Collins et al. [11]. The Bootstrap_0.094 network refers to the high-confidence subset (BT cut-off: 0.094) of the Bootstrap network derived by Friedel et al. [17]. The Inferred network comprised of computationally inferred interactions from the Predictome database [40]. The Gavin+Krogan+Inferred network was generated by augmenting the Gavin+Krogan network with these inferred interactions. The ICD(Gavin+Krogan+Inferred) and FSW(Gavin+Krogan+Inferred) networks were generated by scoring this augmented network using the Iterative-CD and FS Weight systems, respectively.

ISSN: 1471-2105