How complete are current yeast and human protein-interaction networks?

Table 1 Yeast protein-interaction assay false-positive rates: yeast datasets

Dataset	Number of interactions	Derived false-positive rate* (%)	Published false-positive rate (%)	Average false-positive rate (%)
Uetz et al. [35]	854	46 [32]	32 [24]^†,47 [44], 50 [37], 51 [42]	45
Ito [36]	4,393	89 [32]	71 [24]^†, 78 [41], 85 [37], 91 [44]	83
Gavin et al. [16]	3,180	68 [32]	14 [24]^†, 22 [4], <72 (upper bound [20])	35
Ho et al. [17]	3,618	83 [32], 81, 82, 80	55 [24]^†, <97 (upper bound [20])	76
Jansen et al. [22]	15,922	81, 79	-	80
Gavin et al. [27]	18,137	78, 82, 86^‡	-	82
Krogan et al. [28]	14,317 (7,123 core)	75, 79, 66^‡ (59, 65, 37^‡ core)	-	73 (54 core)
Overall	51,419			72

*This interaction assay false-positive rate is taken from D'haeseleer and Church [32] or derived using the method therein. Multiple values derive from choosing either the GRID [2] or MIPS [33] reference sets. ^‡This interaction assay false-positive rate is calculated with the EPR server of Deane et al. [42]. ^†The mean of four values estimated from Table S3 of Lee et al. [24] by fitting the interaction set as a linear combination of true-positive (small scale interactions) and false-positive (random pairs) interactions.

ISSN: 1474-760X