Table 5.
Dataset (ref.) | N | m | PFP | E [Y1] | E [Y2] | U obs | R obs |
ItoFull [1] | 720 | 258,840 | 0.0008 | 414 | 0.17 | 435 | 68 |
ItoCore [1] | 128 | 8,128 | 0.0025 | 41 | 0.05 | 43 | 36 |
Uetz et al. [6] | 108 | 5,778 | 0.003 | 35 | 0.05 | 36 | 10 |
Gavin et al. [10] | 852 | 362,526 | 0.0017 | 1230 | 1.10 | 1201 | 743 |
Krogan et al. [11] | 1,458 | 1,062,153 | 0.0019 | 4,029 | 3.80 | 3945 | 538 |
Shown are the expected number of false positive (FP) errors on the filtered datasets for [1,6,10,11]. N is the number of proteins within each filtered dataset. The values for PFP and m are estimated upper bounds obtained by setting PFN = 0 and using the solution curves of Figure 6c,d. Denote Y1 as the random variable for the number of unreciprocated FP observations, and Y2 for the number of reciprocated FP observations. The variables Uobs and Robs show the observed number of unreciprocated and reciprocated interactions from the data, respectively. The table implies that even in the worst case scenario for maximal PFP, reciprocated edges mostly report true interactions.