Evidence of probabilistic behaviour in protein interaction networks

BMC Systems Biology

Table 1 Properties of PPI networks

Network	Number of proteins	Number of interactions	r*	θ ^†	γ^† (× 10^-5)	γ(cal)^‡ (× 10^-5)
H. sapiens	9263	34564	0.99	1.00 ± 0.02	1.49	1.45
D. melanogaster	6736	20308	0.99	1.01 ± 0.02	2.25	2.46
Yeast
- DIP	4617	16311	0.99	1.02 ± 0.02	2.98	3.07
- CORE	2449	5579	0.99	1.03 ± 0.06	8.41	8.97
- Y2H	3277	4393	0.99	1.15 ± 0.08	7.22	11.4
E. coli	1473	5709	0.97	1.06 ± 0.04	6.09	8.79
Worm
- Y2H	2624	3967	0.98	1.08 ± 0.07	9.59	12.6
- CORE	727	814	0.99	1.03 ± 0.09	53.5	61.7
P. falciparum	1304	2745	0.99	0.99 ± 0.05	18.0	18.3

*Pearson correlation coefficient for test of association between P(k₁, k₂) and k₁k₂, where P(k₁, k₂) is the probability of interaction between two proteins of degrees k₁ and k₂.
^† Fitted values (from the Log-Log plots in Figure 1) occurring in the expression P(k₁, k₂) = γ(k₁k₂)^θ. θ is given with 99% confidence intervals.
^‡ Prediction of γ via γ(cal) = E/Σ_i<j(k_ik_j), where E is the number of interactions in the network and the summation is over all pairs of proteins.

ISSN: 1752-0509