Estimating the size of the human interactome

Abstract

After the completion of the human and other genome projects it emerged that the number of genes in organisms as diverse as fruit flies, nematodes, and humans does not reflect our perception of their relative complexity. Here, we provide reliable evidence that the size of protein interaction networks in different organisms appears to correlate much better with their apparent biological complexity. We develop a stable and powerful, yet simple, statistical procedure to estimate the size of the whole network from subnet data. This approach is then applied to a range of eukaryotic organisms for which extensive protein interaction data have been collected and we estimate the number of interactions in humans to be approximately 650,000. We find that the human interaction network is one order of magnitude bigger than the Drosophila melanogaster interactome and approximately 3 times bigger than in Caenorhabditis elegans.

Fig. 1.

Performance of the estimator, Eq. 8, for the yeast network. Here, the DIP dataset was taken as a gold-standard “true” interaction network. (A) True network size (red bars) and histograms of predicted sizes for subnets that were created by sampling 20%, 40%, 60%, and 80% of nodes with equal probability. (B) Fraction of estimates obtained from 1,000 independent subnets (covering 20%, 40%. 60%, and 80% of the nodes in the true network) where the empirical 95% bootstrap confidence interval (based on 1,000 replicates) contains the true value (green).

Fig. 2.

Estimated interactome sizes for humans and three other eukaryotic species for which high-throughput interaction data are available. The letters denote the approximate position of the point estimate, M̂_𝒩, and the horizontal bars indicate the range of the approximate 95% CIs (obtained from 10,000 bootstrap replicates; see SI Text for details). (The yeast and human datasets are largely independent but there is large overlap between the datasets for D. melanogaster and especially C. elegans.)