Quantifying Systemic Evolutionary Changes by Color Coding Confidence-Scored PPI Networks

Abstract

A current major challenge in systems biology is to compute statistics on biomolecular network motifs, since this can reveal significant systemic differences between organisms. We extend the "color coding" technique to weighted edge networks and apply it to PPI networks where edges are weighted by probabilistic confidence scores, as provided by the STRING database. This is a substantial improvement over the previously available studies on, still heavily noisy, binary-edge-weight data. Following up on such a study, we compute the expected number of occurrences of non-induced subtrees with k ≤ 9 vertices. Beyond the previously reported differences between unicellular and multicellular organisms, we reveal major differences between prokaryotes and unicellular eukaryotes. This establishes, for the first time on a statistically sound data basis, that evolutionary distance can be monitored in terms of elevated systemic arrangements. © 2009 Springer Berlin Heidelberg.