FUSE: a profit maximization approach for functional summarization of biological networks

<p>Abstract</p> <p>Background</p> <p>The availability of large-scale curated protein interaction datasets has given rise to the opportunity to investigate higher level organization and modularity within the protein interaction network (<smcaps>PPI</smcaps>) using graph theoretic analysis. Despite the recent progress, systems level analysis of <smcaps>PPIS</smcaps> remains a daunting task as it is challenging to make sense out of the deluge of high-dimensional interaction data. Specifically, techniques that automatically abstract and <it>summarize </it><smcaps>PPIS</smcaps> at multiple resolutions to provide high level views of its functional landscape are still lacking. We present a novel data-driven and generic algorithm called <smcaps>FUSE</smcaps> (<b>Fu</b>nctional <b>S</b>ummary G<b>e</b>nerator) that generates <it>functional maps </it>of a <smcaps>PPI</smcaps> at different levels of organization, from broad process-process level interactions to in-depth complex-complex level interactions, through a pro t maximization approach that exploits <it>Minimum Description Length </it>(<smcaps>MDL</smcaps>) principle to <it>maximize information gain </it>of the summary graph while satisfying the <it>level of detail constraint</it>.</p> <p>Results</p> <p>We evaluate the performance of <smcaps>FUSE</smcaps> on several real-world <smcaps>PPIS</smcaps>. We also compare <smcaps>FUSE</smcaps> to state-of-the-art graph clustering methods with <smcaps>GO</smcaps> term enrichment by constructing the biological process landscape of the <smcaps>PPIS</smcaps>. Using <smcaps>AD</smcaps> network as our case study, we further demonstrate the ability of <smcaps>FUSE</smcaps> to quickly summarize the network and identify many different processes and complexes that regulate it. Finally, we study the higher-order connectivity of the human <smcaps>PPI</smcaps>.</p> <p>Conclusion</p> <p>By simultaneously evaluating interaction and annotation data, <smcaps>FUSE</smcaps> abstracts higher-order interaction maps by reducing the details of the underlying <smcaps>PPI</smcaps> to form a <it>functional summary graph </it>of interconnected <it>functional clusters</it>. Our results demonstrate its effectiveness and superiority over state-of-the-art graph clustering methods with <smcaps>GO</smcaps> term enrichment.</p>

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.