Optimal Network Alignment with Graphlet Degree Vectors

摘要: Important biological information is encoded in the topology of networks. Comparative analyses networks are proving to be valuable, as they can lead transfer knowledge between species and give deeper insights into function, disease, evolution. We introduce a new method that uses Hungarian algorithm produce optimal global alignment two using any cost function. design function based solely on network use it our alignment. Our applied networks, not just ones, since only topology. align protein-protein interaction eukaryotic demonstrate exposes large topologically complex regions similarity. At same time, biologically valid, many aligned protein pairs perform From alignment, we predict yet unannotated proteins, which validate literature. Also, apply find topological similarities metabolic different build phylogenetic trees score. The obtained this way bear striking resemblance ones by sequence alignments. detects similar statistically significant. It does independent or other external