Semantic visualization of multifunctional PPI networks by edge coloring

Abstract

Protein-protein interaction (PPI) network data contains interaction among proteins in organisms. PPI networks are used to many cell behavior studies, and visualization of the network is essential for clear and efficient analysis. Gene Ontology (GO) data contains known features of gene products, such as proteins those are represented as nodes in PPI networks. So integrating PPI network data and Gene Ontology information extends the amount of information of nodes while makes more challenging to visualize. In this research, we suggest new method to visualize multifunctional PPI networks built by integrating PPI and GO data, especially using biological process terms. We annotate GO terms to edges by comparing GO annotation terms of two nodes connected by the edges. We visualize functional information from GO data using edge color, differently from previous methods, which visualize the data by modifying visual variables of nodes. In addition, we visualize edge betweenness centrality of edges respectively for functional networks. And we assign similar colors to similar GO term considering semantic similarity among GO terms. This method can visualize multifunctional PPI networks and represents some information of the networks, which other previous methods cannot represent.