Centrality Analysis in PPI Networks

The need to investigate the atomic constituents of biological network has led to increasing popularity of topological studies. A biological network is a graphical representation of the interactome with nodes and edges. Therefore, graph theoretic measures can be applied to such networks. We have performed an empirical study comparing a number of centrality measures, viz., betweenness centrality, eigenvector centrality, pagerank centrality, closeness centrality and radiality associated with a graph in the context of PPI networks. This empirical study shows the superiority of pagerank and radiality measure. We believe that their superiority can be leveraged in the analysis of other networks such as gene networks or metabolic networks.