# Protein interaction network

The definition and generation of good null models to assess the statistical relevance of observed features in protein interaction networks (PIN) is a well known problem in systems biology, where the aim is to understand how the structure of PINs relates to their biological functionality. By using statistical mechanics techniques, Annibale and Coolen recently derived explicit formulae for complex network ensembles with built-in topologies beyond the degrees, as proxies for real networks; provided exact algorithms for generating random graphs from such ensembles; and derived information-theoretic tools to quantify distances between networks. The application of these tools has enabled them and their coworkers to quantify detection biases in PIN data sets and species consistency.