Pattern mining across many massive biological networks

The rapid accumulation of biological network data is creating an urgent need for computational methods on integrative network analysis. Thus far, most such methods focused on the analysis of single biological networks. This chapter discusses a suite of methods we developed to mine patterns across many biological networks. Such patterns include frequent dense subgraphs, frequent dense vertex sets, generic frequent patterns, and differential subgraph patterns. Using the identified network patterns, we systematically perform gene functional annotation, regulatory network reconstruction, and genome to phenome mapping. Finally, tensor computation of multiple weighted biological networks, which filled a gap of integrative network biology, is discussed.