The Connectivity of Large Genetic Networks: Design, History, or Mere Chemistry?

I review evolutionary explanations of broad-tailed connectivity or degree distributions observed in metabolic networks and protein interaction networks. Self-assembled chemical reaction networks show degree distributions similar to those observed for metabolic networks, which argues against the postulated role of natural selection in maintaining this degree distribution. In addition, metabolic networks contain traces of their ancient history in the form of highly connected metabolites. Similarly to the degree distribution of metabolic networks, that of protein interaction networks can be explained without resorting to natural selection on the network level. I present data suggesting that highly connected proteins are not distinguishably older than other proteins, and explain this finding with a simple model of how a protein’s degree changes in evolutionary time.