Michihiro Tanaka (email@example.com)
Takuji Yamada (firstname.lastname@example.org)
Masumi Itoh (email@example.com)
Shujiro Okuda (firstname.lastname@example.org)
Susumu Goto (email@example.com)
Minoru Kanehisa (firstname.lastname@example.org)
Bioinformatics center, Institute for Chemical Research, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan
Recent evidence points to the existence of scale-free properties in many biological networks. By topological analysis, several models including preferential attachment and hierarchical modules have been proposed to explain how these networks are organized. On the other hand, analyses using dynamics have suggested that gene expression and metabolic networks have been organized with the scale-free property by the other models such as “rich-travel-more” and “log-normal dynamics.” Because most of these approaches are based on comparative genomics of extant species, and did not consider evolutionary events such as horizontal gene transfer, gene loss and gene gain, we have analyzed transition of metabolic networks from the vertical point of view of evolution. First, to identify metabolic networks of common ancestors, we applied a parsimony algorithm for the enzymatic reaction set. Then by comparing the estimated metabolic networks among common ancestors, we investigated the transition of metabolic networks along the evolutionary process. As a result, we estimated enzymatic reaction contents of 227 common ancestors from 228 extant species, and found that links of several specific metabolites have frequently changed during the course of evolution.