Comparing metabolic networks at a global level

The metabolic networks analysis is an important process to discover similarities and differences between metabolisms of different organisms. When comparing metabolic networks we need to take into account the computational complexity due to the large size of such nets. In this thesis we propose a metabolic networks comparison method based on graphs representing an abstraction of the two metabolisms. This allows us to reduce the size of the data structures and makes feasible the computational problem. We build the two networks using the information in the KEGG data base and we compare them considering both the topology of the net and the information in each pathway. At the structural level the similarity is computed considering both a directed and an undirected graph representation. The proposed methods have been implemented in Java and integrated in a tool for metabolic networks comparison.