A computational method for reconstructing gapless metabolic networks

Abstract

We propose a computational method for reconstructing metabolic networks. The method utilizes optimization techniques and graph traversal algorithms to discover a set of biochemical reactions that is most likely catalyzed by the enzymatic genes of the target organism. Unlike most existing computational methods for metabolic reconstruction, our method generates networks that are structurally consistent, or in other terms, gapless. As many analyses of metabolic networks, like flux balance analysis, require gapless networks as inputs, our network offers a more realistic basis for metabolic modelling than the existing automated reconstruction methods. It is easy to incorporate existing information, like knowledge about experimentally discovered metabolic reactions or metabolites into the process. Thus, our method can be used to assist in the manual curation of metabolic network models as it is able to suggest good candidate reactions for filling gaps in the existing network models. However, it is not necessary to assume any prior knowledge on composition of complete biochemical pathways in the network. We argue that this makes the method well-suited to analysis of organisms that might differ considerably from previously known organisms. We demonstrate the viability of our method by analysing the metabolic network of the well-known organism Escherichia coli. © Springer-Verlag Berlin Heidelberg 2008.