Metabolic Reconstructions

A central goal of integrative systems biology is the accurate representation of molecular interaction networks. Ultimately, such networks can be used to underpin mathematical models that permit the simulation of biological behaviour. The first step in generating such models is constructing a network of biochemical reactions and interactions between molecular components of the system to form a qualitative model.

Our group has been involved in this research since 2007, when we organized a Jamboree for unification of the various yeast metabolic reconstructions. It was recognized then that the proliferation of different and, to some extent, incompatible reconstructions of S. cerevisiae metabolism was an obstacle. Since then we have been involved in updating the reconstructions of yeast and human metabolism, and standardized existing reconstructions of E. coli and CHO cells.

Human Metabolic Reconstruction (Recon2)

Recon 2 was constructed by combination of five different sources (Recon1, the Edinburgh Human Metabolic Network, HepatoNet1, a small intestine enterocyte reconstruction, and models of macrophages, hepatocytes and kidney cells). This effort took some five years and several "jamboree" meetings (the first organized by us in Manchester). The result — Recon2 — is also specialized to represent 65 different cell types. This resource is a great tool to map the human genome to physiology, allowing for prediction of the consequences of mutations (such as inborn errors of metabolism), as well as the action of drugs and other environmental challenges.

Recon2 is available as a public resource, and also in the BioModels database as MODEL1109130000. Recently, Kieran Smallbone released an updated version of this reconstruction, fully carbon balanced and also fully elemental balanced, both are available in SBML format through the BioModels database (MODEL1311110000 and MODEL1311110001).

Yeast Metabolic Reconstruction

We update the most comprehensive and documented reconstruction of yeast metabolism. This stemmed from a community-wide collaborative project from which a consensus model emerged. We have subsequently continued to update this reconstruction by filling in gaps and adding parts of metabolism that were previously not well covered by reconstructions. The most up to date reconstruction is available in SBML format from yeast.sf.net.

Other reconstructions

Kieran Smallbone has encoded in SBML with MIRIAM annotations the metabolic reconstructions of:

Software

Neil Swainston constructed the SuBliMinaL Toolbox which facilitates the reconstruction process by providing a number of independent modules to perform common tasks, such as generating draft reconstructions, determining metabolite protonation state, mass and charge balancing reactions, suggesting intracellular compartmentalisation, adding transport reactions and a biomass function, and formatting the reconstruction to be used in third-party analysis packages. The individual modules manipulate reconstructions encoded in Systems Biology Markup Language (SBML), and can be chained to generate a reconstruction pipeline, or used individually during a manual curation process.

Bibliography