Abstract
We introduce a systematic approach to constructing entire pathway model of whole cell metabolism: (1) Top down modeling from genomic information, (2) Bottom up modeling from metabolome analysis, and (3) Closing gap by bioinformatics. We have developed a computer program named GEM system that automatically constructs whole-cell-level metabolic pathway model from genome sequence data of a given organism, based on general enzymatic information published in credible database systems, such as COG, SWISS, KEGG and EMBL. Models so constructed are incomplete due to incomplete knowledge in the published databases. To fill the gaps in the incomplete pathway models, we next identify all metabolites in the cell by CE/MS (for charged metabolites) and LC/MS (for neutral metabolites), and then bioinformatics algorithms try to connect them to complete the pathway model. The bioinformatics methods to predict unknown pathways are based on chemical structure comparisons and correlations of dynamic changes in quantity among metabolites.
