A Method For Determining The Elementary Flux Modes In Genome-based Large-scale Metabolic Network

Elementary flux mode (EFM) analysis is a useful tool for characterizing thestructure and function of metabolic network. However, a direct calculation of EFMsof genome-based large-scale metabolic network is presently prohibitive due to theproblem of combinatory explosion of computational time and memory capacityrequired. Overcoming this computational problem is desired for a more efficient andtimely exploitation of genomic information for a system level understanding ofcellular metabolism. In this work we present a method to calculate the EFMs of metabolic networkreconstructed from genome data. The method is composed of four major steps: (1)graphic representation and analysis of the metabolic network to reduce the networkscale and to define the external metabolites, (2) division of the reduced network intosub-networks of suitable sizes, (3) calculation of EFMs for the subnetworks and (4)assembling of EFMs for the subnetworks with consideration of EFMs spanning thesubnetworks. The method is demonstrated for the calculation of EFMs of metabolicnetwork of Streptococcus pneumoniae. The calculated EFMs are 111429 and 180 inthe right subnetwork and in the left subnetwork, respectively. After combining EFMsspanning across the two subnetworks, we obtained 268343 EFMs in the wholemetabolic network of S.pneumoniae. We have analyzed enzyme subsets, and thecycles of bioreactions, and pathway redundancy. We find there are big differencesbetween the structures of pathway producing different products, using differentsubstrates and using different metabolites, which are potential substrates or products. Metabolic fluxes of organisms realized under given growth conditions arecombinations of different EFMs. With the determination of the total number of EFMs,it is desirable to know how these EFMs contribute to a given steady state fluxdistribution and thus to know how fluxes are regulated at EFMs level. However, nowork has been done in literature in this regard. In this work, a method is presented todetermine the contribution of EFMs to a given steady state flux distribution in asimplified metabolic network of Escherichia coli.