论文部分内容阅读
Background: Flux variability analysis is usually used to determine the minimum and maximum flux values that the reactions in the metabolic network can carry under various simulation conditions.While several approaches have been proposed for that purpose,they are computationally too inaccurate to qualitatively classify reactions in genome-scale metabolic network models(GSMs).Therefore,developing preciser methods to analysize robustness of metabolic network models is urgent.Methods: We develop a new tool for qualitative classification of reactions base on flux variability analysis called QCFVA,which overcomes the problem of calculation accuracy of the exiting methods.The implementation of QCFVA was mainly divided into three steps: Firstly,the target reaction,such as biomass growth,was set as the objective function and FBA was used to maximize the flux of the target reaction(mentioned as the first simulation); Secondly,the flux of target reaction was set to the maximum value calculated in the first simulation; Finally,all the n reactions were considered as objective function in turn and FBA was again used to calculate the maximum and minimum fluxes of each reaction in GSMs,respectively.Results: We show that the improved implementation of QCFVA makes classification of reactions,calculation of the bounds on the reaction fluxes and flux balance impact degree(FBID)of single reaction deletion more exactly,greatly expanding the application of flux variability analysis in systems biology from prediction of genetic manipulation targets in genome scale to integration of high-throughput biological data to construct multi-scale whole-cell network model.