Abstract: In recent years constraint based methods such as Flux Balance Analysis (FBA) has widely applied for computation of flux distributions in the metabolic networks. In this study, the effects of changes in the intercellular concentration on the Gibbs free energy of the system and subsequently on the model’s fluxes have studied. This method which makes a correlation between the flux directions and metabolite concentrations, has applied to large scale metabolic network, Escherichia coli iAF1260. The biomass and succinate fluxes have selected as objective functions and the multi objective genetic algorithm has used to optimization of the E. coli iAF1260 network. The obtained results revealed that a living system such as E. coli and its mutants are not stable from thermodynamic point of view before reaching to the cellular threshold concentrations. Also the behavior of the mutants of microorganism involving their return to the wild type phenotypes could be justified.