Abstract: In this work, we present a time-scale analysis based model reduction and parameter identifiability analysis method for metabolic reaction networks. The method uses the information obtained from short term chemostat perturbation experiments. We approximate the time constant of each metabolite pool by their turn-over time and classify the pools accordingly into two groups: fast and slow pools. We performed a priori model reduction, neglecting the dynamic term of the fast pools. By making use of the linlog approximative kinetics, we obtained a general explicit solution for the fast pools in terms of the slow pools by elaborating the degenerate algebraic system resulting from model reduction. The obtained relations yielded also analytical relations between a subset of kinetic parameters. These relations also allow to realize an analytical model reduction using lumped reaction kinetics. After solving these theoretical identifiability problems and performing model reduction, we carried out a Monte Carlo approach to study the practical identifiability problems. We illustrated the methodology on model reduction and theoretical/practical identifiability analysis on an example system representing the glycolysis in Saccharomyces cerevisiae cells. (C) 2008 Elsevier Inc. All rights reserved.
Abstract: In this work, the biomass growth and the TaqI endonuclease production by recombinant Esherichia coli were studied using artificial neural networks. The effects of the medium components on biomass growth and enzyme yield were modeled by various networks. After the most successful networks were statistically determined, they were used to extract additional knowledge such as the possible correlations between the biomass growth and the enzyme yield, and the relative significance of the medium components. It was found that the change of the biomass growth and the enzyme yield with the change of KH2PO4 concentration was strongly correlated with an R-value of -0.954. Some mild correlations were also observed for the other components. It was also found that the relative significances of the medium components were in the same order for both outputs: (NH4)(2)HPO4 Concentration was determined as the most important parameter followed by the glucose, KH2PO4 and MgSO4 concentrations. (C) 2008 Elsevier B.V. All rights reserved.
Abstract: The effect of bioprocess conditions (pH and temperature) on the growth and alkaline protease production of halotolerant Bacillus licheniformis BA17 bioreactor cultures have been systematically analyzed using response surface methodology in order to assess the importance of these generally disregarded parameters. Two models were proposed differing by the choice of response variable. Under optimized bioprocess conditions, whole alkaline protease activity was about 3 fold higher than the activities obtained in the preliminary studies. Results of this study not only highlight the importance of pH and temperature for further engineering purposes but also serve as basis for understanding the true mechanism lying under the relation between these process parameters and growth and whole alkaline protease production.
Abstract: The effect of bioprocess conditions (pH and temperature) on the growth and alkaline protease production of halotolerant Bacillus licheniformis BA17 bioreactor cultures have been systematically analyzed using response surface methodology in order to assess the importance of these generally disregarded parameters. Two models were proposed differing by the choice of response variable. Under optimized bioprocess conditions, whole alkaline protease activity was about 3 fold higher than the activities obtained in the preliminary studies. Results of this study not only highlight the importance of pH and temperature for further engineering purposes but also serve as basis for understanding the true mechanism lying under the relation between these process parameters and growth and whole alkaline protease production.
Abstract: Background: Dynamic modeling of metabolic reaction networks under in vivo conditions is a crucial step in order to obtain a better understanding of the (dis) functioning of living cells. So far dynamic metabolic models generally have been based on mechanistic rate equations which often contain so many parameters that their identifiability from experimental data forms a serious problem. Recently, approximative rate equations, based on the linear logarithmic (linlog) format have been proposed as a suitable alternative with fewer parameters. Results: In this paper we present a method for estimation of the kinetic model parameters, which are equal to the elasticities defined in Metabolic Control Analysis, from metabolite data obtained from dynamic as well as steady state perturbations, using the linlog kinetic format. Additionally, we address the question of parameter identifiability from dynamic perturbation data in the presence of noise. The method is illustrated using metabolite data generated with a dynamic model of the glycolytic pathway of Saccharomyces cerevisiae based on mechanistic rate equations. Elasticities are estimated from the generated data, which define the complete linlog kinetic model of the glycolysis. The effect of data noise on the accuracy of the estimated elasticities is presented. Finally, identifiable subset of parameters is determined using information on the standard deviations of the estimated elasticities through Monte Carlo (MC) simulations. Conclusion: The parameter estimation within the linlog kinetic framework as presented here allows the determination of the elasticities directly from experimental data from typical dynamic and/or steady state experiments. These elasticities allow the reconstruction of the full kinetic model of Saccharomyces cerevisiae, and the determination of the control coefficients. MC simulations revealed that certain elasticities are potentially unidentifiable from dynamic data only. Addition of steady state perturbation of enzyme activities solved this problem.
Abstract: Factorial designs and second order response surface methodology (RSM) for medium optimization were employed for the growth of recombinant Escherichia coli cells carrying a plasmid encoding TaqI endonuclease as a part of the fermentation strategy for general recombinant protein production. The method used was effective in screening for nutritional requirements using limited number of experiments. The concentrations of carbon source (glucose), inorganic nitrogen ((NH4)(2)HPO4), potassium (KH2PO4) and magnesium (MgSO4 center dot 7H(2)O) sources in medium were changed according to the central composite rotatable design consisting of 29 experiments, and the biomass yield was calculated. The optimum medium composition was found to be 15gL(-1) glucose, 6.6gL(-1) (NH4)(2)HPO4, 20.1 gL(-1) KH2PO4 and 1.7 gL(-1) MgSO4 center dot 7H(2)O. The model prediction of 2.72 gDCW L-1 biomass at optimum conditions was verified experimentally as 2.68 gDCW L-1 which is much higher than any value obtained in initial experiments as well as in studies carried out previously. The correlation between biomass growth and TaqI endonuclease enzyme yield obtained under the same medium compositions was also analyzed. (c) 2006 Elsevier B.V. All rights reserved.
Abstract: The effect of medium composition on the TaqI endonuclease production, by recombinant Escherichia coli cells carrying a plasmid encoding TaqI endonuclease, was investigated using response surface methodology. The concentration of glucose, di-ammonium hydrogen phosphate, potassium di-hydrogen and magnesium sulphate in media were changed according to a central composite rotatable design consisting of 29 experiments and enzyme yields were determined. The results were fitted to a second order polynomial with an R-2 of 0.828. The model equation was then optimized using the Nelder-Mead simplex method to maximize enzyme yield within the experimental range studied. The optimum medium composition was found to be 6 g L-1 glucose, 1.5 g L-1 (NH4)(2)HPO4, 8 g L-1 KHPO4, and 0.8 g L-1 MgSO(4)center dot 7H(2)O. The model prediction of 179 x 10(6) U g DCW-1 enzyme yield at optimum conditions was experimentally verified. This value is higher than any value obtained in the initial experiments as well as in the previously reported studies. The response surface methodology was found to be useful in improving the production of recombinant TaqI endonuclease in E. coli. (c) 2004 Elsevier Ltd. All rights reserved.
