Sensitivity analysis of a mechanistic growth model of Escherichia coli

A detailed mechanistic model, with a large number of metabolites and parameters, describing growth of a single cell of the bacterium Escherichia coli is simulated and the resulting predictions are compared with available reported experimental observations. The effects of small time-invariant perturbations in each of the parameters on final model predictions of metabolite concentrations and doubling times are systematically investigated using the theory of sensitivity analysis. This is quantified by defining (linear) sensitivity coefficients $\underline{\Lambda},$ and are subsequently used to identify sensitive parameters in the model description. Growth rate effects due to parametric perturbations are studied by comparing the total phase shift obtained with that for a nominal system description. The prevailing behavior of kinetic and polymerization rate constants compared to that of saturation constants is established. Results obtained are successfully compared with available experimental data. Implications to Metabolic Control Analysis are also discussed