Construction of Kinetic Model Library of Metabolic Networks

Kinetic model identification of metabolic networks is important, but still represents a significant challenge. The difficulties faced vary from the vast kinetic parameter search space to the lack of complete parameter identifiability. To meet these challenges, an incremental modeling approach is proposed here, including two key components—dynamic flux calculation and flux-based kinetic parameter estimation. In essence, the identification method relies on time-course concentration data to generate the family of consistent metabolic flux values and by doing so, the parameter estimation step can be done one flux at a time. The key contribution of the method is an efficient generation of a library of kinetic models with similar goodness of fit to the provided data. The performance of this identification method is demonstrated using a generic branched metabolic pathway model and the glycolytic pathway model of Lactococcus lactis (L. lactis).