Ensemble modeling and related mathematical modeling of metabolic networks

Metabolic engineering involves the manipulation or modification of key genes in metabolic pathways to accomplish the goal of increasing the production of a substance of interest. While candidate genes for manipulation are commonly chosen through intuition, many effective targets are non-trivial and thus serve as the motivation for mathematical modeling. Effective mathematical modeling strategies allow for a greater insight into the behavior of the biological system, while helping to identify candidate genes for manipulation. In this article, we summarize the advantages and limitations of various modeling methodologies, including both stoichiometric and kinetic approaches, using a simple model system to illustrate their application. As mathematical modeling of metabolic systems becomes more widespread and easy to implement, the iteration between modeling and experiment has the opportunity to become increasingly symbiotic, allowing for experiment to be driven by computation, and computation to be updated and refined by the results of experiment. In this light, we focus our analysis on the newly introduced ensemble modeling methodology. This approach allows for a tighter relationship between computation and experiment, bringing this goal a step closer to common practice.