Event driven modeling for the accurate identification of metabolic switches in fed-batch culture of S. cerevisiae

Mechanistic model-based methods are indispensable tools for characterization, monitoring and control in biopharmaceutical industry. However, the complexity of mechanistic models is restricted by the availability of process analytics. As a result, biological reactions are often lumped and only central metabolic pathways and extracellular analytics are considered. Moreover, due to process dynamics during typical batch and fed-batch cultivations, intracellular phenomena can often not be neglected. Typical examples are the Pasteur effect, Crabtree effect, and diauxic growth. A solution to this is to formulate discontinuous (piecewise) growth models and to incorporate metabolic switches expressed as logical operations. This contribution discusses the application of a piecewise kinetic growth model in the context of an industrial relevant case study. Targeted Saccharomyces cerevisiae lab scale experiments were conducted with different glucose and ethanol fluxes to trigger switches between metabolic pathways. We propose to use an event driven method to accurately identify the location and sequence of these switches, and the duration of active metabolic pathways during the time course of an experiment. It turns out that, compared with a standard implementation without active event location, the proposed approach leads to more accurate identification of switches and model parameters and thus, to more accurate model predictions.