Control of the production of Saccharomyces cerevisiae on the basis of a reduced metabolic model*

Saccharomyces cerevisiae is a species of yeast with a long tradition in human history and a growing demand in industry and research. The yeast cells are produced in a series of fed batch reactors which are fed with oxygen and glucose as the main carbon source. One problem during the production process is that the cell culture can switch to the undesired production of ethanol leading to a lost batch. For improving the production process a suitable modeling and control strategy is needed that should cover the switch to ethanol production and should be able to describe the growth of the cell culture so that the operating policies can be optimized. This work presents a novel method that uses dynamic flux balance analysis to derive a reduced metabolic model from a full biochemical stoichiometric network which is then used within a model predictive control. The reduced metabolic model covers the gene regulation by using the redox metabolites as key regulators. It is shown that this modeling approach is very flexible and can be used to control and to monitor the process.