Two stage optimization of biochemical pathways using parallel runs of global stochastic optimization methods

A disadvantage of global stochastic optimization methods is the stochastic convergence of the best value of the objective function to the global optimum. In spite those methods are widely used optimizing kinetic models of biochemical pathways. Moreover, the growing number of adjustable parameters in the model leads to combinatorial explosion of combinations of adjustable parameters if the best combination per number of adjustable parameters in the combination has to be found. Two stage optimization using automatic optimization termination criteria of parallel optimization runs in case of consensus is proposed in this paper to reduce the duration of scanning. The application of this method is demonstrated solving biotechnologically relevant task - maximization of bioethanol production rate in the yeast glycolysis model. The experimental results underline the efficiency of two stage approach using parallel runs and consensus criteria.