A hybrid approach to modeling metabolic systems using genetic algorithm and simplex method

The genetic algorithm is applied to the parameter estimation problem to optimize a model of the glucose cycle of an E. Coli cell. Since the evaluation of the model is computationally expensive, a hybrid algorithm is proposed which grafts a proposed variant of J.A. Nelder and R. Mead's (1965) downhill simplex-called concurrent simplex-with the genetic algorithm by using the simplex as an additional operator. The addition of the operator speeds up the rate of convergence of the genetic algorithm in some cases. The advantages and disadvantages of the simplex hybrid are discussed and the hybrid is tested against several different problem sets to verify its improvement over the generic genetic algorithm.<
>