Hybrid approach to modeling metabolic systems using genetic algorithm and simplex method

Genetic algorithms (GA) have been demonstrated to be a promising search and optimization technique that is more likely to converge to a global optimum than most alternative techniques. In an attempt to apply GA to estimate parameters of a metabolic model, however, we found that the slow convergence rate of GA becomes a major problem for its applications to model identification of dynamic systems due to the high computational costs associated with the evaluation of models. To alleviate this difficulty, we developed a hybrid approach that combines Nelder and Mead's downhill simplex method with the genetic algorithm. We evaluated the hybrid approach by extensively comparing its performance with pure GA and pure simplex approaches for the metabolic modeling problem and a function optimization problem. As expected, the hybrid approach not only speeds up GA's rate of convergence, but also improves the quality of the solution found by pure GA.