ESTIMATING PARAMETERS OF A MODEL OF THIN FILAMENT REGULATION IN SOLUTION USING GENETIC ALGORITHMS
B.Stojanovic, M.Svicevic, Dj.Nedic, M.Ivanovic, S.M.Mijailovich
The estimation of chemical kinetic rate constants for any non-trivial model is complex due to the nonlinear effects of second order chemical reactions. To accomplish this goal we have developed an algorithm based on genetic algorithms (GA) and then tested the effectiveness of this method on the McKillop–Geeves (MG) model of thin filament regulation. This method have shown better accuracy than deterministic methods, the Damped Least Squares (DLS), quasi-Newton (QN) and simulated annealing (SA). However, it requires large number of evaluations of candidate solutions which take longer CPU time. In this paper, a platform for distributing evaluation of different sets of parameters (i.e. individuals in genetic algorithm) over a number of threads on a single computer is presented. Performed tests have shown that parallelized evaluation provides significant speedup with better accuracy and estimation times comparable to deterministic methods.