Threonine Biosynthesis Pathway Simulation Using IBMDE with Parameter Estimation

摘要

When analysing a metabolic pathway through mathematical model, it is important that the significant parameters are being correctly estimated. However, this process often comes across problems such aseasily being trapped in local minima, repetitive exposure to worse results during the search process, and occurrence of noisy data. Thus, an improved Bee Memory Differential Evolution algorithm (IBMDE), which is a hybrid of the Differential Evolution algorithm (DE), the Kalman Filter (KF), Artificial Bee Colony algorithm (ABC), and a memory feature is presented this paper. IBMDE is an improved estimation algorithm as previous work only utilised DE. The threonine biosynthesis pathway is the metabolic pathways used in this paper. For metabolite O-Phosphohomoserine production simulation, the IBMDE able to produce the estimated optimal kinetic parameter values with significantly reduced error rate (63.67%) and shows a faster convergence time (71.46%) compared to the Nelder Mead (NM), the Simulated Annealing (SA), the Genetic Algorithm (GA), and DE respectively. In addition, IBMDE demostrates to be a reliable estimation algorithm.