Author/Authors :
Ng, Siew Teng UniversitiTeknologi Malaysia - Faculty of Computing - Artificial Intelligence and Bioinformatics Research Group, Malaysia , Chong, Chuii Khim UniversitiTeknologi Malaysia - Faculty of Computing - Artificial Intelligence and Bioinformatics Research Group, Malaysia , Choon, Yee Wen UniversitiTeknologi Malaysia - Faculty of Computing - Artificial Intelligence and Bioinformatics Research Group, Malaysia , Chai, Lian En UniversitiTeknologi Malaysia - Faculty of Computing - Artificial Intelligence and Bioinformatics Research Group, Malaysia , Deris, Safaai Universiti Teknologi Malaysia - Faculty of Computer Science and Information Systems - Artificial Intelligence and Bioinformatics Laboratory, Malaysia , Illias, Rosli Md UniversitiTeknologi Malaysia - Faculty of Chemical Engineering - Department of Bioprocess Engineering, Malaysia , Shamsir, Mohd Shahir Universiti Teknologi Malaysia - Faculty of Biosciences and Medical Engineering - Department of Biosciences and Health Sciences, Malaysia , Mohamad, Mohd Saberi Universiti Teknologi Malaysia - Faculty of Computing - Artificial Intelligence and Bioinformatics Research Group, Malaysia
Abstract :
Parameter estimation is one of nine phases in modelling, which is the most challenging task that is used to estimate the parameter values for biological system that is non-linear. There is no general solution for determining the nonlinearity of the dynamic model. Experimental measurement is expensive, hard and time consuming. Hence, the aim for this research is to implement Particle Swarm Optimization (PSO) intoSBToolbox to solve the mentioned problems. As a result, the optimum kinetic parameters for simulating essential amino acid metabolism in plant model Arabidopsis Thaliana are obtained. There are four performance measurements used, namely computational time, average of error rate, standard deviation and production of graph. As a finding of this research, PSO has the smallest standard deviation and average of error rate. The computational time in parameter estimation is smaller in comparison with others, indicating that PSO is a consistent method to estimate parameter values compared to the performance of Simulated Annealing (SA) and downhill simplex method after the implementation into SBToolbox.
Keywords :
Parameter estimation , PSO , SBToolbox , Arabidopsis Thaliana
