S. Kimura, M. Hatakeyama, T. Kawasaki, T. Naka, and A. Konagaya (Japan)
Biochemical Pathway, Parameter Estimation, Optimiza tion, Computer Simulation, Bioinformatics
The computational simulation of biochemical pathways re quires a technique for estimating unmeasurable parameters from observable data. In many cases this can be prob lematic, as parameter estimation problems for biochemi cal pathways tend to have difficult properties. This paper proposes two techniques that can be used to overcome the difficulties of such problems. The first, a problem decom position, is used to divide the parameter estimation prob lem into several subproblems. The second, a logarithmic transformation, is used to decrease the flat area of the ob jective function of this problem. We apply these two pro posed techniques to estimate the parameters of a model of HRG-stimulated ErbB4 receptor signaling.
Important Links:
Go Back
