Study On Identification Method Of Gene Regulation Network

This paper introduces the research status of gene regulatory networks at home and abroad, describes the significance of the research of gene regulatory network. Based on the existing gene regulatory network model and the principle of system identification to find out the method for estimating the parameter of linear regression model and the differential equation of the gene regulatory network, and carried out the simulation analysis. The main contents of this paper are as follows:(1) From the research background, research significance, research status and trend to describe the types and its construction method of gene regulatory networks from the existing models.(2) Introduces the identification process of system identification and describes two identification algorithms: least square algorithm and the recursive least squares identification algorithm.(3) Building two kinds of linear model of gene regulatory networks: the output error model and the coupling linear model, finding the identification method to identify the models of these gene regulatory networks through the least squares identification algorithm principle and verifying the validity of the algorithm through simulation example.(4) Establish a linear differential equations model of the gene regulatory network and using the Euler discrete principle change differential equations into difference equation then get the linear regression identification model. Finally using the least square method for parameter identification of the linear differential equations and modeling the gene regulatory network.(5) Establish a nonlinear differential equation of the gene regulatory network, using the Euler discrete principle change the continuous differential equations into a discrete difference equation, then estimate the parameters of the linear and nonlinear part of the model respectively and get the identification algorithm. Finally prove the feasibility of the algorithm though a example.