The Research Of Constructing Gene Regulatory Network Based On Model Of Dynamic Bayesian

After the completion of gene sequencing, the emergence of gene chip technology of large-scale determination of gene expression level and the use of high-performance computer make it possible to large-scale study on the regulation of gene expression using simulation methods, and reconstruction of gene regulatory network became One of the most challenging issues in functional genome. Using Bayesian Networks to construct gene regulatory networks is currently the bioinformatics research focus. DBNs can incorporate the timing characteristics of data into models, to overcome the static Bayesian Network's acyclic graph shortcomings and to express the feedback of genes ,so DBNs is more suitable for handling the complex biological phenomena -gene regulatory networks.The main contents of this dissertation are as follows:(1) A survey about the research on Bayesian networks was made, including the background, the current research state and development trend of Bayesian networks, the basic principle of Bayesian networks,the Principles of Dynamic Bayesian Network and the advantages of Construction gene regulatory networks.(2) Dynamic bayesian network (DBN) is apowerful modeling tool for gene rugulation network.Missing data in building gene regulation network is better dealt with SEM(Bayesian structre expectation maximization) algorithm,however,the result of learning by SEM algorithm has strong depengence on the iniaial parameters.This dissertation proposed an improver SEM algorithm,which randomly generated a number of candidate initial parameters and selectde the best parameters and selectde the best parameter as whole model's initialparameter to erecute basic SEM algorithm after a iterative process.Comparing gene regulation network constructed with yeast cycle gene expression data by improved SEM algorithm with existing literature improve the accuracy of constructing regulati on net work.(3) Most research work in learning gene networks assumes that either there is no time delay in gene expression or there is a constant time delay. However, the biological literature shows that different gene pairs have different time delays for gene regulation. This dissertation proposed an effective method to model mutiple time unit delayed gene regulatory network using DBNs. First of all, estimate transcription time delay of each regulate-target pair gene based on gene expression data and determine the set of potential regulatory genes,and then re-organization the gene expression data matrix on the basis of the estimated time delay, Finally the re-organized gene expression data is applied to the traditional dynamic Bayesian model to construct gene regulatory networks. This method is applied to the Saccharomyces cerevisiae cell cycle expression data, and to model mutiple time unit delayed gene regulatory network,and construct the delay of the gene regulatory networks, showing the effectiveness of the algorithm.