On the Genetic Regulation of Bayesian Networks
Yan Zhang · 2019
To study the network regulatory relationship of genes, the probability knowledge and graph theory is combined by the Bayesian network method. The Bayesian network model of genes is constructed effectively and the reasoning is carried out. For a group of leukemia data, firstly, the data preprocessing such as standardization and discretization is carried out. Secondly, the order of the nodes between genes is obtained by the decision tree ID3 algorithm. And the structure learning of the Bayesian network is studied by the K2 algorithm to find out the network topology of genes. Then the parameter learning is used by the maximum likelihood estimation to find out the probability dependence relationship between the parent nodes and the child nodes in the network. Finally, the validity of the Bayesian network model is verified. The analysis of the test data shows that the Bayesian network has high accuracy to predict and analyse the regulatory relationship of genes.