Research On Identification And Application Of Multiplex Boolean Networks

As a discrete model of time and state,Boolean network is one of the important models to study gene regulatory networks in systems biology.It is a relatively simple logical dynamic system based on directed graph.It can simulate some complex networks of biological systems.At the same time,with the development of systems biology,the research and development of genetic network model and network reconstruction have made remarkable progress,and the identification of multi-layer biological system model has become a hotspot in the field of research.As the complex regulatory network is tested in layers,the observation data shows that the network with more complex network structure is needed for identification.Therefore,it is of great significance to study the identification of Boolean multiplex networks based on half tensor product in this paper.This thesis focuses on Boolean multiplex network identification.First of all,based on the observed data,and the semi-tensor product of matrices,the Boolean network is expanded to the Boolean multiplex network identification.In view of the multilayer complex network,there is a global state layer,the network dynamic model is established based on the observation data,the multiplex Boolean network is identified to obtain the relationship between nodes in each layer and the relationship with the global state layer,construct its algebraic form,and then convert it back to the logical form.Furthermore,the necessary and sufficient conditions for Boolean multiplex network identification are given.Secondly,considering the randomness of real biological systems,the multiplex probabilistic Boolean network identification is proposed based on the Boolean multiplex network identification.Based on the observed data and using the method of matrix semi-tensor product,the given data are processed and written into the form of vector.According to the minimum degree modeling algorithm,the structural matrix is identified and the logical expression is obtained.Then,the maximum likelihood estimation method is applied.Find the probability that the data in the conflicting column is selected.Finally,it was applied to models of gene expression in the tissues of breast cancer patients and whether there was a Shared pathogenesis between the pathogenic micro RNA and different cancers.The validity of the result is further showed by these two examples.