Research On Gene Network Reconstruction Based On Prior Biological Knowledge

In recent years,with the completion of the human genome project,progress in molecular biology techniques,especially in next-generation sequencing(NGS)refor-m principles of biological research,innovations in these methods allow scientists to spend a lot of time and costs,however,the analysis of such data becomes a new challenge in the field of Bioinformatics[2].All data mining task,gene regulatory net-work(GRN)inference and prediction has become one of the most exciting topics[3].After the sorting and recognition of biological organisms,researchers interested in genes how to adjust to each other.ENCODE recently completed task reveals most of non-coding DNA involved in the regulation of gene expression,further empha-sized the importance of gene regulatory networks.In short,study on inference of gene regulatory networks in the next ten years will still be very important.Al-though there are many studies on inference of gene regulatory networks,but only a few people focused on the emerging data time series gene expression data.With the help of next-generation sequencing technology,researchers can easily get from a biological process such data,for example,the cell cycle.Almost each individual's biological phenomena is a dynamic process involving time,biologists are increasingly interested in the collection and analysis of time series gene expression data.In fact,known as gene expression database has been collected in recent years an increasing number of time series data.Therefore,the development of reliable gene regulatory network inference and time series algorithms for time series gene expression data has become an urgent task.Between the study of gene regulatory networks from gene interactions reveal the complexity of the phenomenon of life,is an important content of functional genomics research,is also the current frontiers of bioinformatics.Ap-plication of gene chip technology in bioinformatics to gene network reconstruction provides a number of available research and analysis of the underlying data.This paper according to DBN network modeling method and the lack of mutual information of network reconfiguration method for unbalanced data sets is proposed(n t)approach.Specific do has following several items work;first introduced two Yuan causal relationship model and the multiple causal relationship model;second,through joined prior knowledge using Ling return found gene Zhijian causal rela-tionship;last,calculation each article gene side of significantly sex level,through significantly sex level and Pearson related coefficient filter gene side,on gene func-tion for comments,on significantly exists of gene relationship through Cytoscape software for show.