Construction Of A Gene Regulatory Network For Arabidopsis Anther

Bioinformatics is a new subject that is studied by scientists who master knowledge involving in biology, science of computer and mathematics. Based on computers and networks, people use the approach of mathematics and informatics science to research biologic molecules and the regulatory mechanism between molecules in bioinformatics. Gene regulatory network is an active area of research in the post-genome research. The expression of a gene may be regulated or influenced by other genes, this gene in turn has the potential to regulate of influent additional genes. By identifying and organizing these transcriptional relationships, a gene regulatory network can be constructed. Understanding mechanisms of gene expression would not only facilitate the acquaintance of the process of real regulatory, but would also provide valuable insight into the development of therapeutic drugs and the field of biomedicine in general. In addition, gene regulatory network also play an important role in revealing complicated phenomenon of lives because of the regulatory or influent relationships between genes in organism.In this paper, first, we describe the history and model of some represent gene regulatory network, such as static and dynamic Bayesian network models, Boolean network models, Differential equation models and Neural network models. For these models, we further study their disadvantages and give some papers that focuse on the design of effective methods.Arabidopsis thaliana, the model plant, is a good example of the challenges of network reconstruction. The growth of Arabidopsis anther is regulated by gene networks which are know rarely by researches. In the present paper, based on the Maximum-clique algorithm, we used a bioinformatics approach that integrates the analysis of gene expression data with the prediction of transcription factor binding sites in the promoter regions, to construct a gene regulatory network. Using bioinformatics, a total of 6836 TF-gene pairs were analyzed, 95 of which were characterized as highly confident, and 5 were confirmed by previously published experimental data. These results suggest that the predictions by this model are reliable has the potential to improve our understanding of the role of these processes in plant development.Using the bioinformatics method, a large-scale and more accurate gene regulatory network can be constructed. A significant advantage of this method is more efficient and has a higher throughput capacity. We hope that the gene regulatory network we built will provide academic guide that are used to predict the relationships between target genes and TFs for biologists, while the experimental results can guide the construction of gene network. So, the mutual effects may lead to rapid development of bioinformatics and biology.