Identify The Transcriptional Factor Binding Sites Base On A Graph Method

Since the last 90s century, the development of Human Genome Project indicated that biology has stepped into the age of genome, bioinformatics has been one of the important field of science. The post genome era starts with more and more model organisms sequenced. And the transcriptional regulation is an important research topic in the post-genomic era. A large number of studies have shown that gene expression is effected each other. Study of this complex relationship has become an important field. Transcription factor binding sites is a segment of DNA and located in the upper area of gene and control the transcription. Identifying Transcription factor binding sites have an irreplaceable role on the Construction of transcriptional regulatory networks. Predicting highly-quality transcription factor binding sites has become an unavoidable and urgently research topics.There are a large number of methods for identifying transcription factor binding sites, and it can be divided into two ways: conservative motif based and comparative genomics. The map method is more popular in recent years. In this article a graph-based approach is used to identify transcription factor binding sites. The scientists found that in general transcription factor binding sites have a highly conserved short sequence of 3-5bp as core. Based on this feature, firstly in the upstream region of a group of co-regulated gene highly conserved core is searched; then, according to the user's request, they are extended on both sides to be transcription factor binding sites. This article makes experiments with yeast genetic data. Experimental results show that the method can obtain more good results in the limited experimental conditions. And compare with three classic algorithms, there is a few better performances. The results show that the method is valid.