Research On The Association Relationship Of Transcription Factor With LncRNA And The Construction Of Regulatory Network

With the development of the next generation sequencing, we find that in the eukaryotic genome, not all of RNAs can code proteins. RNAs transcribed by most of genes can’t code proteins, named as non-coding RNAs(ncRNAs). One of the emerging non-coding RNA is long non-coding RNA(lncRNA) whose length is longer than 200 nt. LncRNAs are almost as many as protein-coding RNAs. Recent studies have shown that in the eukaryotic genome lncRNAs have various biological functions.They can play a part in the regulation of gene expression in various ways. LncRNAs also are generated by DNA sequences according to complementary base pairing rules,and the transcription also is regulated by transcription factors. The previous studies usually only researched the regulation of protein-coding genes by transcription factors,while few studies paid attention to the regulation of lncRNA genes by transcription factors.In this paper, we analyzed the difference between the expression of lnc RNA genes and the protein-coding genes in some different cell lines. Based on SVM model,we used six kinds of histone modifications(H3K4me1, H3K4me2, H3K4me3,H3K9 ac, H3K27 ac and H3K27me3) to analyze the promoters of lncRNA genes and protein-coding genes. Finally, based on mixed na?ve Bayesian model, using transcription factors PWM scores, DNase I footprint and six histone modification, we constructed a classification model to predict the regulation of lnc RNA genes by transcription factors.We found that the expression level and expression ratio of lnc RNA genes were general lower than those of protein-coding genes, which indicated that eukaryotes mainly expressed the protein-coding genes, and also indicated that lnc RNA genes had more tissue specificity and cell specificity. We also found that, lnc RNA genes had the similar promoter regions to protein-coding genes, therefore we could analyze the promoter regions of lnc RNA genes using the way of analyzing protein-coding gene promoters. In the end, We found that our Bayesian model could predict the regulation of lnc RNA genes by transcription factors accurately. Our research can help people to understand the regulatory mechanisms and biological functions of lnc RNA genes more thoroughly and construct more detailed gene regulatory network.