Bioinformatical Analysis Of Gene Regulatory Network Consisting Of Transcription Factor And MicroRNA

Gene regulation is a key factor in gaining a full understanding of molecular biol-ogy. By studying gene regulation, we reveal the mechanisms underlying gene expres-sion, and learn more about a variety of biological process as embryonic developmentand disease pathogenesis.There are two important class of regulators in eukaryote, as transcription factor(TF) and microRNA (miRNA). TF regulates the transcription of their target genes byspecifically binding to gene’s promoter region. TFs have widely co-operation in theirregulation by forming cis-regulatory modules (CRM), which consist of multiple TFbinding sites. miRNA is a novel class of gene regulator. miRNAs are22nt smallnon-coding RNAs. They bind to the3’-untranslated region of target mRNA and fa-cilitate mRNA’s degradation or inhibit translation to regulate gene expression at thepost-transcriptional level. It has been established that TF, CRM and miRNA have acrucial function in a wide range of biological process.The interaction and combinatorial regulation of TFs and miRNAs have beenwidely identified. They form a complex regulatory network. There is a network motiftermed as feed forward loop, which plays a crucial role in gene expression stabilization.Therefore, systematically investigating the gene regulatory network of TFs and miR-NAs and discovering their network motifs are essential to elucidate the gene regulatorymechanism.The regulatory network is typically constructed by computational approachesbased on sequence analysis. However, it has been recognized that these computationalapproaches for TF and miRNA target prediction have high false-positive rate. With thedevelopment of high throughput technology, more and more expression profiles havebeen available to study gene regulation. Here we propose a novel approach named as graphical adaptive LASSO (GALASSO). GALASSO incorporates adaptive LASSOpenalties with Gaussian graphical model, and integrates the computational predictionswith gene expression profiles to systematically study the gene regulatory network.We apply GALASSO to construct the regulatory network of breast cancer. Wereveal the structure of the regulatory network, and explore the role of feed forwardloops in gene regulation. In addition, we discuss the combinatorial regulatory effectbetween TF and miRNA, and provide detail analysis of their role in cancer.Meanwhile, we develop a new combinatorial regulation paradigm which is formedby CRM and miRNA. We examine the expression pattern of its target genes, and investi-gate the regulatory network composed of CRM and miRNA to discover the mechanismunderlying their co-regulation and interaction. Furthermore, we discuss miRNA andCRM’s effect on embryonic development.Generally, we provide a comprehensive and detail analysis of gene regulatory net-work of TF, CRM and miRNA. This study helps us to gain further understanding ofgene regulation, and facilitates us to explore the mysteries of life and provide valuablesuggestions on clinical study.