Regulation Model Recognition Of Biological Interaction Network And Application In Drug Screening

It is known that the physical function of organisms usually perform through functional pathways to consist components, such as kinases and transcription factors, execut certain interactions between proteins and other cellular components, and to form the genetic changes and cellular development in life systems. The transcripton, which contains the transcriptional level regulation by the transcription factors and the post-transcriptional network of miRNA regulation and the epigenetic modifications, is one of the most common phenotypes of the disease research. With the development of life science technologies, the measurements to examine and understand the changes in life systems, including the cancer development and effective treatment, are becoming more and more advanced. Through the high-throughput microarray experimental techniques combined with computational analysis methods, we are able to level up our research from the traditional study that partially targets on singular cells to a systematical study that acts on multiple targets. Since the drug induced multiple targets are also influenced by a variety of regulating components, biological process and functions, a complex network is formed between the interactions of drugs and target genes, as well as those of genes and functional modules.The research of this paper is based on the following aspects in the study of regulatory interactions of molecular and functional modules in the biology network. First, we focused on the changes of characters in the network caused by different disease conditions, and the changes of the same network under the effect of drug perturbation. Secondly, we extracted useful interaction patterns to search for the efficient, accurate modules and the interaction between them in the life systems, so as to discover the biological knowledge of the drug action mechanism. In the end, we discussed the application of this strategy in the translational medicine.1. Based on the regulation between the regulatory component and its target genes, integrated with the context information of the target genes, we created the new network based on the new regulations.The new index was created to analyze the specific function in regulatory network, and was applied to drug screening. According to the specific context of the target genes, we defined a new vector to address the miRNA’s regulation in a specific context. The specific context is defined according to the biological processes and cellular components in Gene Oncology Term. Previously we have introduced the concept of CoMi (Context-Specific miRNA Activity), and it has provided a novel perspective on drug mechanisms of action. In this work, we intended to construct a virtual drug screening system and test whether CoMi could be used to capture the common features of approved cancer drugs, in order to discriminate them from other candidate chemicals. The CoMi refers to the influence of miRNA’s regulation on specific modules through the regulation of its target genes. According to the definition of CoMi, a differential disease model for breast cancer VS cancer tissues is established by microRNA and the context module, and a two-vector network is formed. Through the analysis of topological structure characteristics of the network, we selected the nodes with high connection degrees to identify the oncogene that plays an important role in the cancer prognosis. Finally, we discussed the function of miRNA regulation on specific contexts in the disease model, and used it in drug screening models.2. Based on the different definitions of the gene models, we constructed a new linteraction network based on the gene models, and used it in a drug discriminate model.First, we redefined and explored some new functional modules. According to the different molecules and their functions, we defined new functional modules based on different interactions between molecules and built new interaction networks. For example, according to the prognostic information, we selected the most related cellular component Terms and the biological processes Terms from the Gene Oncology, and define them as the prognostic related modules. After that we conducted correlation analysis by integrating more than100of drug’s sensitive data (GI50) with gene expression data from diverse cancer cell lines in NCI60datasets, and generated the drug sensitivity gene sets. Then we defined a perturbation Index by integrating the breast cancer survival data with the drug sensitivity information, representing the regulation between the gene modules of different biological meanings. In the next step, we investigated the regulations between the different gene modules, and searched for the relationship patterns between drug and disease by the perturbation Index as vectors. In the end, we successfully built a drug discrimination system using the perturbation Index as a vector to search for the most suitable treatment for a specific individual patient. Among140drug candidates, we are able to filter4FDA approved drugs and could identify2breast cancer drugs among4known breast cancer therapeutic drugs in total.3. Metagene approaches predict responses based on gene expression signatures derived from an associative analysis of clinical data. They can identify chance associations caused by the heterogeneity of a tumor, leading to reproducibility issues in independent validations. In this study, to incorporate information from drug mechanisms of action, we explore the potential of microRNA regulation networks as a new feature space for identifying predictive markers. Based on the previous work, we conducted a comprehensive study of the Paclitaxel, summarized the important role of Paclitaxel in breast cancer treatment, and defined new features based on interactions, extracted specific interaction characteristics that associated with the drug action mode of Paclitaxel. We introduced a measuring term CoMi (Context-specific-miRNA-regulation) pattern to represent a descriptive feature of the miRNA regulation network in the transcriptome. We examined whether the modifications to the CoMi pattern on specific biological processes are a useful representation of drug action by predicting the response to neoadjuvant Paclitaxel treatment in breast cancer and showed that the drug counteracts the CoMi network dysregulation induced by tumorigenesis. We then generate a quantitative testbed to investigate the ability of the CoMi pattern to distinguish FDA approved breast cancer drugs from other FDA approved drugs that are not related to breast cancer. We also compare the ability of the CoMi and metagene methods in predicting the response of neoadjuvant Paclitaxel treatment in clinical cohorts. We found that the CoMi method outperforms the metagene method. Furthermore, several of the predicted CoMi features highlighted the network-based mechanism of drug resistance. Thus, our study suggests that explicitly modeling the drug action using network biology brings forth a promising approach to predictive marker discovery and provides useful information for molecular biologists.To summary, in this paper we combined the fundamental knowledge of biology and systematic methods of bioinformatics, to analyze the topological properties of complex networks and the interactions between a variety of molecular and biological functional modules in the transcriptional regulatory network. In order to facilitate the development of transaction medical research, we aim to establish an effective mathematical model to achieve effective control and design of living systems.