基于互信息基因调控网络构建及其在甲状腺肿瘤基因分析中应用的研究

发布时间：2018-09-12 10:32

【摘要】：随着生命科学和计算机科学迅速的发展,使得生命科学和计算机科学相互结合形成一门新学科,即生物信息学。生物信息学通过研究生物信息的采集、处理、存储和传播等,分析和解释生物学秘密。然而要从系统的角度研究生物系统中的信息,研究生物信息系统中的所有组成成分(基因、mRNA、蛋白质等)之间的相互作用关系和作用机制成为一个必要的过程。生物体遗传、发育、进化和疾病等过程中的基因与基因的相互作用或因果关系可以概括为基因调控,这是生物信息学研究的一个重要课题。通过对基因调控网络的研究可以帮助我们理解并认识基因的发育和基因调控过程的因果关系,能够帮我们进一步对生物功能、行为的研究起到决定性的作用。本文研究发现,通过对患病基因和非患病基因调控网络的研究可以帮助我们找到跟疾病密切相关的基因以及基因的表达水平,从而从基因层面找到致病的原因,进而通过分析尝试对基因进行药物或是物理方式的治疗,达到治疗疾病的效果。然而现阶段仅仅通过基因表达数据构建调控网络的思路并不具有较强的生物学意义,所以本文研究将RNA-Seq数据、miRNA数据和合理的生物学信息都包含到基于互信息的网络构建过程中,通过合理的使用大范围的数据使构建出的网络具有更强的生物学意义;同时在进行患病基因差异筛选的过程中不仅仅使用简单的基因差异筛选算法进行差异筛选,通过利用引入Bootstrap自助法多次抽样进行基因差异筛选。最终将此思路运用到甲状腺肿瘤数据中,探索更加有意义、更加准确的信息,能够对预测、治疗甲状腺疾病提供有科学依据的建议。本文研究过程如下:(1)数据的获取。从TCGA平台获取甲状腺肿瘤的RNA-Seq数据和miRNA数据并进行预处理。(2)对RNA-Seq数据进行正常与患病差异基因的筛选,寻找miRNA的靶基因。(3)对以上数据通过基于互信息的BC3NET基因调控网络构建方法进行网络构建,并对构建的基因调控网络进行模块划分。(4)对划分的子网模块进行验证和分析,得出对预测、治疗甲状腺疾病具有建设性意义的结论。(5)总结以上研究工作的过程形成一套可行的从构建基因调控网络到对疾病进行预测和治疗的方法或思路。在未来的工作中,还可以通过加入甲基化数据,Gene Card上的数据使得我们构建的基因调控网络更加具有研究价值。通过加入更多的生物学信息,可以针对性地对疾病从分子的层面提出建设性的预防和治疗的建议和依据。
[Abstract]:With the rapid development of life science and computer science, life science and computer science combine to form a new subject, bioinformatics. Bioinformatics analyzes and explains biological secrets by studying the collection, processing, storage and dissemination of biological information. However, it is necessary to study the interaction and mechanism of all components (gene mRNAs, proteins, etc.) in biological information systems from a systematic perspective. The interaction or causality between genes in the processes of heredity, development, evolution and disease can be summarized as gene regulation, which is an important subject in bioinformatics. The study of gene regulation network can help us to understand and understand the causal relationship between gene development and gene regulation process, and can help us to further study biological function and behavior. In this study, we found that the study of the regulatory network of diseased and non-diseased genes can help us to find the genes closely related to disease and the level of gene expression, and thus to find the cause of the disease at the gene level. Through analysis, we try to treat the gene by medicine or physical method, and achieve the effect of treating disease. However, the idea of constructing regulatory networks based on gene expression data is not of great biological significance. Therefore, this paper studies the inclusion of RNA-Seq data and reasonable biological information into the process of network construction based on mutual information. Through the rational use of a wide range of data to make the constructed network has stronger biological significance, and in the process of disease gene differential screening, not only the use of simple genetic differential screening algorithm for differential screening, Genetic diversity screening was carried out by introducing Bootstrap self-help method. Finally, this idea is applied to thyroid tumor data to explore more meaningful and accurate information, which can provide scientific advice for prediction and treatment of thyroid diseases. The research process of this paper is as follows: (1) data acquisition. The RNA-Seq data and miRNA data of thyroid tumor were obtained from TCGA platform and processed. (2) screening the difference gene between normal and diseased RNA-Seq data. To find the target gene of miRNA. (3) to construct the BC3NET gene regulatory network based on mutual information, and divide the gene regulation network into modules. (4) verify and analyze the subnet module. The conclusion is constructive to predict and treat thyroid diseases. (5) A set of feasible methods or ideas from constructing gene regulatory network to predicting and treating thyroid diseases are formed. In the future work, we can also add methylation data to gene Card data to make our gene regulatory network more valuable. By adding more biological information, we can provide constructive advice and basis for prevention and treatment of disease at molecular level.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R736.1

【参考文献】