基于集成网络的胶质母细胞瘤失调模块检测

发布时间：2018-02-08 20:16

本文关键词： 集成生物网络多形性胶质母细胞瘤失调模块 eQTL 随机游走　出处：《西安电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：疾病是制约人类寿命的最重要因素,以癌症为代表的复杂疾病严重影响着人类的正常生活和工作。生物医学家研究发现,人类所有健康和疾病状态都与基因有密切关系,如何诊断和治疗以癌症为代表的多基因复杂疾病一直是科学家研究的热点和难点之一。长久以来,由于生物实验技术和实验结果分析技术都不太成熟,对复杂疾病的认识不够深入,导致复杂疾病的研究一直没能取得有效进展。高通量生物技术的快速发展为复杂疾病的研究提供了可靠的数据来源,同时以蛋白质相互作用网络为代表的生物网络为复杂疾病的研究提供了很好的数据支持。目前,对复杂疾病的研究多集中于单个致病基因预测,而忽略了生物大分子之间的相互调控关系。研究发现不同基因的扰动会导致相同的表型,且这些基因倾向收敛于相同的细胞路径,基于此本文的研究重点由致病基因的预测转移到失调模块的检测,针对复杂疾病多形性胶质母细胞瘤进行失调模块检测。鉴于单个生物网络的数据不够完整,实验结果很大程度依赖特定数据集,本文整合蛋白质-蛋白质相互作用网络、转录因子-蛋白质相互作用网络以及蛋白质磷酸化网络,构建出集成生物网络。首先针对多形性胶质母细胞瘤的基因表达数据,使用SAM软件检测出差异表达基因作为目标基因;然后使用表达数量性状定位分析技术(eQTL)定位控制这些目标基因差异表达的遗传因子作为候选因果基因;最后在集成生物网络中,使用随机游走方法检测出因果基因以及最终的失调模块。本文最终成功检测出由172个基因及206条相互作用链接组成的失调模块,其中包含OMIM、KEGG以及AceView数据库中已知与多形性胶质母细胞瘤密切相关的基因,如CDKN2A、EGFR、TP53、AKT1、RB1、PTEN、PRKCA等。而基因RHOBOTN2、CEBPA和GBAS虽然不是OMIM、KEGG以及AceView数据库中收录的基因,但是本文通过文献挖掘找到了其与多形性胶质母细胞瘤密切相关的有力证据。而且,该失调模块包含了使用单个生物网络分析所得到的大部分致病基因。通过统计理论和生物意义两方面对实验结果进行分析验证,揭示出该失调模块与多形性胶质母细胞瘤的致病机制密切相关,同时也说明本文所提基于集成生物网络的方法是一种完整,有效的失调模块检测方法。
[Abstract]:Disease is the most important factor that restricts the life of human being. The complex disease represented by cancer seriously affects the normal life and work of human beings. Biomedical research has found that all human health and disease state are closely related to genes. How to diagnose and treat polygenic complex diseases represented by cancer has always been one of the hot and difficult points in the research of scientists. For a long time, because of the immaturity of biological experiment technology and experimental result analysis technology, The lack of understanding of complex diseases has led to a lack of effective progress in the study of complex diseases. The rapid development of high-throughput biotechnology provides a reliable source of data for the study of complex diseases, At the same time, biological networks, represented by protein interaction networks, provide a good data support for the study of complex diseases. At present, the researches on complex diseases are mostly focused on the prediction of single pathogenic genes. The study found that disturbances of different genes lead to the same phenotypes and that these genes tend to converge to the same cellular pathway. Based on this, the emphasis of this study is transferred from the prediction of pathogenic gene to the detection of maladjustment module, and the maladjustment module is used to detect multiple glioblastoma of complex diseases. In view of the incomplete data of single biological network, The results depend heavily on specific data sets. This paper integrates protein-protein interaction networks, transcription factor-protein interaction networks, and protein phosphorylation networks. Firstly, according to the gene expression data of glioblastoma pleomorphic, the differentially expressed genes were detected by SAM software as the target gene. Then we use quantitative trait mapping technique (QTL) to localize the genetic factors that control the differential expression of these target genes as candidate causality genes. Finally, in the integrated biological network, A random walk method was used to detect the causal gene and the final disorder module. In the end, the disorder module composed of 172 genes and 206 interaction links was successfully detected. It contains OMIMN-KEGG and genes known to be closely related to pleomorphic glioblastoma in the AceView database, such as CDKN2ANEGFRN TP53AEGFRT53AKT1rrrrrrrrrrrrbt1, PRKCA, etc. The genes RHOBOTN2CEBPA and GBAS are not included in OMIMKEGG and AceView databases, although they are not included in the OMIMKEGG and AceView databases. However, this paper has found strong evidence of its close relationship with glioblastoma pleomorphic by literature mining. The disorder module contains most of the disease-causing genes obtained by using a single biological network analysis. The experimental results are verified by statistical theory and biological significance. It is revealed that the disorder module is closely related to the pathogenesis of glioblastoma pleomorphic, and that the method proposed in this paper based on integrated biological network is a complete and effective method for the detection of maladjustment modules.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R739.41

【相似文献】