基于网络分析的肥胖和相关疾病的关系研究

发布时间：2019-01-24 18:48

【摘要】：随着基因组测序的完成和新一代测序技术的发展,人类已经掌握了大量生物数据,而且,基因以及蛋白质与蛋白质相互作用网络方面的数据也在不断更新及丰富。通过这些数据来分析肥胖与疾病之间的关系和发现影响这种联系的关键基因,是研究与肥胖相关的疾病机理的重要方法,对基因组学和医学也具有现实意义。众所周知,肥胖与多种疾病有关,是许多疾病的主要危险因素,如II型糖尿病、冠心病和心血管疾病等。然而,肥胖在相关疾病的发展中起着重要的作用还没有被很好地理解。而且,目前也缺乏对于肥胖和相关疾病之间的全面研究。为了解决这个问题,我们构造三种不同的网络分析算法,第一个名字是OBNet,它主要是基于一个类似基因集富集分析和一个随机游走过程的算法;第二个算法叫OBsp,是一种基于最短路径的算法;最后一个叫OBoverlap,是基于直接求交集算法。我们分析比较了三种算法,发现基于扩展的模块化网络的OBNet方法是最优的,然后我们用这种方法来进一步研究肥胖与其相关疾病之间的分子层次的关联和潜在的功能联系,并有助于临床医学的深入认识。本文主要基于肥胖基因和疾病基因数据,提出了一种新的研究肥胖和疾病全局关系的网络分析方法,主要完成了以下两个方面工作:1)提出三种不同的网络分析算法从全局角度来分析肥胖和疾病之间的关系。通过比较三种不同算法的结果,选择OBNet-基于扩展的模块化网络可以更好的鉴定肥胖和疾病之间的关系。根据OBNet-基于扩展的模块化网络方法,我们可以找到与肥胖关系密切的一些疾病,以及与肥胖相关的疾病在哪些通路或子网络上与肥胖显著富集关联。最后具体分析了两个特定疾病,预测了调节这两个疾病与肥胖间关系的关键驱动基因。2)基于乳腺癌基因表达数据,WGCNA算法可以得到29个模块,我们抽取其中与乳腺癌显著相关的前10个模块;然后根据OBNet-基于扩展的模块化网络方法,我们可以得到的乳腺癌最显著富集的前10个子网络;把这10个字网络与WGCNA的前10个模块相比较,发现了两者有显著重叠,这说明我们的OBNet方法在不依赖基因表达谱的情况下,也可以找到疾病高度相关的一些驱动基因和模块。
[Abstract]:With the completion of genome sequencing and the development of new generation of sequencing technology, human beings have mastered a large number of biological data, and the gene and protein-protein interaction network data are constantly updated and enriched. The analysis of the relationship between obesity and disease and the discovery of the key genes affecting the relationship are important methods to study the mechanism of obesity related diseases, and also have practical significance for genomics and medicine. Obesity is known to be a major risk factor for many diseases, such as type II diabetes, coronary heart disease and cardiovascular disease. However, the important role of obesity in the development of related diseases has not been well understood. Moreover, there is a lack of comprehensive research on obesity and related diseases. In order to solve this problem, we construct three different network analysis algorithms. The first is called OBNet, which is mainly based on a similar gene set enrichment analysis and a random walk process algorithm. The second is called OBsp, which is based on the shortest path, and the last is called OBoverlap, which is based on the direct intersection algorithm. We analyzed and compared three algorithms and found that the OBNet method based on extended modular network is optimal. Then we use this method to further study the molecular level association and potential functional association between obesity and its associated diseases. It is helpful for the further understanding of clinical medicine. Based on the data of obesity gene and disease gene, this paper proposes a new network analysis method to study the global relationship between obesity and disease. The main contributions are as follows: 1) three different network analysis algorithms are proposed to analyze the relationship between obesity and disease from a global perspective. By comparing the results of three different algorithms, the extended modular network based on OBNet- can better identify the relationship between obesity and disease. Based on OBNet- 's extended modular network approach, we can find out some diseases closely related to obesity, and which pathways or sub-networks are significantly associated with obesity enrichment for obesity related diseases. Finally, two specific diseases are analyzed, and the key driving genes to regulate the relationship between the two diseases and obesity are predicted. 2) based on the breast cancer gene expression data, the WGCNA algorithm can get 29 modules. We extracted the top 10 modules significantly related to breast cancer; Then, according to OBNet- 's extended modular network method, we can get the first 10 subnetworks that are most significantly enriched in breast cancer. Comparing these 10 word networks with the first 10 modules of WGCNA it is found that there is a significant overlap between them which indicates that our OBNet method can also find some driving genes and modules which are highly related to the disease without relying on gene expression profiles.
【学位授予单位】：河北科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R589.2

【参考文献】