基于生物网络的疾病microRNA挖掘技术研究

发布时间：2018-06-02 21:43

  本文选题：生物信息学 + 生物网络　； 参考：《哈尔滨工业大学》2010年博士论文

【摘要】： 非编码RiboNucleic Acid (RNA)是生物信息学领域当前的研究热点。步入21世纪以来,非编码RNA的相关研究连续获得Science评选的年度十大科学突破,并在2006年获得了诺贝尔生理学或医学奖。MicroRNA是一类重要的非编码RNA,它的异常能导致人类疾病的发生、发展。通过生物实验的方法能够挖掘疾病microRNA,但是实验方法代价高、周期长。本文从生物信息学的角度提出四种疾病microRNA的挖掘方法,挖掘出潜在的导致该疾病发生的microRNA,从而为生物学、医学研究者有针对性地进行microRNA生物实验提供一定指导,进而为药物开发、临床诊断治疗提供一定的依据。本文的主要内容包括: (1)挖掘及分析已知的microRNA与疾病关系 自2002年以来,越来越多的研究证明microRNA失调有助于疾病的发生发展,然而这些已知的疾病与microRNA关联关系分散在已发表的文献当中,目前还没有研究机构建立在线共享的数据库,收集、存储、管理这些数据;科研人员不易获取这些已知的microRNA与疾病的关联信息。因此我们先从文献中挖掘已知的疾病microRNA ,构建了全球首个microRNA与疾病关系数据库(miR2Disease),并对数据进行管理。对miR2Disease中的数据进行分析,发现多种疾病往往共享一些致病microRNA,拥有部分相似的发病机制;此外,总结出疾病microRNA失调的三种机制:首先,疾病microRNA常位于与疾病有关的基因座内,例如杂合子缺失的微小区域、微小扩增区域或断裂位点等脆性位点区域;其次,疾病microRNA失调是由异常的表观遗传信息改变所致;例如DNA异常甲基化、组蛋白异常修饰等等;最后,疾病microRNA失调是由参与microRNA生物合成的酶的功能异常所致。 (2)提出基于布尔网络的疾病microRNA挖掘技术 生物网络在挖掘编码蛋白的疾病基因方面发挥了重要作用,然后在疾病microRNA挖掘领域,至今还未提出基于生物网络的疾病microRNA挖掘方法。因此本文提出了构造布尔型的功能相关microRNA网络的算法,以网络的形式来研究microRNA。通过对网络的分析,我们发现布尔型microRNA网络像其他生物网络一样,网络的度服从幂分布,网络具有层次模块性等特点。我们进一步构建了phenome-microRNAome网络,在此网络上,对已知的疾病microRNA进行分析,发现“功能相关的microRNA失调倾向于导致表型相同或相似的疾病”这一规律。以此为理论基础,提出了基于布尔型生物网络的疾病microRNA挖掘算法,并验证了算法的有效性。 (3)提出基于权重型网络的疾病microRNA挖掘技术 基于布尔网络的疾病microRNA挖掘技术在构造布尔型microRNA网络只需根据靶基因重叠的显著性来确定二个microRNA之间的关联关系。当知道microRNA对靶基因的抑制强度信息时,可以利用该信息构建权重型网络。因此,我们提出了基于权重型网络的疾病microRNA挖掘方法,取得了很好的性能。 (4)提出基于支持向量机的疾病microRNA挖掘方法 为了直接从数据出发挖掘疾病microRNA,我们把疾病microRNA的挖掘问题转化为一个分类问题,提出了基于支持向量机的疾病microRNA挖掘方法,把数据挖掘、机器学习的思想引入到疾病microRNA的挖掘中并交叉验证了方法的有效性。 (5)提出基于基因组数据融合的疾病microRNA挖掘技术 统计数据表明近三年获得的生物医学数据超过过去四万年的总和,数据呈爆炸增长,面对浩瀚的生物学数据海洋,如何把这海量的数据转化为有意义的医学诊断和治疗信息并惠及人类自身的健康是21世纪生物医学信息学面临的严峻挑战。本章整合了多种生物数据资源,构建了全人类基因组范围的基因功能相关网络,在此网络基础上,提出了利用microRNA的靶基因与已知的感兴趣疾病的致病基因之间在网络上的功能关系来挖掘新的潜在的疾病microRNA的算法,将算法应用到结肠癌上验证了算法的有效性。
[Abstract]:Non coded RiboNucleic Acid (RNA) is the current research hotspot in the field of bioinformatics. Since entering twenty-first Century, the related research of non coded RNA has continuously obtained the ten major scientific breakthroughs of the annual Science selection, and in 2006 the Nobel prize for physiology or medicine.MicroRNA is an important non coded RNA, and its abnormality can lead to human beings. The occurrence and development of the disease can be developed through biological experiments, but the experimental method is able to excavate the disease microRNA, but the experimental method is expensive and the cycle is long. From the perspective of bioinformatics, this paper proposes a mining method of four diseases, microRNA, to excavate the potential microRNA that causes the disease to occur, from the biological, and the medical researchers are targeted to the mic. RoRNA biological experiments provide some guidance for further development of drugs and clinical diagnosis and treatment.
(1) mining and analyzing the known relationship between microRNA and disease
Since 2002, more and more studies have shown that microRNA disorders contribute to the development of the disease. However, the relationship between these known diseases and microRNA is scattered in the published literature. There is no research organization to establish online shared databases, collection, storage, and management of these data; researchers are not easy to obtain these already. We know the association information between microRNA and disease. So we first excavated the known disease microRNA from the literature, constructed the world's first microRNA and disease relational database (miR2Disease), and managed the data. We analyzed the data in miR2Disease and found that a variety of diseases often share some of the pathogenic microRNA and have some similarities. In addition, three mechanisms of disease microRNA disorders are summarized: first, the disease microRNA is often located in the loci related to the disease, such as the tiny region of the heterozygote deletion, the small amplification area or the fracture site and other fragile sites; secondly, the disorder of the disease microRNA is caused by abnormal epigenetic information; Abnormal methylation of DNA, histone modification, and so on. Finally, the disorder of microRNA is caused by dysfunction of enzymes involved in microRNA biosynthesis.
(2) propose microRNA mining technology based on Boolean network.
Biological network plays an important role in mining the disease genes encoding protein, and then in the field of disease microRNA mining, so far, the disease microRNA mining method based on biological network has not been proposed. Therefore, this paper proposes a algorithm for constructing a Boolean functional related microRNA network, which is used in the form of network to study microRNA. through the form of network. Network analysis, we find that Boolean microRNA networks like other biological networks, the degree of the network is subordinate to power distribution, and the network has the characteristics of hierarchical modularity. We further construct the phenome-microRNAome network, on which the known disease microRNA is analyzed, and that "functional related microRNA disorders tend to lead to guidance." On the basis of the theory, a disease microRNA mining algorithm based on Boolean biological network is proposed, and the effectiveness of the algorithm is verified.
(3) put forward the microRNA mining technology based on weight heavy network.
The disease microRNA mining technology based on Boolean networks can determine the association between two microRNA based on the significance of target gene overlap in constructing Boolean microRNA networks. When the microRNA is known to suppress the target gene, we can use the information to build a heavy network. Therefore, we propose a weight based heavy weight network. Network disease microRNA mining method has achieved very good performance.
(4) propose a method of disease microRNA mining based on support vector machine.
In order to mine disease microRNA directly from data, we transform the problem of disease microRNA into a classification problem, and propose a disease microRNA mining method based on support vector machine, which introduces the idea of data mining, machine learning to the mining of disease microRNA and cross verifies the effectiveness of the method.
(5) propose microRNA mining technology based on genome data fusion.
Statistics show that the biomedical data obtained in the last three years are more than the sum of the past forty thousand years, and the data are increasing. In the face of the vast ocean of biological data, how to translate this mass of data into meaningful medical diagnosis and treatment information and benefit human health is a severe challenge for biomedical informatics in twenty-first Century. In this chapter, a variety of biological data resources are integrated, and the gene function related network of the whole human genome is constructed. On the basis of this network, a new algorithm for mining potential disease microRNA using the functional relationship between the target gene of microRNA and the pathogenetic genes of the interested diseases is proposed, and the algorithm should be used. The effectiveness of the algorithm was verified with colon cancer.
【学位授予单位】：哈尔滨工业大学
【学位级别】：博士
【学位授予年份】：2010
【分类号】：R341

【引证文献】

相关硕士学位论文 前1条

1 刁丽红;蛋白质相互作用网络在癌症研究中应用以及金属抗癌配合物的合成[D];广西民族大学;2012年

，

本文编号：1970302