【作者】 吴迪；

【导师】 孙咏梅；

【作者基本信息】 北京邮电大学 ， 通信与信息系统， 2015， 硕士

【摘要】 随着社会的进步与发展,信息技术和生命科学两个学科对人类社会产生了越来越深远的影响。随着相关研究不断深入,生命科学工作者所需处理的数据量越来越庞大,使用信息技术进行数据处理成为了主流趋势。在分子遗传学领域,受限于知识结构,RNA编辑事件相关研究人员并不能有效使用计算机等现代化工具对信息进行有效的处理,因此一套完整、清晰的信息化解决方案将具有重要的研究意义。本论文研究目标为：将信息处理技术应用于RNA(核糖核酸)编辑事件的相关研究,提出一种RNA识别机制,并对识别机制获取的结果位点进行特性分析。本论文提出的识别机制克服了传统识别机制零散和不规范等缺点,对海量原始信息进行统一、标准化的处理；在处理过程中,更加注重整体数据流程的透明性、可调节性,保留了流程中关键步骤的处理结果,便于识别机制的分析、改进,以及后续的升级、扩展功能等工作的展开。该机制可分为两大部分,即规则型滤除算法和统计型滤除算法。前者主要分析测序数据的质量、RNA编辑候选位点所处基因区域特点等方面,对不合规则的位点予以滤除；后者主要分析候选位点统计学特点,从显著性水平分析等方面入手,提高识别结果的有效性。为了体现RNA编辑事件的结果位点在分子遗传学领域的价值,本论文还对其进行了特性分析。本论文的源数据来自合作医学研究机构,在获得结果位点后,首先将其与传统处理分析软件的处理结果进行对比,以验证识别机制的效率和有效性；其次,分析RNA编辑事件结果位点的所属基因功能区域,并且对多位病人癌组织和正常组织进行了统计分析,呈现了一定数量的有价值的分析结果,最后依此作出了一些合理的结论判断。更多还原

【Abstract】 With the progress and development of society, information technology and life sciences are producing more and more profound impact on human life. With the deepening of life sciences research, the amount of data needed to process is larger and larger and the use of information technology for data processing has become a mainstream trend. In the field of molecular genetics, limited by the knowledge structure, researchers on RNA editing events can not effectively use computers to process data. Therefore, a complete and clear information solution will make sense.The research objective of this thesis is as follows:applying information processing technology to research of RNA (ribonucleic acid) editing events. Besides, algorithm for RNA recognition is proposed and characteristics analysis for RNA editing sites is performed.Recognition algorithm proposed in this thesis overcomes the shortcomings of traditional method which doesn’t have a unified and standardized process. It maintains the overall data flow adjustable and transparent, which facilitates analysis, improvement and subsequent upgrades. This algorithm includes two main parts, that is, rule-based filtering algorithms and statistics-based filter algorithms. Rule-based filter algorithms mainly take quality of sequencing data and candidate RNA editing sites’location in gene regions into consideration. Statistics-based filter algorithms analyze characteristics of candidate sites to filter out statistically unreliable data, which can improve the validity of the results.In order to reflect the value of RNA editing sites in the field of molecular genetics, this thesis also analyzes their characteristics. After processing the sequencing data from cooperative research agencies, this paper compare results got from the algorithm we proposed with those got from traditional method to verify the efficiency and effectiveness. Then, we analyze the candidate sites of RNA editing events on the classification distribution, and number of cancer and normal tissues from the patients. In the end, we present valuable analysis results in forms of graph and tables, and make several related conclusions.更多还原