面向宫颈癌HPV分型中序列结构分析方法研究与突变分析

发布时间：2018-02-25 19:11

本文关键词： 人乳头瘤病毒 HPV分型结构相似性比较 HPV突变结构域　出处：《浙江理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：宫颈癌是全世界女性最常患的肿瘤类妇科疾病之一,其发病率仅次于乳腺癌。大量的基础与临床研究发现,高危险型HPV持续感染是诱发宫颈癌的关键因素之一。近年来,中国报道了大量的HPV高危型的突变数据,与野生型病毒相比,感染宫颈的HPV高危型存在多种突变模式。目前,大部分宫颈癌HPV的研究只关注通过HPV序列本身,忽略了临床病变信息。本文以为宫颈癌HPV分型中蛋白质序列、结构为对象,围绕蛋白质结构比较、分型预测模型,以及HPV突变与序列、结构之间的关系开展研究。具体工作如下:1、综述了人乳头瘤病毒与宫颈癌的基础知识,包括人乳头瘤病毒分型、结构、功能、人乳头瘤病毒导致的相关疾病、人乳头瘤病毒与宫颈癌的关系,为本文接下来的研究工作提供了理论基础和依据。2、提出了一种基于马尔科夫随机场的蛋白质结构相似性分析方法。在距离矩阵分布和不同节点的邻域系统的基础上,建立了改进的接触图矩阵,并通过计算马尔科夫随机场中的条件概率度量蛋白质结构之间的差异。结果表明,本文提出的蛋白质结构比较方法可以有效地度量不同多肽或者蛋白质结构之间的差异。此外,本文还发现alpha-C、O、和N端包含重要的结构信息,而侧链的原子集团会影响到模型的效率;通过分析马尔科夫随机场邻域系统的阶数,发现效率最高的马尔科夫随机场往往采用2个节点的邻域系统。3、构建了一个基于氨基酸特性的宫颈癌HPV分类预测模型。本文采用氨基酸的物化性质对20种常见氨基酸进行约化,6种特征信息提取方法提取蛋白质序列信息,利用支持向量机实现对宫颈癌HPV分型预测。实验表明,本文提出的预测模型可以准确地识别高危型HPV和低危型HPV,比现有的方法更有效。此外,本文还发现若利用E5、E6、E7、L1和L2蛋白质对HPV分型,最好选择氨基酸的beta类物化性质进行约化;若利用E1、E2、E4、E5和E7蛋白质,则PRseAAC蛋白质特征表现最优;而对于E6、L1和L2蛋白质,RTCD这类蛋白质特征的表现优于其余特征。4、研究了HPV突变与序列、结构之间的关系。本文通过文献检索,整理了大量的国内宫颈癌HPV的突变数据,并研究了突变位点与序列保守区域、结构保守区域的关系。结果表明,E6、E7和L1蛋白质中突变个数分别是134、86和166,远远大于其余蛋白质的突变数量;E2 N端的3159突变可以改变蛋白的免疫功能;HPV低危型E6有11个突变,其中9个突变落在p53蛋白结合区域或者是抗原决定簇区域;HPV高危型的E6有91个突变,有49个突变种类落在p53蛋白结合区域或者是抗原决定簇区域;突变落在E7功能域内的比例最高,大约93%以上的突变都落在了功能域区。
[Abstract]:Cervical cancer is one of the most common tumorous gynecological diseases in women in the world, and its incidence is second only to breast cancer. A large number of basic and clinical studies have found that persistent infection of high-risk HPV is one of the key factors to induce cervical cancer. China has reported a great deal of mutation data of HPV high-risk type. Compared with wild-type virus, there are many mutation patterns in HPV high-risk type infected with cervix. At present, most studies on cervical cancer HPV are focused only on the HPV sequence itself. The clinical pathological information was ignored. The protein sequence and structure in HPV typing of cervical cancer were considered as the object. The protein structure was compared, the prediction model of typing, and the mutation and sequence of HPV were discussed. The relationship between human papillomavirus and cervical cancer is summarized as follows: 1. The basic knowledge of human papillomavirus and cervical cancer is summarized, including human papillomavirus typing, structure, function, and related diseases caused by human papillomavirus. The relationship between human papillomavirus and cervical cancer, This paper provides the theoretical basis and basis for the next research work in this paper. 2. A method of protein structure similarity analysis based on Markov random field is proposed. Based on the distance matrix distribution and the neighborhood system of different nodes, An improved contact graph matrix is established, and the difference between protein structures is measured by calculating the conditional probability in Markov random field. The protein structure comparison method proposed in this paper can effectively measure the differences between different peptides or protein structures. In addition, we also find that alpha-CO and N-terminal contain important structural information. The cluster of atoms in the side chain will affect the efficiency of the model, and by analyzing the order of Markov random field neighborhood system, It is found that the most efficient Markov random field usually uses the neighborhood system of two nodes. 3. A classification and prediction model of cervical cancer based on amino acid characteristics is constructed. In this paper, the physicochemical properties of amino acids are used to predict 20 kinds of common ammonia. The protein sequence information was extracted by six kinds of feature information extraction methods. The prediction of cervical cancer HPV classification using support vector machine is realized. The experimental results show that the proposed prediction model can accurately identify high-risk HPV and low-risk HPV, which is more effective than the existing methods. It is also found that if we use E5, E6, E7, L1 and L2 proteins to type HPV, it is better to select the physicochemical properties of amino acids for beta reduction, and if we use E1, E2, E4, E5 and E7 proteins, the characteristics of PRseAAC proteins will be the best. The relationship between HPV mutation, sequence and structure was studied. A large number of HPV mutation data of cervical cancer in China were collected by literature retrieval. The mutation sites and conserved regions were studied. The results showed that the number of mutations in E6 E7 and L1 proteins was 134n86 and 166respectively, which was much larger than that in the other proteins. 3159 mutations at the N terminal of E2 could change the immune function of HPVE6. There were 11 mutations in E6. Among them, 9 mutations were found in p53 protein binding region or antigen determinant region. There were 91 mutations in E6, and 49 mutations in p53 protein binding region or antigen determinant cluster. The proportion of mutations in the E7 functional domain is the highest, and about 93% of the mutations fall into the functional domain.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R737.33

【相似文献】