凝胶图像间的蛋白点匹配方法研究

发布时间：2019-04-22 11:45

【摘要】：人类基因组全序列测定的完成标志着后基因时代的到来，生命科学从基因组序列分析转向基因功能的研究。蛋白质组学是后基因时代研究中的重要组成部分，双向凝胶图像分析技术是蛋白质组学研究的重要技术，差异蛋白点提取和分析是双向凝胶图像分析技术的核心内容，蛋白点匹配又是差异蛋白点提取的关键环节。蛋白点匹配算法可用于鉴定潜在差异蛋白质分子，能够为疾病诊断、药物研制和环境污染分析提供依据。本论文以凝胶电泳图像蛋白点为研究对象，探索凝胶图像间的蛋白点匹配方法。主要研究工作、成果和创新点如下：首先，介绍了蛋白组学的发展和主要关键技术，概述了本课题的研究背景意义，综述了凝胶图像蛋白点匹配方法。然后，通过添加Landmark标记点和运用近邻点机制，提出了基于Landmark的两凝胶图像间的蛋白点半自动匹配算法。算法通过采用Landmark区域划分法和对应区域和跨区域蛋白点匹配提高了算法精度，并从Landmark数目和图源两角度验证了算法的有效性。但此方法为半自动方法，，需要人工介入。其次，依据凝胶电泳图像蛋白点的分布特征，提出了先特征蛋白点的粗匹配再两图像间蛋白点精匹配的自动匹配算法。首先采用统计学原理计算匹配距离阈值；然后采用区域划分和相似性原理进行特征点提取及粗匹配；其次利用匹配到的特征点建立两图像间映射关系；最后实现两图像蛋白点间的精确匹配。通过真实凝胶图像验证了算法的有效性。最后，从医学分析效果角度，采用了先组内两两匹配并生成合成胶，再合成胶蛋白点匹配的多幅凝胶图像间蛋白点匹配策略。将凝胶图像间的蛋白点匹配方法嵌入到凝胶图像蛋白点分析软件Protein Master中，并对Protein Master软件匹配系统中的各子模块进行了简单介绍。
[Abstract]:The completion of complete sequencing of the human genome marks the arrival of the post-gene era, and life science has shifted from genomic sequence analysis to gene function research. Proteomics is an important part of post-gene age research. Bi-directional gel image analysis is an important technology in proteomics research. Differential protein spot extraction and analysis is the core of bi-directional gel image analysis. Protein dot matching is also the key to differential protein spot extraction. Protein dot matching algorithm can be used to identify potentially differential protein molecules, which can provide basis for disease diagnosis, drug development and environmental pollution analysis. In this paper, the protein spots of gel electrophoresis images are studied, and the matching methods of protein spots between gel images are explored. The main research work, achievements and innovations are as follows: firstly, the development and key technologies of proteomics are introduced, the background significance of this subject is summarized, and the methods of protein dot matching in gel image are summarized. Then, a semi-automatic protein point matching algorithm between two gel images based on Landmark is proposed by adding Landmark labeling points and applying the nearest neighbor point mechanism. The algorithm improves the accuracy of the algorithm by using the Landmark region partition method and the matching of the corresponding and cross-region protein spots. The effectiveness of the algorithm is verified from the point of view of the number of Landmark and the source of the map. However, this method is semi-automatic and requires manual intervention. Secondly, according to the distribution characteristics of protein spots in gel electrophoresis images, an automatic matching algorithm based on coarse matching of feature protein spots and fine matching of protein spots between two images is proposed. Firstly, the matching distance threshold is calculated by using the statistical principle; then the feature points are extracted and coarse matching is carried out by using the region division and similarity principle; secondly, the mapping relationship between the two images is established by using the matching feature points. Finally, the exact matching between the two image protein spots is realized. The validity of the algorithm is verified by real gel images. Finally, from the point of view of the effect of medical analysis, a multi-image matching strategy of protein spots between gel images was adopted, in which pairwise matching was first used to generate synthetic glue and then synthetic gelatinous spot matching. The protein point matching method between gel images is embedded into the gel image protein dot analysis software Protein Master and the sub-modules of the Protein Master software matching system are briefly introduced.
【学位授予单位】：南昌航空大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：R346

【参考文献】