最大熵算法在SNP与疾病关联性研究的应用

发布时间：2018-01-03 15:06

本文关键词：最大熵算法在SNP与疾病关联性研究的应用　出处：《西安电子科技大学》2011年硕士论文　论文类型：学位论文

更多相关文章： SNP 稳定性 最大熵 解交叉 bootstrap

【摘要】：单核苷酸多态性(SNP)是一种由人类DNA的等位基因变异导致的DNA的多态性。医学研究证明,SNP与人类的许多疾病都有着密切的联系,对SNP与人类疾病之间的相关性进行研究能够提早发现可能在患者身上出现的潜在疾病,在医学研究上能够对一些疑难杂症的形成原因提供理论依据,在对威胁人类健康的重大疾病的预防与治疗提供更加深入的诊断方式。同时这一研究目前在世界范围尚属前沿领域,具有重大意义。本文研究了通过最大熵对SNP与疾病相关性建模的方法。针对实验数据中存在的小样本、信息交叉的情况,提出了bootstrap、稳定性计算及解交叉等数据处理方法,分析了各个方法的原理及实验中的作用,通过这些方法对实验数据进行数据扩展、稳定性计算、信息解交叉等工作。同时从理论上分析了最大熵方法在本文工作中的适用性,并利用最大熵方法建立SNP与疾病相关性的模型。通过对真实SNP样本数据进行实验,求解出每个SNP模型的致病概率,并利用测试样本对结果的性能进行评价。通过实验验证了bootstrap最大熵方法对SNP与疾病相关性模型的优化作用,对比解交叉前后的实验结果,验证了解交叉方法对实验性能的提升作用。对实验中出现的问题做了分析,指出了SNP与疾病相关性研究的下一步工作思路。
[Abstract]:Single nucleotide polymorphism (SNP) is a kind of polymorphism caused by human DNA allelic variants of DNA. Medical studies have shown that many diseases and human SNP are closely linked, the correlation between SNP and human disease research can be found early in patients with underlying disease may appear on the body that can provide the theoretical basis for the formation of some Difficult miscellaneous diseases in medical research, provide more in-depth diagnostic methods in the prevention and treatment of major diseases threatening human health. At the same time, the present research is still in the world before the field is of great significance.
This paper studies the method of SNP and disease correlation modeling through the maximum entropy. According to the existing experimental data in small sample information, cross, put forward bootstrap, stability calculation and processing method of cross data analysis solutions, principle and experiment of each method in the role of experimental data expansion, stability calculation through these methods, information and so on. At the same time cross solution from the theoretical analysis of the applicability of the maximum entropy method in this work, and the maximum entropy method is used to establish the relationship between SNP and disease model. Through the experiment of real SNP data, calculate the pathogenic probability of each SNP model, and evaluate the performance of the the results of the test samples was verified by experiments. The optimization function of bootstrap maximum entropy method to SNP and disease correlation model, compared the experimental results before and after the solution of cross validation. Understand the improvement of cross method on the performance of the experiment. Analyze the problems in the experiment, and point out the next step of SNP related research.

【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：R394

【相似文献】