單一核甘酸多型性與子宮頸癌癌化因素之分析; Using a Machine Learning Approach to Derive Single Nucleotide Polymorphisms and Microsatellite Cofactors of Cervical Cancer

NCU Institutional Repository > 資訊電機學院 > 資訊工程研究所 > 博碩士論文 > Item 987654321/8706

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/8706

题名:	單一核甘酸多型性與子宮頸癌癌化因素之分析;Using a Machine Learning Approach to Derive Single Nucleotide Polymorphisms and Microsatellite Cofactors of Cervical Cancer
作者:	胡凱智;Kai-Chih Hu
贡献者:	資訊工程研究所
关键词:	微衛星重複性序列;單一核甘酸多型性;子宮頸癌;決策樹;貝式網路;microsatellite;Bayesian network;decision tree;Cervical cancer;single nucleotide polymorphism
日期:	2003-06-25
上传时间:	2009-09-22 11:33:19 (UTC+8)
出版者:	國立中央大學圖書館
摘要:	子宮頸癌是台灣以及全世界女性中主要癌症之一。近十幾年來的研究，某些人類乳突狀病毒已經被確定子宮頸癌的主要造成原因。雖然乳突狀病毒為子宮頸癌發生所必需，但是卻非充分的條件。很顯然有其他環境上的因素以及宿主反應在子宮頸上皮癌化的過程之中也扮演著重要的角色。人類基因體定序的完成以及家族血統相關研究也指出子宮頸癌中基因上的變異性擁有決定性的因素。因此，本研究將單一核甘酸多型性加入分析，以找出哪些基因性因素以及人類乳突狀病毒的組合與子宮頸癌的引發過程中有關。本研究將分析11個單一核甘酸多型性標記，4個microsatellites，年齡與人類乳突狀病毒，使用J48，PART，Id3和PRISM等決策樹分析演算法來分析。同時也比較了這些決策樹演算法的優劣。結果提供我們一個概念，就是我們的確可以將單一核甘酸多型性加入子宮頸癌因素分析，讓我們經由這些因子來預測並分析結果提供醫療用途。 Cervical cancer is a common cancer among women worldwide. Recently, infection with high-risk types of human papillomavirus (HPV) has been identified as the central cause of cervical cancer. Although HPV has been identified as playing the central role in cervical cancer, it is still insufficient to confer the cervical cancer. Obviously, there are other environmental and host factors which are involved in the progression of HPV infection to cancer. Recent population based twins and family studies have showed the hereditary component of cervical cancer, indicating genetic susceptibility plays an important role. Thus, to take some SNP markers and microsatellite into account as the genetic factors might be helpful to find out some combinations of these genetic factors and HPV that are involved in the progress of precancerous changes into cervical cancer. We take patients’ age, 11 SNP markers and 4 microsatellites into account and perform decision tree analysis using different learning algorithms. We also compare the performance among these learning algorithms. We anticipate that the results of this study will open the door for investigations of identifying the combinations of genetic factors such as SNPs and microsatellites that interact in a non-additive or nonlinear manner to influence risk of common complex multifactorial disease.
显示于类别:	[資訊工程研究所] 博碩士論文

文件中的档案:

档案	大小	格式	浏览次数

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....