利用資料採礦技術於巴金森氏症之微陣列基因篩選

Using Data Mining Technique to Rediscover Parkinson's Disease Candidate Genes from Microarray Data

10.6338/JDA.201312_8(6).0005

陳佳怡(Chan-Yi Chen)；謝邦昌(Ben-Chang Shia)；侯藹玲(Ai-Ling Hour)

巴金森氏症 ； 微陣列 ； 資料採礦 ； 判別分析 ； 羅吉斯迴歸 ； 支援向量機 ； 隨機森林 ； Parkinson's Disease ； Microarray ； Data Mining ； Discriminant Analysis ； Logistic Regression ； Support Vector Machine ； Random Forests

Journal of Data Analysis

8卷6期（2013 / 12 / 01）

91 - 109

繁體中文

隨著醫療科技發展、教育和經濟變動等因素，導致全球逐漸邁入高齡化社會，重視老年相關疾病議題已成為社會關注的焦點，然而近年來巴金森氏症已被列為三大老年人疾病之一，其罹患人口增加的最主要關鍵在於老年人逐年增加，故巴金森氏症為高齡化社會不可忽視的問題。而目前巴金森氏症之主要治療方法還是以內科為主、外科為輔方式進行治療，且尚無藥物可以完全治癒此疾病。本研究利用Affymetrix GeneChip Human Genome U133 Plus 2.0 Array此基因晶片的Microarray資料進行研究，分別為GSE6613和GSE8397兩組資料集，本研究先利用判別分析與二元羅吉斯迴歸篩選出可用於辦別出罹患巴金森氏症之基因，並使用支援向量機與隨機森林進行模型之建立，最後利用GO基因功能分類資料庫，其中GSE6613為血液樣本，故期盼給予未來生物與醫學之研究上，亦可以藉由血液樣本來辨別出是否為罹患巴金森氏症之高風險群。

With the development of medical technology, education and economic, the global gradually become an aging society. Age-related diseases have become the focus of attention. In recent years, Parkinson's disease has been listed as one of the three major age-related diseases. The number of patients increasing mainly due to the annual increment of the elder people. Therefore, Parkinson's disease cannot be ignored in the society. So far, the main treatment of Parkinson's disease is the medical and supplemented by the surgical. However there is no drug that can completely cure the disease. We used the Affymetrix GeneChip Human Genome U133 Plus 2.0 Array, GSE6613 and GSE8397 data sets downloaded from NCBI. We selected the genes significantly differential expression for discriminant analysis and binary logistic regression. Then, we build the model with support vector machine and random forest. Finally, we found out the gene ontology terms the significant genes enriched, where GSE6613 is blood samples provided some useful index for researchers in distinguish Parkinson's disease in the future.

1. http://www.parkinson.org/About-Us/Press-Room/NPF-In-The-News/2011/April/People-With-Signs-of-PD-Wait-to-See-Doctor
2. http://www.nhi.gov.tw/webdata/webdata.aspx?menu=6&menu_id=168&WD_ID=168&webdata_id=2924
3. http://thchou.blogspot.tw/2009/03/logistic-regression.html
4. http://www.doh.gov.tw/CHT2006/DM/DM2_p01.aspx?class_no=24&nw_fod_list_no=10973&level_no=2&doc_no=74601. [March 15, 2010 accessed]
5. http://www.epochtimes.com/b5/9/5/15/n2527596.htm
6. http://www.genome.jp/kegg/
7. http://www.pdcenterntuh.org.tw/ContentAspx/Browser00.aspx?isClass=3&setTable=1&type=3
8. Bonifati, V.,Rizzu, P.,Baren, M. J.,Schaap, O.,Breedveld, G. J.,Krieger, E.,Dekker, M. C. J.,Squitieri, F.,Ibanez, P.,Joosse, M.(2002).Mutations in the DJ-1 Gene Associated with Autosomal Recessive Early-Onset Parkinsonism.Science,299,256-259..
9. Maraganore, D. M.,Lesnick, T. G.,Elba, A.,Chartier-Harlin, M. C.,Gasser, T.,Kruger, R.,Hattori, N.,Mellick, G. D.,Quattrone, A.,Satoh, J.(2004).UCHL1 is a Parkinson's disease susceptibility gene.Annals of Neurology,55,512-521.
10. Pankratz, N.,Pauciulo, MW.,Elsaesser, VE.,Marek, DK.,Halter, CA.,Wojcieszek, J.,Rudolph, A.,Shults, CW.,Foroud, T.,Nichols, W.C.(2006).Mutations in DJ-1 are rare in familial Parkinson disease.Neurosci Lett,408,209-213.
11. Polymeropoulos, MH.,Higgins, JJ.,Golbe, LI.,Johnson, WG.,Ide, SE.,Di Iorio, G.,Sanges, G.,Stenroos, ES.,Pho, LT.,Schaffer, AA.(1996).Mapping of a gene for Parkinson's disease to chromosome 4q21-q23.Science,274,1197-1199.
12. Ramirez, A.,Heimbach, A.,Grundemann, J.,Stiller, B.,Hampshire, D.,Cid, LP.,Goebel, I.,Mubaidin, AF.,Wriekat, AL.,Roeper, J.(2006).Hereditary parkinsonism with dementia is caused by mutations in ATP13A2, encoding a lysosomal type 5 P-type ATPase.Nat Genet,38,1184-191.
13. 朱迺欣(1998)。巴金森病：認識與面對。台北市:健康世界雜誌社。
14. 黃春常(1994)。高齡化社會。嘉義市:紅豆出版社。
15. 謝邦昌(2001)。資料採礦入門及應用：從統計技術看資料採礦。臺北市:資商訊息顧問。
16. 謝邦昌、鄭宇庭、蘇志雄(2009)。Data Mining 概述以 Clementine 12.0 為例。臺北縣新莊市:中華資料採礦協會。