一套關於全基因相關性分析的標準流程

Abstract

有越來越多的證據顯示，全基因相關性研究在找出與常見人類疾病相關的基因是有效的方法。很多研究成功的利用全基因相關性研究找出新的致病位置。然而，目前並沒有一套一致性的分析程序。在這個研究中，我們檢閱目前的全基因相關性研究去整理出一套標準流程並利用WTCCC的真實資料去驗證。我們的流程包括四個部分：資料處理、預處理分析、相關性檢定、以及視覺化結果呈現。為了得到疾病與單體核苷酸多樣性的真實關係，我們做了兩個預處理程序，品質控制與母體分層。除此之外，我們可以畫Q-Q圖及Manhattan圖去視覺化我們的相關性分析結果。在研究的最後，我們成功的(1)確認必要且重要的分析程序(2)確認目前可用來做這些分析的軟體(3)利用WTCCC的資料完成了分析(4)提供了執行全基因相關性研究的一般指導方針。

There are increasing evidences that genome-wide association (GWA) studies represent a powerful approach to the identification of genes involved in common human diseases [1]. Many studies had successfully performed the GWA study to identify novel susceptible loci. However, there is a lack of agreement about what constitutes an adequate analytic procedure. In this study, we review existing genome-wide association studies to identify such a procedure and implement the built procedure to real datasets from the Wellcome Trust case-control Consortium. Our procedure includes four steps: data management, preliminary analysis, association testing and result visualization. In order to get the true association between disease and SNP, we execute 2 preliminary processes, the quality control (QC) and population stratification. Furthermore, we can plot the quantile-quantile (Q-Q) plot and Manhattan plot to visualize association analysis results. At the end of the study, we have successfully (1) identified the necessary and important analyses for GWA, (2) identified currently available software for these analyses, (3) performed the analysis on the Wellcome Trust case-control Consortium data, and (4) provided general guidelines for performing GWA.