合併染病同胞對資料之標識基因下傳-不下傳訊息與同源全等基因訊息的連鎖分析方法(II)：一個最小合併變異數方法

Abstract

傳遞不平衡檢定(transmission/disequilibrium test,簡稱TDT)，與均值檢定(mean test)是目前兩種主要利用有染病子代的核心家庭資料於連鎖分析(linkage analysis)的無母數方法。傳遞不平衡檢定方法檢定標識對偶基因(marker allele)與疾病對偶基因(disease allele)一起下傳或不下傳至子代的機率是否相等，均值檢定是比較染病同胞對(affected sib pair)標識對偶基因的同源全等(identical by descent,簡稱 IBD)數目之觀察值與沒有連鎖時之期望值是否相等。由於連鎖不平衡(linkage disequilibrium)程度弱時，均值檢定的檢定力比傳遞不平衡檢定高；連鎖不平衡強時，傳遞不平衡檢定的檢定力比均值檢定高，所以發展合併此兩種檢定統計量的方法是值得探討的問題。本研究藉由合併傳遞不平衡訊息及同源全等數目的訊息發展出一個合併二者訊息的新的檢定連鎖方法。合併統計量的方法是經由使合併統計量變異數最小，產生合併二不同訊息之權數得到統計量。本研究利用模擬的方式比較傳遞不平衡檢定、均值檢定與這裡所提方法在不同的遺傳模式下檢定力的表現。

The transmission/disequilibrium test (TDT) and mean test are the two nonparametric methods for test of linkage for affected sib pairs data. The TDT tests the equality of the probability of a marker allele transmitted or not transmitted with the disease gene, and the mean test tests whether the observed number of marker identical by descent in affected sibs is equal to the expected. The mean test and TDT have different performances in power. The power of mean test is higher than that of TDT when the linkage disequilibrium is weak, but, in contrast, the power of TDT is higher than that of mean test when linkage disequilibrium is strong. In this paper we proposed a minimum variance combining method to combine the two sources of information from the TDT aspect and the mean test aspect. We derived the combining weights for the two test statistics. Simulation studies of the powers of TDT, mean test and the new statistic in various genetic models were conducted.