中文摘要：

数量基因位点（quantitative trait loci,QTL）分析是利用图位法克隆控制数量性状主效基因的前提和基础。采用BoxCox公式进行数据正态转换对提高QTL分析效率有显著作用。笔者以杨树为实例显示了BoxCox公式在数据正态转变中的应用及数量性状表型分布对QTL分析的影响。结果发现,数据是否符合正态分布对发现控制数量性状的基因位点有显著影响。如果不对性状的表型值进行分布检测并进行正态转变,可能无法发现某些有显著效应的遗传位点。通过比对转化前后QTL分析的LOD值变化曲线得知,曲线变化的趋势在转化前后是一致的,曲线上各个小峰的位置也是稳定的,但转化后LOD峰值显著提高。如第4染色体上,转化前后LOD值变化曲线上在40～60、80～100及130～150cm区间均有3个独立峰值出现,数据转化后对应峰值升高,其中第1个峰的峰值约增加3倍,数据转化后该峰值LOD支持度达到了极显著水平,显示该位置存在一个控制该性状的较强的遗传位点。第8染色体上也出现相似的情况。

英文摘要：

Quantitative trait loci （QTL） analysis is the precondition and basis for positional cloning of major genes underlying quantitative traits. Normality transformation of quantitative trait data using BoxCox formula has significant effect on QTL detection. In this paper,a case study on poplar is carried out to demonstrate the application of BoxCox formula for normality transformation and the effect of data distribution on QTL analysis. The results are as followings： QTL detection is significantly affected by the data distribution. Without normality transformation,some significant QTLs may be missed. Similar variation trends are observed on the heat plots of LOD scores and LOD peaks are found to appear in the same positions in the charts established by LOD scores derived before and after the normality transformation. On chromosome 4,3 LOD peaks appear at 40—60 cm,80—100 cm and 130—150 cm intervals respectively both before and after transformation. However,the corresponding LOD peaks are found to increase to the significant level after transformation. Especially, the increment of the first peak is about 3 folds higher than that before BoxCox transformation,which means a relatively strong QTL in the corresponding position. The same scenario is also observed on chromosome 8 in this study.