Biostatistics and Bioinformatics Unit, Cardiff University, Wales College of Medicine, Heath Park, Cardiff CF14 4XN, UK.
BMC Genet. 2005 Dec 30;6 Suppl 1(Suppl 1):S45. doi: 10.1186/1471-2156-6-S1-S45.
We evaluate a method for the incorporation of covariates into linkage analysis using the Genetic Analysis Workshop 14 simulated data. Focusing on a randomly chosen replicate (42) we investigated the effect of the 12 subclinical phenotypes, sex, population, and parent-of-origin on the linkage signal from a model-free linkage analysis of Kofendrerd Personality Disorder.
We detected a linkage peak on chromosome 1, at about 175 cM, which varied depending upon individuals' status for subclinical phenotype b. A linkage peak on chromosome 3 (310 cM) was found not to depend upon subclinical phenotype status. Further peaks were found on chromosomes 5 (12 cM), 9 (4 cM), and 10 (95 cM), depending on the status of subclinical phenotypes a, k, and c/d/g, respectively.
Retrospective comparison of our results with the simulation model showed correct identification of disease loci D1-5 on chromosomes 1, 3, 5, 9 and 10, respectively.
我们评估了一种将协变量纳入连锁分析的方法，使用了遗传分析研讨会 14 号模拟数据。我们关注一个随机选择的副本（42），研究了 12 种亚临床表型、性别、人群和母源效应在无模型连锁分析中对科芬德个性障碍连锁信号的影响。
我们在 1 号染色体上发现了一个大约 175 cM 的连锁峰，其取决于个体 b 亚临床表型的状态。在 3 号染色体上（310 cM）发现的连锁峰不依赖于亚临床表型状态。进一步的峰出现在 5 号染色体（12 cM）、9 号染色体（4 cM）和 10 号染色体（95 cM）上，分别取决于亚临床表型 a、k 和 c/d/g 的状态。
我们的结果与模拟模型的回顾性比较表明，正确识别了染色体 1、3、5、9 和 10 上的疾病位点 D1-5。
