De Bakker Paul I W, Graham Robert R, Altshuler David, Henderson Brian E, Haiman Christopher A
Broad Institute of Harvard and Massachusetts Institute of Technology, Cambridge, MA, USA.
Pac Symp Biocomput. 2006:478-86.
Genetic association studies can be made more cost-effective by exploiting linkage disequilibrium patterns between nearby single-nucleotide polymorphisms (SNPs). The International HapMap Project now offers a dense SNP map across the human genome in four population samples. One question is how well tag SNPs chosen from a resource like HapMap can capture common variation in independent disease samples. To address the issue of tag SNP transferability, we genotyped 2,783 SNPs across 61 genes (with a total span of 6 Mb) involved in DNA repair in 466 individuals from multiple populations. We picked tag SNPs in samples with European ancestry from the Centre d'Etude du Polymorphisme Humain, and evaluated coverage of common variation in the other samples. Our comparative analysis shows that common variation in non-African samples can be captured robustly with only marginal loss in terms of the maximum r2. We also evaluated the transferability of specified multi-marker haplotypes as predictors for untyped SNPs, and demonstrate that they provide equivalent coverage compared to single-marker tests (pairwise tags) while requiring fewer SNPs for genotyping. The efficacy of a tagging-based approach in studying genotype-phenotype correlations in complex traits is strongly supported by our empirical results.
通过利用附近单核苷酸多态性（SNP）之间的连锁不平衡模式，基因关联研究可以变得更具成本效益。国际人类基因组单体型图计划（International HapMap Project）目前在四个人类群体样本中提供了一张覆盖全人类基因组的高密度SNP图谱。一个问题是，从人类基因组单体型图这样的资源中选择的标签SNP在独立疾病样本中能多好地捕获常见变异。为了解决标签SNP的可转移性问题，我们对来自多个群体的466名个体中涉及DNA修复的61个基因（总跨度为6 Mb）的2783个SNP进行了基因分型。我们从人类多态性研究中心（Centre d'Etude du Polymorphisme Humain）的欧洲血统样本中挑选标签SNP，并评估其他样本中常见变异的覆盖率。我们的比较分析表明，非非洲样本中的常见变异可以被可靠地捕获，在最大r2方面仅有微小损失。我们还评估了指定的多标记单倍型作为未分型SNP预测指标的可转移性，并证明它们与单标记测试（成对标签）相比提供了同等的覆盖率，同时需要进行基因分型的SNP数量更少。我们的实证结果有力地支持了基于标签的方法在研究复杂性状的基因型-表型相关性方面的有效性。
