Department of Statistics, The Texas A&M University, College Station, TX 77843-3143, USA.
Eur J Hum Genet. 2010 Oct;18(10):1148-59. doi: 10.1038/ejhg.2010.60. Epub 2010 May 12.
One of the surest signatures of recent positive selection is a local elevation of advantageous allele frequency and linkage disequilibrium (LD). We proposed to detect such hitchhiking effects by using extended stretches of homozygosity as a surrogate indicator of recent positive selection. An extended haplotype-based homozygosity score test (EHHST) was developed to detect excess homozygosity. The EHHST conditioned on existing LD and it tested the haplotype version of the Hardy-Weinberg equilibrium. Compared with existing popular tests, which usually lack clear distribution, the EHHST is asymptotically normal, which makes analysis and applications easier. In particular, the EHHST facilitates the computation of an asymptotic P-value instead of an empirical P-value, using simulations. We evaluated by simulation that the EHHST led to appropriate false-positive rates, and it had higher or similar power as the existing popular methods. The method was applied to HapMap Phase II data. We were able to replicate previous findings of strong positive selection in 17 autosome genomic regions out of 20 reported candidates. On the basis of high EHHST values and population differentiations, we identified 15 new candidate regions that could undergo recent selection.
近期正选择的一个可靠标志是有利等位基因频率和连锁不平衡(LD)的局部升高。我们建议通过使用纯合性扩展片段作为近期正选择的替代指标来检测这种 hitchhiking 效应。提出了一种基于扩展单倍型的纯合性评分检验(EHHST)来检测过度纯合性。EHHST 受现有 LD 条件的限制,并检验了 Hardy-Weinberg 平衡的单倍型版本。与通常缺乏明确分布的现有流行检验方法相比,EHHST 渐近正态,这使得分析和应用更加容易。特别是,EHHST 可以使用模拟计算渐近 P 值,而不是经验 P 值。我们通过模拟评估发现,EHHST 导致了适当的假阳性率,并且它具有与现有流行方法相同或更高的功效。该方法应用于 HapMap Phase II 数据。我们能够复制之前在 20 个报告候选者中的 17 个常染色体基因组区域中发现的强烈正选择的结果。基于高 EHHST 值和种群分化,我们确定了 15 个新的候选区域,它们可能经历了近期的选择。