Reich D E, Goldstein D B
Whitehead Institute/MIT Center for Genome Research, Cambridge, Massachusetts 02142, USA.
Genet Epidemiol. 2001 Jan;20(1):4-16. doi: 10.1002/1098-2272(200101)20:1<4::AID-GEPI2>3.0.CO;2-T.
Case-control studies are subject to the problem of population stratification, which can occur in ethnically mixed populations and can lead to significant associations being detected at loci that have nothing to do with disease. Here, we describe a way to measure and correct for stratification by genotyping a moderate number of unlinked genetic markers in the same set of cases and controls in which a candidate association was found. The average of association statistics across the markers directly measures stratification. By dividing the candidate association statistic by this average, a P-value can be obtained that corrects for stratification.
病例对照研究存在群体分层问题，这种情况可能发生在种族混合的群体中，并可能导致在与疾病无关的基因座上检测到显著关联。在此，我们描述了一种方法，通过在发现候选关联的同一组病例和对照中对适度数量的不连锁遗传标记进行基因分型来测量和校正分层。跨标记的关联统计平均值直接衡量分层情况。通过将候选关联统计量除以该平均值，可以获得校正分层后的P值。
