Center for Research on Genomics and Global Health, National Human Genome Research Institute, Bethesda, Maryland 20892-5635, USA.
Genet Epidemiol. 2011 Nov;35(7):632-7. doi: 10.1002/gepi.20612. Epub 2011 Sep 15.
Testing for Hardy-Weinberg equilibrium (HWE) is commonly used as a quality control filter in genome-wide scans for markers with experimentally determined genotypes. In contrast, for markers with imputed genotypes, there are post-imputation metrics of quality that can be used as screens but there are no formal tests of deviation from HWE. Similarly, there are no formal tests of deviation from HWE for probabilistic genotypes that are generated by sequencing projects. Here, I describe generalizations of the approximate χ(2) and exact tests of HWE for use with uncertain genotypes. The tests fully account for the probabilities of all possible genotypes at a marker for each individual. By computer simulation, the approximate and exact tests are shown to maintain valid control of the type I error rate. Calculations of the loss of power as the uncertainty in genotypes increases are illustrated. The tests are compatible with chip-based genotypes for single-nucleotide polymorphisms and copy number polymorphisms, imputed genotypes, and probabilistic assignments of genotype from variable-coverage sequence data.
检测 Hardy-Weinberg 平衡（HWE）通常被用作全基因组扫描中用于具有实验确定基因型的标记的质量控制筛选。相比之下，对于具有推测基因型的标记，可以使用推断后质量的度量标准作为筛选，但没有偏离 HWE 的正式检验。同样，对于通过测序项目生成的概率基因型，也没有偏离 HWE 的正式检验。在这里，我描述了用于不确定基因型的 HWE 近似 χ(2)和精确检验的推广。这些检验充分考虑了每个个体在标记处所有可能基因型的概率。通过计算机模拟，证明近似和精确检验可以有效地控制第一类错误率。说明了随着基因型不确定性的增加，效力损失的计算。这些检验与基于芯片的单核苷酸多态性和拷贝数多态性的基因型、推测的基因型以及来自可变覆盖序列数据的基因型概率分配兼容。
