Comparison of Hardy-Weinberg Equilibrium Methods for Survey Data

• Activity Number: 134 - Recent Development in Methods for Statistical Genetics
• Type: Contributed
• Date/Time: Monday, July 30, 2018 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): John Pleis* and Donald Malec and Rong Wei and Vladislav Beresovsky and Bill Cai and Te-Ching Chen and Yulei He and Peter Meyer and Lyna Schieber and Hee-Choon Shin and Ajay Yesupriya and Guangyu Zhang
• Companies: National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics and CDC/NCHS and CDC/NCHS and CDC/NCHS and National Center for Health Statistics and Centers for Disease Control and Prevention and CDC/NCHS and State University of New York at Albany and National Center for Health Statistics
• Keywords: Genetics; Survey; NHANES; NCHS

Abstract

In population-based genetic association studies, testing for departure from Hardy-Weinberg Equilibrium (HWE) is commonly performed as an initial evaluation of the data. Random samples of the population, such as seen in household surveys, can avoid selection bias seen in observational studies. Several methods for HWE for household survey data have been proposed, which consider the clustering, stratification, and unequal probabilities of selection of such sample designs. Some of these methods include those proposed by Moonesinghe et al. (2010), Li and Graubard (2009), Li (2013), She et al. (2009), Li et al. (2011), and Wang et al. (2016). For this paper we review these various methods using genetic data from the Third National Health and Nutrition Examination Survey (NHANES III). Of particular interest is the impact that extreme values of the sampling weights can have on the various HWE methods.