Individual Hardy-Weinberg Proportion Testing
John J. Chen
Division of Biostatistics
University of California at Berkeley
Berkeley, CA 94720-3140
It is often of interest to detect the deviation from Hardy-Weinberg
Proportion (HWP) for a particular heterozygote.
Hernandez and Weir (1989) suggested a disequilibrium coefficient
approach and proposed a one degree of freedom chi-square test. We derive
the appropriate variance for performing this individual HWP test.
In practice, however, the number of genotypes in a sample
can be quite large relative to the sample size. Therefore, the validation
of this asymptotic test is in doubt. We have proposed a Monte Carlo
method for performing such tests. Powers of the individual tests are
compared for the large-sample approach and the Monte Carlo approach,
and for different test statistics. Results of applying the individual
test to several human genetic data are also presented.