Question: Hardy-Weinberg Equilbrium for multiple loci
3.7 years ago by

I would like to test to see if a list of variants are in HWE

Say I have 100 loci with biallelic (AA/Aa/aa) genotypes for 1,000 samples. I would like to test that on average are the alleles in HWE.

Do I calculate a chi-squared statistic across each allele and then take the mean/median Chi-Sq stat to find the Pvalue?

I am assuming that each region is independent of each other (no linkage)

hwe • 1.4k views
ADD COMMENTlink modified 3.7 years ago by WouterDeCoster44k • written 3.7 years ago by QVINTVS_FABIVS_MAXIMVS2.4k
gravatar for anp375
3.7 years ago by
anp375170 wrote:

This sort of looks like a Multiple Testing problem. Get the chi-square/something-else p-values for all locations. With an alpha-threshold of 0.05, you can expect to get 5 p-values below 0.05 by chance.

The rest of this is pure speculation and I have no idea how correct it is:

Use a binomial distribution function to calculate the probability of obtaining your number of significant p-values (k) out of the total number of p-values (n=100), given an expected proportion of significant p-values of 0.05 (p=5/100).

If your probability is less than 0.05, then on average, your set of loci is not in Hardy-Weinberg equilibrium.

This is almost definitely not the most powerful test for your scenario, so you probably won't get a significant result.

ADD COMMENTlink written 3.7 years ago by anp375170
gravatar for WouterDeCoster
3.7 years ago by
WouterDeCoster44k wrote:

I seemed to have missed this question 11 days ago. I would definitely recommend using plink for jobs like this, and similar. The tool is intuitive and the documentation is amazing, see hardy-weinberg equilibrium test in plink

ADD COMMENTlink written 3.7 years ago by WouterDeCoster44k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 781 users visited in the last hour