GWASpower calculation
1
0
Entering edit mode
3.3 years ago
putty ▴ 40

Dear All,

Sorry , sounds like a stupid question but been struggling with this for a while. I am trying to do post-hoc power calculation for my GWAS using GWASpower software for quantitative trait (i know many don't prefer post-hoc power calculation, but reviewer asked me despite clearly mentioning it in the limitations section). Can anyone please clarify what we are supposed to enter in the effect size (heritability) box (image attached) because GWAS plink output gives only Beta, SE, CI, p values ?

BTW I do know what an "effect size" and " heritability" mean. My confusion is whether I need to calculate "effect size/heritability" from my OWN study, or can i just use an estimate of effect size/heritabilty from previous studies ? Also i am not sure whether we can estimate heritability from GWAS, i thought heritability came from family/twin studies.

Hope i have made myself clear .

Thank you allenter image description here

GWASpower post-hoc • 1.4k views
ADD COMMENT
1
Entering edit mode

Not familiar with the GWAPower calculation software. But you can definitely estimate Heritability using your GWAS data (e.g. using LDSC or GCTA)

ADD REPLY
2
Entering edit mode
3.3 years ago
Lemire ▴ 940

That paper is a hot mess with their use of "heritability". In practice, you are dealing with a SNP that will explain a proportion of the variance of the trait (that proportion is the R2 estimate you should get in your .assoc file). That's your effect size. Enter that value in the form.

Alternatively you could use the genetic power calculator to confirm.

http://zzz.bwh.harvard.edu/gpc/qtlassoc.html

Enter your R2 in the "Total QTL variance" box, select singletons, enter your SNP allele frequency (use the same freq for M1 and D-prime of 1). The results won't be affected by the sibling correlation since you are using singletons, so enter something like 0.5 (but check that nothing is negative in the output).

ADD COMMENT
0
Entering edit mode

Thank you, that was really helpful. I eventually ended up using gpc for my power/sample size calculation. Seems the GWApower tool link is broken (can't access it ).

ADD REPLY

Login before adding your answer.

Traffic: 2233 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6