Requesting Clarification on GWAS Power Calculation from lander et al. 2008

Question

[aakashi]

Hi,

I was wondering how the attached figure is constructed.

The figure originates from this 2008 paper: https://pubmed.ncbi.nlm.nih.gov/18988837/

Unfortunately, there is no reference to how it was constructed. Normally when calculating GAWS power, the prevalence of the disease affects the power. For example Online GWAS Power Calculator

In this paper, they are not looking at a single disease but all diseases, and have no reference to prevalence. How is it possible to create a graph in this manner? Is there an obvious rule of thumb or any assumptions that are made that I am missing? Is there a separate power test other than the one proposed in Skol et al. (2006)?

Thanks!

Answer 1

[Kevin Blighe]

Your questions are fundamentally related to statistics; so, it would be best served on the CrossValidated forum.

The grey borders around the figures tell me that they used STATA.

If you want to mess around with these calculations yourself, then try out CaTS Power Calculator.

How is it possible to create a graph in this manner?

With a lot of assumptions going into the calculations - these are generic Power calculations that don't fit all situations.

Kevin