Hi,
I am interested in using Coloc which requires MAF as input. However, my GWAS summary statistics only have p values, OR and N (including N of cases and controls). I wonder if it would be possible to calculate MAF, and how to do so?
Thank you very much!