Home
Please see the main README for information about the use of this module.
In my work with GRASP it became obvious that the database is missing many SNPs. They have no SNPs with a p-value greater than 0.05, but beyond that, many studies only have a few hundred SNPs, and the distribution is not at all what you would expect given a truly random sample of SNPs with a p-value less than 0.05. For example, the best overlap I could find in any phenotype I examined between Europeans and Africans was for BMI, and that looked like this:
Not very convincing at all. There should be a greater cluster around 0, and more points on the line; not to mention thousands more points.