I am writing to find out what the most common or most widely accepted algorithm for calculating Meff, or the number of linkage independent effective SNPs.

I am authoring a research article on a study conducted via the immunochip, and have found SEVERAL articles that use widely varying figures for Meff, everything from LD independent ~67000 to ~14000 SNPs. Needless to say, since these numbers differ by a factor of 4 or so, and since that directly affects the Bonferonni adjusted threshold, making the wisest choice here is important.

To help me make that choice, I was hoping to ask 3 questions:

1. Is there a most common way of doing this for Ichip studies?
2. Many times, expert opinion converges on a solution, but that convergence is not reflected in the literature for several years. At this point, is it already clear which solution is the most accurate for estimation of Meff?
3. Pursuant to answering 1 and/or 2., are there any good review articles that could help me understand the assumptions on which these rely?