Specifying sample size from GWAS summary statistics for meta-analysis

Question

[helen]

The meta-analysis tool I am using (METAL) requires users to specify Weight (sample size) from each GWAS summary statistics file. I have a sumstat file contains columns Nca, Nco, and Neff, but the Neff does not equal to the value of Nca + Nco. I was wondering what the sample size means here? Should I create a new column "N" to combine Nca + Nco? Thank you in advance!

Answer 1

[Sam]

Neff is the effective sample size, you can read more here and here. You can use that as your weight column for METAL if I remember correctly.

[helen]

Thank you, Sam. Then what if I have another sumstat that doesn't have any information regarding the sample size. Would you recommend me to calculate the effective sample size from the raw sample size mentioned in the original paper and transform it using the statistic methods like in the linked articles?

[Sam]

Do you have the SE or the case control sample size? You can calculate the effective sample size from the number of cases and control, or you can use the SE for the meta analysis.