statgen / raremetal

A flexible tool for meta-analysis
8 stars 9 forks source link

different SKAT results #41

Open smhaider opened 1 year ago

smhaider commented 1 year ago

Hello,

I have a confusion. I have done a sex combined and sex-specific gene burden test using EPACTS (SKAT) and RAREMETAL(SKAT). I got different results such as.

with Epact's SKAT test: the PLIN5 gene is significant in the sex combined burden test with a p-value of 10-7 and not significant in the sex-specific analysis.

while with the Raremetal SKAT test: the PLIN5 gene is not even close to being significant in the sex combined analysis, but is significant in females with a p-value of 10-8 in the sex-specific analysis.

So I am confused as to why two of the same tests give different results and wonder which result I should consider more reliable.

p.s. The parameters were the same, the number of genes in the groups were same for both tools.

looking forward to hearing from you and thanks in advance.

Yours sincerely, Haider