Opened this issue 3 months ago · 3 comments
This will require subsetting into cohorts and applying a model/ statisical test.
also see https://github.com/halbritter-lab/stoneformers-multisample-vcf/issues/4
Relates to #23
Also add counts for cohort sizes (control vs. affected)
I think this should be for now a simple test like Fishers exact probands vs. controls with FDR correction of the values just on the merged variant counts.