Hello, I have question about MAF filtering in GWAS.

As I know of, MAF-based variants filtering is a crucial step for GWAS since it filters out potential false-positive derived from sequencing and genotyping error. (in addition, low MAF would have low statistical power)

However, now I work on rare disease that may caused by rare variants, so that I want to remain those rare(low MAF) variants for downstream analysis such as PRS and burden test.

I wonder if it is the appropriate way because it means I'm taking a risk of having potential false-positives.

I read some research tried remaining rare variants and found way to emphasize it(such as burden test) but still I'm not confident about remaining rare variants.

I want to hear your advice and opinions. Thank you.