I have been using GLIMPSE2 to impute my low-coverage samples for quality control (to proceed with subsequent association analysis). Unfortunately, I found that about 80% of the SNP mutation sites have a MAF value of 0. When I set the MAF threshold to 0.05, more than 90% of the SNP sites were eliminated. I would like to ask if this phenomenon is normal? Is it a contamination issue in my sample detection, a normal occurrence after imputation, or is there a mistake in my imputation steps?
Thecover
commented
1 week ago
Hello,
I have been using GLIMPSE2 to impute my low-coverage samples for quality control (to proceed with subsequent association analysis). Unfortunately, I found that about 80% of the SNP mutation sites have a MAF value of 0. When I set the MAF threshold to 0.05, more than 90% of the SNP sites were eliminated. I would like to ask if this phenomenon is normal? Is it a contamination issue in my sample detection, a normal occurrence after imputation, or is there a mistake in my imputation steps?
Best ZW Hu