Effect Allele Frequency in base summary statistics for PRSs calculation

Question

[bigfoot]

I want to use summary statistics containing only the Effect Allele Frequency (EAF) and not the Minor Allele Frequency (MAF) as base dataset for PRS calculation in PRSice2. It would be possible to convert EAF to MAF and then filtering for a MAF threshold of 0.01?

Thanks in advance

Answer 1

[Sam]

If you know the effective allele information, then you can simply filter the effective allele frequency as those > 0.99 and < 0.01. Because if the effective allele frequency is less than 0.01, it's the minor allele.

[gradstudentNew]

Do you mean 0.49 instead of 0.09?

[Sam]

Opps, sorry, should be 0.99. Have now edited

[bigfoot]

Thanks Sam. So, in the end, only SNPs with EAF in the range 0.01-0.99 should be retained for the analyses.. right?

[Sam]

If your goal is to filter by MAF of 0.01, then yes.

[rem]

As a follow-up, what if our summary statistics only contain the alternative allele frequency and we would like to filter on a MAF threshold? Could we go through the same process of do we first need to transform the alternative allele frequency somehow?

[Sam]

You can change the alternative allele frequency to MAF by making sure the frequency is less than 0.5. So for any frequency > 0.5, do 1- frequency and you'll have the MAF