GWAS dataset calculating MAF for European dataset

Question

[v.johnson]

How does frequency of existing variant in the combined 1000 genome European data from Variant effector predicator website equate to MAF. I have a SNP data from GWAS data and need to find the MAF for each of the SNPs. Thank you

Answer 1

[Kevin Blighe]

Hi Victoria, can you please clarify the format of your data, currently? Per any given variant, MAF can be applied specifically to any population, i.e., MAF in the global population, MAF in the Irish population, MAF among sub-Saharan Arficans, et cetera.

[v.johnson]

Dear Kevin, I have a European dataset with SNP data and need to get the MAF for each SNP. I have a lot of data from a GWAM study. Headings are; CHR SNP BP ref_allele alt_allele sample_size case_proportion Thank you

[Kevin Blighe]

Hi, in that case, you already have enough information to calculate the MAF in your own cohort of patients / samples. However, if you want the MAF for your variants from the large public databases, then you could use Ensembl VEP or ANNOVAR. Both of these will take some time to set up, though. There are other solutions, too.

[v.johnson]

As in;

If (AF < 0.5) then MAF = AF

if (AF > 0.5) then MAF = 1.0 - AF

[Kevin Blighe]

More or less, yes! That is, this is your study-specific MAF.

[v.johnson]

Thank you for your help!

[Kevin Blighe]

If you find that the AF is > 0.5, though, then you also have to remember to flip the order of the REF and ALT alleles. These situations are quite common and may arise due to this problem: A: Alternate nucleotide is more frequent than reference nucleotide. OMG I'm dizzy.

I would also verify that some of the AFs in your VCF are actually valid by literally counting across all samples... got to be sure in bioinformatics!