filter 1000 genome vcf file to keep common snps

Question

[shersky]

I want to filter "1000 genome vcf file" to keep only common SNPs. Can someone explain the difference between

1) Filtering by allele frequency AF greater than 0.05 (which I found yields more variants)

bcftools view -i 'AF>0.05'  variants.vcf -o common_variants.vcf

2) Filtering by MAF greater than 0.05

bcftools view -i 'MAF>0.05' variants.vcf -o common_variants.vcf

Which one is correct?

Answer 1

[bk11]

AF is the frequency for an alternate allele. If AF < 0.5, then AF is equal to MAF. Rare variants generally has AF or MAF < 0.05. MAF (Minor allele frequency) refers to the minor allele (least frequent) frequency. Common SNPs have MAF > 0.05. Your filtering method 2 is the correct one to filter common SNPs.

bcftools view -i 'MAF>0.05' variants.vcf -o common_variants.vcf