Posting this as a new topic. I've been trying to run an association test in plink, but it seems the rate of missingness in the merge vcf is high, meaning all copies of the second allele have been replaced with missing genotypes. Therefore, all of my variants have a MAF of 0.
This is the command I've used to create a merged vcf:
bcftools merge --force-samples *vcf.gz -o all_isolates.vcf
Which I then convert to a bed using --make-bed in plink.
Any ideas as to why I'm getting such a high level of missingness in the merged vcf though?