I'm currently trying to do some analysis using the HLA region supplied by the UK Biobank but I have some questions about the output.
From what I understand in this quick example
the way to read this is that:
Individual 1 doesn't have allele A_101 on either chromosome but has two copies of allele A_102.
Individual 2 has one copy of A_101 on one chromosome and potentially has one copy of A_102 on one chromosome.
Individual 3 has two copies of A_101 and either one copy of A_102 on one chromosome and potentially a copy of A_102 on the other or an uncertain probability of having a copy of A_102 on both chromosomes.
Is this a correct way to read this and, if so, do you know why the UKBB say that someone should remove any alleles with probability less than 0.7 and then perform rounding to 0, 1 ,2 for the remaining alleles? Surely all 0 records have already been removed?