gnomAD v4 - bottleneck populations?

General
1

Hi All,

Was wondering if anyone has any insight on what populations are excluded from the FAF in gnomAD v4 due to being bottleneck populations like AJ/Other (Now Remaing)/FIN. Do we know if the Middle Eastern is considered a bottleneck population?
Thanks!

2

Hi @Katiemay8199,

Thank you for you inquiry! This led us to discover we erroneously left the Middle Eastern genetic ancestry group out of our FAF and grpmax calculations in v4.0.

In previous releases, we removed the Middle Eastern genetic ancestry group from FAF as we did not have enough samples to consider them. However, we now have enough samples in v4 and we will be adding them to FAF and grpmax calculations in our v4.1 release. The genetic ancestry groups that will be excluded in v4.1 FAF and grpmax calculations are “asj”, “fin”, “ami”(genomes only), and “remaining”.

Thanks again!
Mike

2 Likes
3

Oh wow! Good to know! Thank you so much! Look forward to seeing this in your v4.1 release!

Thank you for replying and the update. Much appreciated!

-Katie

4

I had the same question after seeing Middle East pop show up in GrpMax for some variants in V4.1. I was confused because I read in the popup explanation for GrpMax that Middle Eastern pop was not part of GrpMax calculation. Just a reminder to change the popup to reflect that MID is not considered a bottleneck pop w/r/to GrpMax. thanks

5

Thank you for reporting this @JJM! We missed this and will update the popup text.

6

Hello,

This did answer one of my questions, but I did have a couple of other questions related to the filtering allele frequency.

  1. I thought that the filtering allele frequency would only use a group if they had a minimum of 5 variant alleles (and a total of 2000 alleles, which I think all groups now have). Is a minimum of 5 alleles no longer necessary?
  2. There are times when the group in the exomes is different from the genomes is different from the total. Shouldn’t they all be consistent?

Thanks,
NM