Dear hive mind,
I was wondering whether you could help me overcome some confusion re the Admixed American population across different gnomAD versions.
I was looking at the most common (European founder) variant for MCAD deficiency that accounts for approximately 2/3 of cases in European and US studies while being virtually absent in Arab or Asian cohorts – so far not surprising.
When looking at the allele frequencies across the different gnomAD builds, the ones in the European subcohort are fairly stable but the ones in the Admixed American subcohort not at all. To test whether this was likely a general big difference due to the more nuanced ancestry within that subcohort inbetween the builds, I looked at a similar case. The CFTR variant F580del which is also European-derived and accounts for about 2/3 of cases shows a different behavior with an, again, comparably stable allele frequency.
The table is attached below. The last column is the calculated disease incidence based on the AFs.
Is there something that I’m missing? Would you have an easy explanation for this observation? Is it just due to the relatively low number of individuals included in v3?
Thanks so much and best,
Florian
