What is the typical rate of sex mismatches found during genetic QC

What is the typical rate of sex mismatches found during genetic QC? I’d like to know what is considered normal and what might be unusually high.

1 or 2 people per 1000

mostly it is data entry error

Or transgender participants

There was a paper about that recently with UK Biobank data, a lot of the “sex discordant” samples had other data to support that the participants may have been transgender or had intersex traits: https://www.pnas.org/doi/abs/10.1073/pnas.2218700120

So in some cases it may not be desirable to exclude them.