Strongly varying percentages between 23andme and AncestryDNA

**~~Harald Meller~~** · 08-13-2020, 09:28 PM

Originally Posted by Götterfunke

It's questionable how much ancient ancestry matters though as 23andme only test ancestry of the recent 200 years or so-and their control groups are considerably younger.

Really? I don't believe that. Explain why Mexicans do not score 100% Mexico in 23andMe, but rather 50% Spain + 50% Native American.

After all, in the last 200 years they are 100% from Mexico, so they should not score like a mix of two things?

**Götterfunke** · 08-13-2020, 09:32 PM

Originally Posted by Harald Meller

Really? I don't believe that. Explain why Mexicans do not score 100% Mexico in 23andMe, but rather 50% Spain + 50% Native American.

After all, in the last 200 years they are 100% from Mexico, so they should not score like a mix of two things?

Nevermind, the 200 years thing is false, this is what they say about their control groups:

"Customers comprise the lion's share of the reference datasets used by Ancestry Composition. When a 23andMe research participant tells us that they have four grandparents all born in the same country — and the country isn't a colonial nation like the US, Canada, or Australia — that person becomes a candidate for inclusion in the reference data. We filter out all but one of any set of closely related people, since including closely related relatives can distort the results. And we remove outliers, people whose genetic ancestry doesn't seem to match up with their survey answers. To ensure a representative dataset, we filter aggressively — nearly ten percent of reference dataset candidates don't make the cut.

We also draw from public reference datasets, including the Human Genome Diversity Project, HapMap, and the 1000 Genomes project. Finally, we incorporate data from 23andMe-sponsored projects, typically collaborations with academic researchers. We perform the same filtering on these public and collaboration reference data that we do on the 23andMe customer data."

So theoretically, I could have been part of the control group before I took the test, as I considered all my four grandparents to be German (and they are, largely at least).