Revised Ethnic Origins at FamilyTreeDNA

Announcement

This morning FamilyTreeDNA, one of the big players in DNA testing, announced that they have revised their ethnic origin percentage estimates:

“myOrigins 2.0 is here!

Discover more of your story with myOrigins!

We’re excited to announce the latest updates to myOrigins, our advanced and popular mapping tool that reveals your ethnic and geographic ancestry.

To experience the latest enhancements, which include new, detailed descriptions of each population cluster, sign in to your account and from your myFTDNA dashboard click on the myOrigins button.

Enhanced Results

This simple tool delivers both new and newly refined reference populations, including smaller trace-percentage results previously not available. These latest additions to our populations and the refinement of existing clusters provide a deeper look into your family history and what makes you…YOU.

New Population Clusters

With the latest reference populations added—including newly refined European, Middle Eastern and Native American clusters—you can now uncover even more clues about your ancestry, confirm family history and gain further insight into your geographic origins.

Everyone has a story…what’s yours?
Discover more of your DNA story today.”

It is expected that each of the DNA testing companies will be revising their origin formulas as more data is available.  It is good to see that FamilyTreeDNA is making these revisions.  It is interesting to see the timing of this announcement, the week after AncestryDNA announced Genetic Communities.  Perhaps FamilyTreeDNA released this origins revision to get their name out and try to be competitive with AncestryDNA.

What changed in the origin estimates?

Earlier I wrote about two individuals who took FamilyTreeDNA tests in DNA Origins: Analyzing Results of two Scandinavians, Norwegian and Swedish.  Let’s compare their new and previous results:

Norwegian

• Scandinavia –         90% vs. previously    55%
• Southeast Europe –  5% vs. previously None
• Finland –                 <2% vs. previously   14% for “Finland and Northern Siberia”
• Central Asia –         <2% vs. previously     2% for “Eastern Middle East”
• British Isles –        None vs. previously    29%

Swedish

• Scandinavia –                          89% vs. previously    87%
• East Europe –                            6% vs. previously None
• Southeast Europe –                   3% vs. previously     1% “Southern Europe”
• Siberia –                                     2% vs. previously     8% for “Finland and Northern Siberia”
• North and Central America – <2% vs. previously None
• Northeast Asia                      None vs. previously     1%
• Central Asia                          None vs. previously     2%

 

The first major difference is that the person with Norwegian background no longer shows British Isles origin.  This is a good correction, since migration was generally the other direction, not this direction.

Next is the difference in Finland contribution.  Previously each had 14% and 8% Finland and Northern Siberia contributions.  This new model gives the Norwegian person a trace amount and the Swedish person no Finland but 2% Siberian contribution.  There is no known Finnish ancestry; perhaps this is a good correction.

Southeast Europe is a new component of the Norwegian’s background, at 5%, and the Swedish person increases slightly from 1% Southern Europe to 3% Southeast Europe.

Eastern Europe is a new component of the Swedish person’s background, now 6%.

The Central Asia component for the Norwegian individual has shifted slightly from the Eastern Middle East region on the map to an area more south-east, more centered on India than Afghanistan, but in the same general area.

The Asian background of the Swedish person is even more interesting.  Instead of Northeast Asia and Central Asia, there is now a trace North and Central America component.  This gives me some pause, since some kind of Asian contribution seems more likely than any actual early American contribution to a Swedish person’s DNA.  Early American DNA is known to be related to Asian DNA.  I would be concerned that the models are misinterpreting Asian DNA and reading it as Native American DNA.

In summary, FamilyTreeDNA has revised their ethnicity origins estimates for its users.  Comparing the results of two individuals before and after the latest change, it appears that Scandinavian and British DNA are being distinguished better, Finnish and Scandinavian DNA are probably being distinguished better, but Asian DNA may be now misreported as Native American DNA.  Of course, this will not and should not be the last time FamilyTreeDNA revises their ethnicity estimates.  Expect more changes, at all of the DNA testing companies.  But if you have tested at FamilyTreeDNA, now is a good time to check out your updated ethnicity estimates.