23andMe Expands its Ancestry Composition Reports

23andMe announced today that they will soon be adding 120 new geographic regions to their Ancestry Composition report (sometimes called an ethnicity estimate). That’s up from their current number of 31 populations. That’s quite an increase!

The Ancestry Composition in the past was estimated by comparing bits of your DNA (called “markers”) with those of individuals with long historical ties to one of those 31 reference populations. Computer algorithms searched those reference data sets for suites of markers that were characteristic of, say, Southeast Asia, and if you also had some of those markers, you might be assigned a percentage from that region. The Ancestry Composition estimates reflected your genetic lineages back roughly 500 years or so.

This update adds another layer of detail onto the existing Ancestry Composition: “recent ancestor locations”. These locations are determined based on exact matches between your DNA and that of 130,000 individuals with known family histories from those 120 geographic regions. If you have an exact match with five or more people from a given region, you will be assigned to that location, with a confidence level based in part on how much DNA you share with the reference individuals. (This sounds a lot like the Genetic Communities that AncestryDNA introduced last year.)

Here’s a sample report from the announcement on 23andMe‘s website. Note that some of the existing 31 populations have been renamed. In the image below, “Manchurian & Mongolian” seems to be the new name for the region previously called just “Mongolian”.

In the sample report, the “recent ancestor locations” look to be listed directly below the broader 31 populations. That is, “Spain” is the recent ancestor location for the Iberian region and “Peru” for the Native American population.

The updated reports will be available to new customers when they receive their results, and existing customers are being phased into the new system.  All 23andMe users should see the new Ancestry Composition update within a few months.

Do you have the new reports yet?  If so, please share your thoughts in the comments.

12 thoughts on “23andMe Expands its Ancestry Composition Reports”

  1. I rather expect that since this improvement was first rolled-out to v5 testees, that 23andMe got access to a bunch of new reference populations from Illumina as part of Illumina’s roll-out of their GSA chip … also known as v5.

    Illumina would have to release substantial documentation of the applicability of their GSA chip to all world populations, populations not covered by HGDP or any other longstanding sample library.

  2. This is for **new** chips only. Do some basic research instead of copying & pasting 23&me information. You obviously don’t comprehend it.

    1. If you log into a 23andMe account and go to the Home page, you will see this announcement:
      “Coming soon!
      23andMe’s Ancestry Composition update with 120 additional regions will be coming soon to all customers.” (emphasis mine)

      1. Agreed, it is being updated for older chips. I have a v4 test, and I got an email from 23andme informing me they would be re-doing my ancestry composition sometime in the next few months. You don’t need to have a v5 test for the updated ancestry composition. Period.

        Now that said, they may be doing the v5 folks first, and have said any kit after 2/28/2018 will get the new results directly. (ie: they will never get the old 31 population report).

        There are numerous places 23andme has said this will apply to “all customers” and they are “gradually rolling out to existing customers”. You can see it on their blog for starters.

        Why be so insistent that it is not? Do you have anywhere that 23andme are on the record saying it will only be new customers?

        1. For what it is worth, my v4 test has been updated to the new populations. That said, it’s really more like sub-populations of the existing ones.

          If you look on your ancestry report after the update, under each population it may list a country that seems to best represent your DNA in that region. So under “France and Germany x%” it may say “Germany” or “Austria”.

          If you go into “show all tested populations” it will show all the sub-populations and the “match strength” of each. For example under “French and German” it will list Austria, Belgium, France, Germany, Luxembourg, Netherlands, and Switzerland, with a 0-to-5 bubbles indicator of “match strength”, but no percentages.

          From what I understand, this is exactly what they do for v5 tests, although the calls may be more (or less) accurate when generated with non-v5 data. For those of European heritage, it is probably about the same, maybe some minor tweaks, but v5 should have better coverage on non-Europeans due to the GSA chip being a bit more “global” in focus.

        2. Thanks for this detailed description! Mine has updated as well. What’s interesting is that under French & German, they only list France, although my match strength is the same for both France and Germany (one dot each). I have similar amounts of ancestry from both countries. I also got “Ireland” (two dots), which agrees with what I know of my family tree.

  3. Version 2 Here and yes it’s been updated. Much like the Ancestry Communities it doesn’t tell me much I didn’t already know but I will say unlike Ancestry Communities it does pick up my Netherlands DNA instead of just my British and Irish.

    It also is closer than FTDNA’s Ancestry and MyHeritage when dealing with more current time frames.

  4. Wow. the science……
    Manchurian and Mongolian are different. There are some history that people confused. You can’t post like <>. It may make some people irritated. The piece of wrong history taught.

    1. The fact that 23andMe groups Manchurian & Mongolian together as a category does not mean that the two are the same; it just means that they don’t have enough data to statistically separate them at this time.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.