Ancestry Set to Expand DNA Circles

AncestryDNAIn a press release – and blogger meetup – from today (see “Ancestry Announces 2015 Product and Content Lineup“), AncestryDNA announced plans for the test in 2015, which includes launching the test in Australia and Canada, as well as a new way to identify ancestors using only DNA.

From Today’s Press Release:

Continued Growth for AncestryDNA

  • With a database of over 700,000 genotyped members, AncestryDNA has generated over one billion cousin connections to date. In 2015, we project this database to grow to exceed well over one million genotyped members, resulting in even more and higher quality cousin matches.
  • Following the successful launch of AncestryDNA in the UK, we will soon be bringing the service to our members in Australia and Canada, and in doing so, will connect the major English-speaking migrations and globally connect families like never before.
  • Building on DNA Circles, in 2015 we will launch a new experience that will use the latest genetic technology to discover new ancestors without the customer having to search records or build a family tree. This new feature will transform how family history research is done by providing valuable hints to help experienced genealogist looking to break through brick walls, as well as open family history to a whole new segment of the population. Through this new experience, AncestryDNA customers will be able to discover new ancestors as far back as the 1700’s by connecting into existing DNA Circles.

DNA Circles Without Family Trees

On the last point, in the coming weeks AncestryDNA will launch an extension of the DNA Circles tool in which they assign you to a DNA Circle without having a family tree connection.

Currently, you must have a decent public tree in order to be put into a DNA Circle based on genealogical relationships. Using this new tool, however, you will (potentially) be put into circles without a tree showing that you belong to the circle (in other words, based only on genetic relationships regardless of the trees).

But note! Just because a person is identified does NOT mean they are a direct-line ancestor! Some of these individuals will be collateral matches (a sibling or cousin of a direct ancestor for example), and others will be due to population effects (such as endogamy). Only a certain percentage will actually be direct-line ancestors. It will be impossible to determine – based on the DNA Circle alone – whether an identified person is direct-line, collateral, or population-based. To the point below, additional research will always, always, be necessary.

Using New DNA Circles

It is vitally important to recognize that these new DNA Circles can ONLY be used as hints for further research, NEVER as proof. They are not definitive proof of a relationship. Instead, placement within a circle means that you have some genetic relationship with one or more other people in the circle, all of whom happen to be descended from the identified individual, and that there is some genetic network.

It will be up to each of us to explore that connection in detail – and independently of the DNA – to determine whether there might be a genealogical relationship with the new individual.





















46 Responses

  1. Margaret McCleskey 12 February 2015 / 4:10 am

    Will we need to be retested or will it work with old tests/results.

    • Blaine Bettinger 12 February 2015 / 6:36 am

      Margaret – this works with existing test results, no need to retest.

  2. Samad 12 February 2015 / 4:47 am

    Informative post this one is! One can appreciate seeing all of one’s closest matches gathered into the group (DNA Circles). Indeed, a genealogical DNA test is wonderful invention for the present time. But It cannot pinpoint the place of origin or social affiliation of even one ancestor with exact certainty.

    • Blaine Bettinger 12 February 2015 / 6:38 am

      Samad – I’ve used DNA testing to find and/or confirm several ancestors. Are you perhaps referring to DNA testing alone? If so, I would agree. DNA testing needs to be confirmed with traditional genealogical research.

      • steve musgrave 13 March 2015 / 12:34 pm

        Does anyone know what’s gong on the FTD DNA it impossible to transfer autosomal DNA over there these days.Is there a service available with a DNA test in the UK.

  3. Jason Lee 12 February 2015 / 6:44 am

    This will be fun. Who among us will be able resist the temptation to enthuse over the sudden appearance of a previously unidentified ancestor? Very few of us, I suspect. All caveats will disappear with a poof! I imagine that total loss of objectivity will be nearly inevitable when the new ancestor seemingly fits into cherished hypotheses — or if the ancestor turns out to be someone with a compelling story. I’m sure that confirmation bias in genealogy is as old as genealogy itself, but these new DNA Circles will put confirmation bias on steroids. Especially for the vast majority of Ancestry’s customers whose understanding of genetics is so rudimentary that they’re incapable of using a chromosome browser.

    • Blaine Bettinger 12 February 2015 / 6:53 am

      I agree that it will be dangerous, particularly to the inexperienced genealogist and test-taker. Accordingly, it will be the responsibility of those with more experience to inform newbies of the limitations and what is needed to be done to examine the DNA Circle hint.

  4. caith culbertson 12 February 2015 / 11:12 am

    Yes, but of the 700,000 base, how many of these have trees; and how many have locked and/so private trees?

    • Blaine Bettinger 12 February 2015 / 12:09 pm

      That’s an ongoing problem at every company, I would give anything if more of my matches linked their tree to their DNA!

  5. steve musgrave 12 February 2015 / 4:09 pm

    I have first cousin whose father is not completely known. I am pretty sure he was born in England.Since she is a first cousin she shows most of the same matches as I do but b very little else,e which also leads me to believe her father was from England (82% Brtish 13% irish by Ancestry.,pretty much the same for Family tree.
    Also my grandmother was born in Ulster and her father is not known exactly (her mom was a gorgeous young maid i working in a peers house.(you get the drift) Maybe some mysteries will be solved.
    Thanks Blaine

    • Blaine Bettinger 13 February 2015 / 11:09 am

      Hopefully this new tool will be beneficial for you! Be sure to let us know how it works out.

  6. Linda Grinnell 12 February 2015 / 11:12 pm

    What will happen to the DNA circles we already have in which we are already given the in-common ancestors?

    • Blaine Bettinger 13 February 2015 / 11:16 am

      Nothing will happen to the existing DNA Circles Linda, except that you might have new members.

  7. Jim Yates 13 February 2015 / 2:17 am

    BLAINE, I suspect you have a bit of pull with the testing companies. Most people are irritated to no end by other members that keep their trees private but benefit from seeing your tree. I expect the companies are reluctant to refuse dues and fee paying customers, even if they keep their trees private.
    My idea is to allow open tree owners to pay an extra fee in order to block their trees from the “chronic blockers”. The extra fees could go to a charitable group which attempts to assist adoptees at birth. Let’s apply some heat on the chronic blockers. I’ve written some of these privateers and asked, “Who is it you’re trying to hide……..Adolph Hitler?” Of course they never replied.

    OK, assuming none of the testing companies agreed to put the squeeze on the “privateers”. Let’s encourage a 3rd party “matching site” to disallow any privateers. Then we could all attack the testing companies by making our trees private and announce that we would go PUBLIC only at “PUBLIC ONLY” sites. That would leave the privateers alone staring at each others’ private blank trees.
    Members posting no trees are another matter. Some people have no knowledge of the identity of their family

    • Blaine Bettinger 13 February 2015 / 11:22 am

      Jim – private trees are a problem for everyone. And you’re absolutely right, people with private trees benefit from our trees while keeping their own private, which is inequitable. I would be willing to bet that you won’t find someone with a private three that has NEVER viewed a matches’ public tree, hence the inequality.

      However, your idea has been implemented to a certain degree by AncestryDNA. At AncestryDNA, there are certain features that are available ONLY to people with public trees. DNA Circles, for example, are only available to people with public trees. And I have a feeling there will be more of these types of tools. This gives people the incentive to make their trees public. And the numbers show that it has indeed been very successful at doing so.

      • Chuck Staples 10 March 2015 / 12:35 pm

        I am one with public (my family) and private (biological family) trees. And one who has tested with all three of the major companies to determine biological ancestry. So, as an adopted person, I experience the frustration with trying to get information on my biological ancestry from 1st-2nd cousin matches unwilling to respond or share their tree (or genome comparisons on other sites).
        Yet, I have a private tree with my known biological relatives to protect and respect the relative (pun more or less intended) anonymity and privacy of both ends of the private matches and keep the public tree out of the mix. I have at least one person provide me information with the stipulation it not be put on an internet-based tree for this very reason (which makes some of the tools harder to use…but I adhere to the promise). does not allow one to truly separate their profiles without having another account (which I really don’t want to pay for). I have no problem with responding to requests at all. Just ask! 🙂 But the total non-responders are just plain rude. Even a “go away” response is preferable to the silent treatment. I know when you last logged on to, the site tells me. For some folks, the private tree thing is important, but it isn’t a total wall. They key to the door is to ask politely.
        If this new tool provides some sort of triangulation, it will help those who are adopted or have ambiguous ancestry to sort things out. I just wish didn’t treat us all like children in regard to genome sharing.

  8. caith culbertson 13 February 2015 / 7:01 am

    Jim Yates, for some of us who have private trees, it has to do with integrity. Why would I put my name on a tree as factual when it is not fully documented, and allow people to use it as the truth and further promulgate lines that are potentially inaccurate? At present, it is a working file, but shared with those who ask, but with my caveats………

  9. Jim Yates 13 February 2015 / 11:07 am

    The very nature of genealogy is “she said/he said”……..not factual.
    Two out my 5 closest DNA cousins had not a single surname in common with my tree. If 1 out of 20 mothers lie about the identity of the father of their child, then NONE of us are exactly who we think we are. The “privateer-blockers” should be confined to their own anonymous site where than can stare each others’ blank trees.
    Someone criticised a line on my tree for not being sufficiently documented.
    I told him that I go with the flow, and copy the predominant lineage.
    The pedigree I was using had over 80,000 users. He replied, “Yeah and over 80,000 could be wrong.”
    I checked my critic’s tree and he had over 30,000 people in his tree. Get real! This guy lives in Hawaii and claims 30,000 ancestors & cousins in his tree. Can’t you see him abiding by his strict standards and having 30,000 in his tree? He’s another holier than thou hypocrite.

    • Blaine Bettinger 13 February 2015 / 11:25 am

      I disagree that genealogy is he said/she said based rather than factual. Genealogy should be based on facts, with reasonable interpretation (where the fats and interpretation is shared so that others can evaluate it for themselves).

      That being said, it is impossible to do that for online family trees, and so everyone must evaluate an online family tree with an enormous grain of salt. And I think that’s fine, as long as we understand that point.

  10. caith culbertson 13 February 2015 / 11:24 am

    Jim, if people deny you access to see their private trees upon your request, well, shame on them.

  11. Jim Yates 13 February 2015 / 2:11 pm

    Figure this one out. My suspected cousin, a Johnson, and I supposedly share a common ggg-grandfather, who should be a Johnson . My cuz has the results back from FTDNA for his Y-DNA. Of my cousin’s closest 11 Y-DNA matches, there are 10 different surnames, only 2 of which were Johnsons. You think there may be a little “she said/he said” in his tree?

  12. Sharon Farmer 13 February 2015 / 2:42 pm

    Hi, Blaine. I am finding this statement vague, even contradictory, and am hoping you can clarify: “placement within a circle means that you have some genetic relationship with one or more other people in the circle, all of whom happen to be descended from the identified individual, and that there is some genetic network”. In the class I took from you, you emphasized the importance of identifying groups of DNA donors who all shared the same piece of DNA—–that this gave the best chance of there being a common ancestor for all the donors. It seems like the above statement could mean that this is how the new DNA circles will be constructed. Alternatively, the statement could mean that a donor will be put in a circle if he/she matches just one of the people in the circle. (Although the phrase “there is some genetic network” seems to imply that you would have to match with more than one person to be in a circle.) Can you provide more information about the basis on which a person will be put into a circle? Thanks! And hope you are surviving the winter!

    • Blaine Bettinger 13 February 2015 / 3:08 pm


      KUDOS!! You made my day. It makes me so happy to see that my class works and that you learned the material so well. And that you’re using it to evaluate new information. Awesome!!

      You are absolutely correct that segment triangulation gives “the best chance of there being a common ancestor for all the donors.” However, AncestryDNA has made the decision not to share segment data, so DNA Circles are not based on segment triangulation. Instead, they are based on what I call “Imperfect Triangulation” in the course. In other words, they’re based on people being genetic matches and being In Common With, but we don’t know if all the individuals share the same segment(s).

      So, for example, I have a DNA Circle from a 5g-grandmother. I share DNA with 2 members of the group, and don’t share DNA with the other 5 members of the group. However, each individual in the group MUST share DNA with at least two other members (each of which must be more distant than a 1st cousin once removed). So, in other words, a “genetic network” is formed, but it isn’t based on the same segments of DNA.

      Great, great job.

    • Jason Lee 13 February 2015 / 7:02 pm

      “…the importance of identifying groups of DNA donors who all shared the same piece of DNA…”

      DNA Circles are formed on the basis of a lower standard.

  13. Jason Lee 13 February 2015 / 7:24 pm

    “However, AncestryDNA has made the decision not to share segment data, so DNA Circles are not based on segment triangulation.”

    Blaine, I need some help understanding this. How does the decision to withhold matching segment information have any bearing on Ancestry’s decision to forgo triangulation?

  14. Blaine Bettinger 13 February 2015 / 8:11 pm

    Jason – please note that this is only trying to explain what I understand AncestryDNA’s reasoning to be, not to convince you of it. The answer is the reason WHY they’ve decided not to share segment data; for privacy reasons (to avoid revealing health-informative SNPs, etc.). By not doing triangulation, then they avoid inadvertently revealing health-informative SNPs.

    You probably understand this now, but I’ll explain for others reading. This is definitely an extreme example.

    Let’s say, for example, that I triangulate a segment on Chromosome 19 from 42,000,000 to 48,000,000 with cousins #1, #2, and #3 on my father’s side of the family. However, let’s say I know that around 45,000,000 on chromosome 19, I inherited a bad ApoE marker from my father. Because I triangulate the segment with cousins #1, #2, and #3, I now likely know half of their ApoE status.

    This is at least one of the reasons AncestryDNA doesn’t base DNA Circles on segment triangulation.

    • Jason Lee 13 February 2015 / 8:49 pm

      Blaine, I’m sure you understand that Ancestry doesn’t need to release matching segment information in order to base DNA Circles on triangulated segments.

      You also know that the privacy issue applies to non-triangulated matching segments. So you undoubtedly understand that if Ancestry wants to avoid a scenario such as the one you described above, they would want to eschew non-triangulated matching segments. That, of course, would be absurd — and it’s a moot point given the fact that Ancestry withholds matching segment information.

      So I still don’t understand your statment:

      “…AncestryDNA has made the decision not to share segment data, so DNA Circles are not based on segment triangulation.”

      I’m not asking you to try to convince me of anything, I’m simply asking for clarification.

      • Blaine Bettinger 13 February 2015 / 9:08 pm

        You’re right that AncestryDNA doesn’t need to release matching segment data to members to use triangulation for DNA Circles, but segment data is at least one of the reasons that they don’t.

        If I correctly understand what you’re saying, the privacy scenario I describe doesn’t apply to non-triangulated matching segments at AncestryDNA, because unless two or more people test somewhere else (in which case all of them consent to the revelation of information, of course), there’s absolutely no way to know what segments they share.

        In contrast, if I know that I triangulate with a group in a DNA Circle, then I can easily figure out the shared segment (and thus potentially health information) of the entire group without their consent by knowing the segment I share with just one member of the group. This does require one minor step at another company or a third-party tool, but the triangulation makes it very easy to learn about the status of another person without their consent.

        • Jason Lee 13 February 2015 / 9:45 pm

          No, I’m saying that your hypothetical privacy vignette is moot because Ancestry withholds matching segment information.

  15. Nita Hanger 15 February 2015 / 11:29 pm

    Jim, Keep in mind some folks have NPE’s in their very recent family history. In some cases, it makes it difficult for them to have a public tree. Kerry Scott and Roberta Estes have written some excellent blogs on this subject. I had a good laugh over the idea of the of the Private Tree folks staring at each others blank trees. When I started out, I had a private tree because I was new and worried about inaccuracies or identity theft. I then figured out it’s better to have a well documented public one for a variety of reasons. A lot of the private tree folks aren’t as good at covering their tracks as they think they are. A few google searches and asking my other close cousins if they also match my “no tree” or “private tree” cousin usually results in my figuring out who they are if the cousin is in the 2nd to 3rd cousin range. I just wish Ancestry would come up with a better way of contacting matches. I wonder what percent of their customers ever even see they got a message requesting contact?

  16. Andrew B 16 February 2015 / 12:53 pm

    Can population-based DNA matches be used as hints that indicate a “tribal” affiliation or a relationship to a larger group of people? Sort of like the nationality makeup ancestry gives, but on a smaller scale?

  17. Madison Torres 22 February 2015 / 1:25 am

    I’ve a problem about the article, where
    am i able to e-mail the creator?

  18. Beverly 25 February 2015 / 6:14 pm

    In the course of this entire thread, I don’t see where anyone has pointed out the most glaring flaw in Ancestry’s Circles – the circles are not only based upon a DNA match but also the horribly flawed trees on Ancestry. Spelling errors, a new trend toward people who I assume are newbies listing the mother’s maiden name before their married name and just the abundance of internet garbage so prevalent in the Ancestry trees make this system ridiculous.

    And the icing on the cake is the “stories” on each ancestor’s circle. They are also based, not upon the 40 years of serious research found in my tree, but the junk that is being spread far and wide with little green hints. The story for one of my gggg grandfather’s gives a name that includes a second given name which appears nowhere in historical records and is based upon online misinformation. His date of marriage is given as 1783, never mind that his wife was born in 1778 making her all of 5 years old at the time. The 1783 date is the date one of his brothers married. I absolutely will not tolerate Ancestry adding this junk to my pages and if I have to make my tree private to stop participating in the circles I will, but more than likely I will try another provider and cancel my Ancestry subscription.

    GEDmatch with all their problems due to volunteer staff and limited resources has been far more help working with my DNA matches than Ancestry has or ever will be based upon such ridiculous enhancements as these circles.

    • Andrew B 27 February 2015 / 11:05 am

      I agree this is certainly a problem. Their expansion of the circles is going to decrease the importance of the trees when it come to inclusion in the circles, so it might actually make things a little better.

  19. Brenda 2 March 2015 / 10:59 am

    Hi, I tested with 23&me and transferred My DNA results to Family Finder. Can I transfer them to Ancestry ? Thanks

  20. Beverly 2 March 2015 / 11:23 am

    I’ve never seen an option to upload your results at Ancestry. You might want to upload them to It is a free service and many of the people who have tested with Ancestry, FtDNA and 23&Me have uploaded their raw data there so there are matches from all these DNA projects.

  21. Betty Arnold Henderson 6 March 2015 / 9:41 pm

    Mr. Bettinger, I would appreciate your thought on what may be a potential problem.
    My great-great-grandfather, James Monroe Arnold, married a full-blood Choctaw woman. Her Choctaw name was “Owah” but. somewhere in her life’s journey, she eschewed her Choctaw name and adopted the name of Telitha Frances Katz. James and Telitha were the parents of five children, four of whom had issue. has many trees posted from family historians who are directly descended from this same couple and every one of those posted trees identifies Telitha Frances Katz as a Native American. After being told all of my life that we had a Native American ancestress, I’m now beginning to wonder. My DNA Ethnicity Estimate indicates I have ZERO percent Native American ancestry. Do you have any thoughts on this conundrum? Thank you, Betty Arnold Henderson

    • Blaine Bettinger 7 March 2015 / 9:24 am

      Betty – I would also upload your raw data to GEDmatch and try the ethnicity calculators there. That being said, if your great-great-grandmother was 100% Native American, it is extremely likely that you would see some Native American in your results. Have any other descendants done DNA testing? If so, have any of them seen NA results?

      I also recommend that you find an unbroken female line, if possible, and try mtDNA testing.

      • Beverly 7 March 2015 / 9:49 am

        Betty and Blaine, I have a similar situation with a Native American 3rd great grandmother. I believe the ethnicity of my DNA results is less than 1% native american and my haplogroup is H. When I initially had mitochrondrial DNA done (in the first Genographic Project) and my haplogroup was found to be H, I accepted that all the family stories, etc. were wrong. However, a few years ago I connected with a cousin who also descends from this 3rd gr grandmother and she knew the surname of her white husband. This cousin’s grandfather had written the surname in a bible he gave her for her high school graduation. He also wrote that she was cherokee. We then found them in census and her name is listed as Unaka sometimes Younacha.

        After years of more research, I have no explanation other than the possibility that she was a white captive or descended from one. There are also new archaeological theories of a European migration to the new world approximately 18000 years ago before the Bering land bridge made migration from Siberia possible and challenging the Clovis first theory of the population of the Americas. That doesn’t really solve our problem but it does open the possibility that Native American DNA may not always be as expected.

        A few years ago my first cousin and I were contacted based upon significant DNA matches by a Native America who is Muskogee Creek. Sadly, we have never found a common ancestor.

        Anyway, my point is that it may still be possible that your ancestor was Native American. Best of luck,


        • Blaine Bettinger 7 March 2015 / 10:39 am

          All great points, Beverly. When an individual or relationship is a handful of generations ago, it can be very challenging to prove or disprove a theory.

          But not impossible! I recommend that Betty test as many descendants as possible. With a sufficiently large network of test-takers, she will almost certainly have her answer.

        • Tina 15 October 2017 / 12:04 pm

          Hello, i am also related to younacha and her husband laxton case. I have only been able to find her in a church record and also the 1850s census.. do you have any info other than that. Or any record that says she is cherokee?

  22. Jane 3 May 2015 / 8:16 pm

    Ancestry’s ethnicity estimates are highly suspect especially for populations for which they had only a few samples. For example, i just checked and found they had 131 samples for Native Americans. Spread out over the entire continent? They didn’t say, but think! How likely is it that Native Americans from Arizona are any more similar to Native Americans from Florida than Norwegians are to Italians? Europeans have been on this continent since the 1500s and most certainly have deposited their dna in many Native American populations. Now let’s examine percent accuracy. Ancestry says I’m 46% Scandinavian. If I click on Range, the prediction is that I am somewhere between 23% and 64% Scandinavian. And just how accurate is this? Based on what I could get out of their white paper example, the 23%-64% interval is about a 75% confidence interval. This means that there is about a 75% chance that somewhere between 23% and 64% of the dna they tested matches that of persons they have identified (correctly or not) as Scandinavian. This means that 0% may not be 0% at all. Now what about my presumed Scandinavian dna? I know all 16 of my great-great grandparents, and none were Scandinavian. However, all had some ancestors (1700s and earlier) from UK, Germany, France, and possibly Scandinavia. And there is probably Viking dna in most Europeans. Beware also of Ancestry’s new ancestors for you. I see that most of the ones they have found for me are the spouses of my cousins and their ancestors, most of whom are NOT related to me.

  23. steve musgrave 4 May 2015 / 3:52 pm

    When I look at the DNA estimate for family tree and the one for ancestry they match each other fairly well, if you understand the movements of people in Western Europe. One says I’m 22% Scandinavian another 5% one has much more British Isles and has a little more Western European etc.
    I don’t know how you could tell a person who’s got his European ancestry mostly through 10 or more generations of Americans can separate Scandinavian from British. Mean the Anglo-Saxons, the Jutes, and Frizians are all from northern Europe in what is now Denmark, Schleswig-Holstein, and Netherlands
    Now if you have someone who’s got a parent who is from Ireland or England or Germany you can see a pretty significant influence in both of those ethnic estimates. I Have a cousin whose father was from England they all show her to be about 80% British. My grandmother was born in Ireland so my ethnic estimate shows a little Irish. Not 25% is my grandmother’s ancestors came to Ireland the 17th century from the Scottish English border.
    Some Native American populations readily intermarried with European settlers, particularly the Cherokee there was also every intermarriage in Eastern Canada between the French and the Indians. When I do a search on Ged match using any number of the programs there I always get a 2% Native American I don’t get it in Ancestry or Family tree DNA and we have no tradition of Native Americans in my family. Recently a l ancestral line of the family that was unclear broke open and I found my maternal grandfather’s side had a strong Cajun ancestry I did not know about. This explained why I had all these cousins with French last names in Ancestry and Family Tree DNA. Following one line I found church records of intermarriage between those French ancestors and Micmac Indians
    In Acadia. Now some of my cousins so much stronger than 2% Native American and some of my cousins show 0%.

    As for the Scandinavian side I can’t find a single one but a few Swiss ancestors on my father’s. I think the trick is larger and larger populations in the mix will get clearer views of localized ethnicities.

  24. CBB 25 May 2016 / 12:22 am

    Id like my dna circle to be able to be seen when I scroll through the list of matches so I know what side of the family someone in on before clicking on them. This would come in handy for like my family were I know almost everything about one side of the family clear back to my 16th great grandmother Queen Regent Catherine DeValois. But onthe other side of the family I know very little and have a lot of faulse leads. By scrolling through the match and easily seeing with dna circle match that side of the family would make. Knowing qhich matches to click easier thereby making research alot easier. Is that a possibility?

  25. drawonceaday 8 September 2016 / 11:01 am

    We are shaped by each other. We adjust not to the reality of a world, but to the reality of other thinkers.

Comments are closed.