Updates to 23andMe Paternal Haplogroup Assignments

With the holiday season upon us, 23andMe is sprucing up its paternal haplogroup tree!

With 23andMe population geneticist and Y-chromosome expert David Poznik at the lead, we’ve updated our Haplogroups paternal-haplo-001Report to reflect significant developments in the field over the past few years. We’re also excited to introduce yHaplo, our new open-source software for researchers.

Major Refinements to the Y-Chromosome Tree

Each generation, fathers pass down copies of their Y chromosomes to their sons. Small variations arise over time and accumulate in patterns that uniquely mark individual paternal lineages. To trace the evolutionary history of these lineages, scientists study DNA sequence differences between and among modern populations and have built a “tree” that shows how global Y chromosomes relate to one another.

However, our understanding of the Y-chromosome tree had, for many years, been limited by our incomplete knowledge of Y-chromosome diversity. Because paternal haplogroup names reflected the structure of the tree, each new insight required renaming haplogroups, and this made it difficult to interpret paternal haplogroup assignments from one year to the next. Recent research, including a study published in Nature Genetics, has drastically refined the structure of the tree. For that work, David and an international team of 42 scientists used complete Y-chromosome sequences from around the world to carry out the largest-ever study of genetic variation within the human Y chromosome (Poznik et al.). This research identified more than 65,000 Y-chromosome genetic variants, vastly increasing our understanding of the tree and setting a new standard for tracing male lineages through migrations that have occurred over the millennia of human history. 

What’s Changing

Male customers on the new 23andMe website experience can expect a couple of changes to their paternal haplogroup assignment with this update, and female customers may see changes to the paternal haplogroup assignments of male relatives and friends in other parts of the website.First, we have substantially updated our Y-chromosome tree to reflect the work of the International Society of Genetic Genealogy (as of January 4, 2016).

In most cases, the updated haplogroup assignments are equivalent to previous assignments or differ only slightly. However, since much more is now known about the tree, we can provide more information about an individual haplogroup’s history and how it relates to others.The second major update is a change to the naming system we use to report paternal haplogroups. Until recently, the convention was to use an often lengthy series of letters and numbers indicating the path of branches from the most recent common ancestor of all men to each haplogroup.

The problem is that these names changed from year-to-year as the tree was refined, making it difficult to know from the name alone which haplogroup male customers actually carry.To reduce confusion, we have moved to a system of shorter and more stable names. Each name uses a letter to identify the major branch of the tree and the name of a genetic marker unique to a specific haplogroup. For example, if we previously reported your paternal haplogroup as “Q1a3a,” we now report it as “Q-M3,” indicating that your Y-chromosome lineage belongs to a subgroup of haplogroup Q that bears the M3 marker. Because this new representation focuses on a specific informative marker associated with your haplogroup, it will be much more stable over time.


A small section of the updated Y-chromosome tree illustrating the marker-based haplogroup naming convention. The structure of the tree was aggregated from the literature by the International Society of Genetic Genealogy.

For more information on the changes coming to the haplogroups report, visit 23andMe’s customer care page, here.

yHaplo™, a New Open-Source Research Tool

The paternal haplogroup update doesn’t end with the tree. As a member of the research team at 23andMe, Poznik has developed a new algorithm to rapidly and robustly identify Y-chromosome haplogroups in very large samples, and he has implemented the algorithm as the yHaplo software package.

This software is very flexible; it runs on full Y-chromosome sequences and on smaller sets of genotyped markers. Furthermore, it is easy to incorporate updates as researchers around the world continue to gather data and learn more about the Y-chromosome tree.At 23andMe, we’re using this software to provide paternal haplogroup assignments to our customers.

As we believe the yHaplo software package can be an extremely useful tool to help drive research, we have made it available under a custom open-source software license for non-commercial research use. To learn more about yHaplo, read our white paper or head to the code repository!

  • 23blog

    Hi Maryyamada,
    I’m not sure of the specifics, but that doesn’t sound like something we would do, unless there were privacy issues around your nephew’s account. I’m forwarding your comments to our customer care to see if they can address the issue.

    • maryyamada

      Yes, it is something they did do. They explained some while ago that because everything was still so new with the new rollout, they had to put my male DNA data on the back burner. I still don’t see it back. My nephew, who has an account with 23andMe (which I paid for), has absolutely no privacy issues on this matter. His name is David Muckey.

      • 23blog

        OK. I think I understand what you are saying, and I think this has to do with assigning a paternal haplogroup to you based on your nephews results. This may have to do with what version of the experience you and your nephew are on, but I will forward this to our customer care team and hopefully they can address it.

        • maryyamada

          Thank you; the matter should have been addressed by now, so I appreciate that it’s been picked up to be worked on. My father is dead; one of his sons had a son, and that is David Muckey.

          • 23blog

            Apologies that this is taking so long. I don’t know all the details, but I know that because customers have been transitioned at different times and in waves, it makes some of the sharing functions and interoperability difficult. I hope we’ll be able to address this for you soon.

  • 23blog

    It’s pretty straightforward to download your raw data. Here is a link to the instructions:
    Basically you navigate to the “tools” menu and then click on the “browse raw data.” Once there, you will see a “download” hyperlink. Click on it and follow the instructions.

  • danodelion

    http://www.jogg.info/12/Athey.pdf). I understand that is new and evolving science, but, 23andMe, can you at least please tell us, I-M223 mutants, if we are descended from I1 or I2 ancestors?

    • 23blog

      “YCC” labels are in perpetual flux. Because they reflect the known
      structure of the tree, they change each time more information is
      discovered. At 23andMe, we rely on ISOGG as the authority on the current
      set of YCC-style haplogroup names. Please see their haplogroup I page:
      http://isogg.org/tree/ISOGG_HapgrpI.html. In it, they list the following
      line of descent for I-M223:
      I I-M170
      I2 I-M438
      I2a I-L460
      I2a2 I-M436
      I2a2a I-M223

  • MichaelSarmat

    I was examined for Gynecomastia. I am male, 78 years of age and in good health. My height is about 5 ft. 11 inches and I weight about 185 pounds. I had painful breast and conditions described by Gynecomastia. However, I was told to stop drinking caffeinated beverages and eating chocolate products. The chest x-ray did not show I have pulmonary complication. I was told to have mammography after 3 weeks. I did.

    Mammography and ultrasound techniques showed I did not have breast cancer. The pain and the breast enlargement had reduced or disappeared.

    The physician told me although the problem is mostly observed among females; some males have genetic linked association with the problem. In these people, including myself, body chemistry treats the caffeinated products as estrogen analog. As we get older, the titer of testosterone naturally would decrease resulting in enhancement of the effect of coffee in those of us who have the genetic predisposition.

    Question: have your research identified the genome associated with Gynecomastia?
    Could you please pose this question to the research group for evaluation? Thank You.

    • 23blog

      Hi Michael,
      Sorry to hear about your condition. That is something that our researchers have not studied. I can ask the question to our research team, however.

  • 23blog

    Hi Isabella,
    No, but knowing your maternal haplogroup can help you when you are looking at your DNA Relative matches. This would allow you to determine which of those matches are on your maternal line. In that way you can at least narrow down the matches that are more likely on your paternal like. Your best bet is going to be some sort of triangulating to figure this information out. So for example if you know what your father’s ancestry was, and if it was distinct from what your mother’s ancestry was, this can also help in identifying DNA Relatives who are more likely on your father’s line.

  • 23blog

    Hi BJ,
    If would be unusual if you were not both on same branch of the paternal tree. There are cases, however, where an individual’s haplogroup assignment might not exactly match those reported to the individual’s parent or sibling. Apparent mismatches arise when more data are available for one member of the family than for another. So if you both have tested on different chips, or if there was a no-call in one of your tests for a relevant SNP.

    You can learn a little more about this here: https://customercare.23andme.com/hc/en-us/articles/202906920-Why-is-my-haplogroup-different-from-a-family-member-s-

  • 23blog

    No it does not. It just changes how we name the haplogroups and this was an attempt to make it more orderly and accurate.

  • Catherine Hays

    My Haplogroup reassignment is L1-M253. How do I determine if DYS455=8?

  • 23blog

    Some of the issues around migration are tied to which market you purchased in. Our domestic US product is different than the one we sell in Canada and the UK, for instance. If you have specific questions about your account and results it would also be more appropriate to ask our Customer Care team directly. (Go here: https://customercare.23andme.com/hc/en-us/requests/new) At the blog we do not have access to your account information.

  • 23blog

    Hi Alta,
    I’m sorry that you are having this issue. I’ve sent a query to our customer care team on your behalf to try and address it. They should be able to help you directly.

    • Susan Solomon

      My understanding is that 23andme is now in a state of failing in terms of making it easy and simple to include a paternal line in a female customer’s account. My brother used to be my connection, now he isn’t actually connected, but still available as a source of the Y Haplogroup. We also lost our Parkinson’s data, rather a long time ago – am I correct on that? I can’t find mine anywhere. What is real good now is having the DNA printed out for each share so you can instantly see the genetic link. Yet overall I question the direction 23andme is going in terms of its internal goals. I feel, for instance, that using ‘Neanderthal’ DNA as part of this project should raise scientific eyebrows; many references to unproven ‘Neanderthal traits’ are intruded into this experience, which sort of lends 23andme a touch of Disney, but hey – maybe that’s exactly what the planners feel will work best.

      • 23blog

        Hi Susan, 23andMe is not failing, and our goals remain what they have been all along. Our mission has always been to help people access, understand and benefit from the human genome. What we can offer customers has changed because we operate in a regulated environment, but we are continually adding new reports, that focus on wellness, carrier status and trait reports. As for the issue with your brother’s paternal haplogroup, this has to do with customers who are sharing data being migrated to the new experience at different times. You should still be able to link to a male relative’s (your brother’s) paternal haplogroup and make it your own. (Here’s a link on how to do that: https://customercare.23andme.com/hc/en-us/articles/202906940-Linking-a-relative-s-paternal-haplogroup). If you have a problem with updating the assignment, you can contact our Customer Care team and they should be able to help. As for our past reports on certain health conditions, customers, like yourself, who had access to those reports are able to download them before migrating to the new experience. Here’s more information on getting access to the reports archive: https://customercare.23andme.com/hc/en-us/articles/219163167-How-to-access-and-share-the-reports-in-your-Reports-Archive

        • Susan Solomon

          I meant the idea of ‘failing’ in a routine, mild way – in that this linking thing seems to be a problem.

          However I’ve clicked on the link you sent in this email many times over, and it does not take me to a page that allows me to link my brother’s paternal haplogroup. The page with maternal and paternal circles does not allow me to click so as to share. I’d love to use it if it did work, and maybe I did something wrong.

          • 23blog

            OK, this may hinge on your brother and whether he is sharing back with you and or on the new experience. But I don’t know so I think it’s best that I let our Customer Care team try to help you out. Here’s a link to where to ask for help on this: https://customercare.23andme.com/hc/en-us/requests/new

  • 23blog

    Hi Truthseeker,
    The two calls are actually equivalent. M132 is a defining marker of the E1a haplogroup, hence “E-M132.”
    See, for example:
    The second row of that table is for haplogroup E-M132, whose YCC and ISOGG haplogroup names are E1a.
    Here is the line of descent from E:
    E E-M96
    E1 E-P147
    E1a E-M132

    Although E (E-M96) and E1 (E-P147) are common, E1a (E-M132), which descends from E1, is rare.

  • Mark Mitchell

    My father and I have gone from y-dna haplgroup R-L2 to R-PF6570. Which is very odd.

    1. Why did my father and I move up the tree from our previous designation of R1b1b2a1a2d3* (R-L2) to R1b1b2a1a2d (R-U152/PF6570)? This is less descriptive and less accurate than before.
    2. Why did you use R-PF6570 when R-U152 is much the better known designation throughout the genetic community? PF6570 is by far the least known of the 3 names used for this SNP. (S28 being the 2nd most common). I’ve been very involved in the U152 community for the last 4 years and had to look up what PF6570 was…almost no one in the genetic community uses this designation. If they do, very few people will know to what they are referring.
    3. For men that were “R1b1b2a1a2d” please change “R-PF6570” to “R-U152” and for all the former “R1b1b2a1a2d3” men at 23andme, please change them to “R-L2”.

    Thank you.

  • blitz

    You have to make sure that he “shares” his DNA with you. You can send him an invite to share. I had the same thing happen with my dad. After he accepted my invite he popped up as my closest relative match. If he shares with you and he still does not show up as a relative you may need to talk to your mom…

    • Melissa Lynn Collier

      He’s already sharing with me…I can see both his ancestry and health results and my ancestry results were updated. But the paternal haplogroup is still not updated. LOL there’s never been a question that he’s not my dad, I’m like the girl version of him.
      I’m wondering if there is a weird glitch that happened because we connected before his results came in…I gave him this as a Father’s Day present so that he could learn more about his ancestry.

      • 23blog

        OK Melissa,
        I’m going to forward this to someone with a little bit more knowledge on how to fix this. So expect to hear from our customer care team soon. Again my apologies for the difficulty.

  • 23blog

    You should be able to do that. First make sure that you’re sharing genomes with your brother. When you next go to your Paternal Line page you will see the option to set any share’s paternal haplogroup as your own.

  • 23blog

    Hi George,
    I don’t think there is a the difference in estimates that you’ve indicated here. You can go here to learn more about this subclade: https://en.wikipedia.org/wiki/Haplogroup_J-M172#cite_note-semino-1

  • Terry Ball

    I am considering having my father complete the 23 and me test. Can you tell me what new information I will gain as his daughter?

  • maryyamada

    I was perking along right nice with my father’s son’s son’s DNA results, until 23andMe yanked it away. Still haven’t gotten it back. How is this new dealeo supposed to help me?

  • 23blog

    When we make an assignment we only go as far was we can with confidence. So in your dad’s case, I would assume that we could not go further on the tree than we went with the T-M70 assignment.

  • Shanna Davis

    My father’s paternal haplogroup has not been updated and therefore neither has mine. Why? Have all of 23andme customers been reassigned yet?

  • joe aufmuth

    Thank you James! Your utility helped us break through to my wife’s maternal biological family! My wife had an FGS test at Familytreedna long ago and was assigned mtDNA haplogroup H27a. She tested at 23andMe, also long ago, and was assigned H. Your utility correctly assigned her H27. The 23andMe chips back then had known problems with mtDNA marker results. And now the “new” chip doesn’t even test for a defining marker of the haplogroup! However your utility assigned the correct haplogoup based on the markers the chip does test.

    I have also suggested to gedmatch that they incorporate your utility to their matching algorithms to improve their relationship projections.

    I can’t thank you enough for such an important utility! Well Done sir!

  • 23blog

    Hi Jeri,
    Yes you can do that if you are sharing with your brother. Here’s a link to the steps you need to take to make the assignment: https://customercare.23andme.com/hc/en-us/articles/202906940-Linking-a-relative-s-paternal-haplogroup