No announcement yet.

Which Y-DNA test to have a potential relative take to establish MRCA

  • Filter
  • Time
  • Show
Clear All
new posts

  • Which Y-DNA test to have a potential relative take to establish MRCA

    Hi all,

    I've done a Y-111 test. Separately, through documentary research, I've identified a potential (but as yet unconfirmed) relative who has a very well documented lineage. Due to geographical proximity and a similar (yet uncommon) surname, it seems relatively likely that we share a common ancestor within the surname period.

    This potential relative has kindly agreed to take a Y-DNA test. If we turn out to be a match, this will be a big step for me towards overcoming a brick wall and hopefully prove a connection to the well documented lineage. In addition to confirming whether or not we share a recent common ancestor, I'd like to estimate how many generations ago our MRCA lived to as accurate a degree as possible, so I can then follow up with targeted documentary research.

    My question, on which I'd very much appreciate advice, is which Y-DNA test I should get my potential relative to take? My instinct is to have them take the Y-37 (given how much cheaper it is), to confirm whether or not we share a common ancestor. If we do, I could then upgrade both my and their tests to a Big Y-700, to get the most accurate prediction of how many generations back our MRCA is. Is my approach the right one?

    In particular, just to check:
    • Whether a Y-37 match is sufficient to prove a recent shared ancestor? (I'm pretty sure on the answer to this one but I'm so new to genetic genealogy I thought it worth asking the question anyway.)
    • How accurately a Y-37 match can predict how many generations back the MRCA is?
    • Whether Big Y-700 is significantly more accurate at predicting how many generations back an ancestor is shared, when compared to Y-111 or Y-37? Assuming so, how accurate is it possible to be?
    Sorry for what are probably quite obvious questions! I have the best of intentions to properly educate myself on genetic genealogy and to start a surname project in due course - so hopefully in a month or two I'll be able to answer these questions myself! But given relatively lengthy turnaround times for tests, I'm keen to get the ball rolling in advance of having a chance to sit down and swot up.

    Thanks in advance!

  • #2
    There is no easy answer. Y37 has price going for it. That's all. A Y37 is a "looks like" test. The result "looks like" a match or not, but you cant be sure. Y37s are generally reliable but there is a nagging real chance of false positive or negatives,

    If your budget allows, get the BigY.700. The SNP part will absolutely verify if you are a true Y match. And you get a Y111 included. So the best of Y-STR and Y-SNP worlds. You will save time and know that you have the best answer possible.

    As for getting an accurate prediction for how many generations, you will only get a best guess, even with the BigY. There are predictors out there that give decent averages over groups of men over long periods of time, but for individuals, its more variable.

    One plus for the Y37. You will get basic haplogroups and if you dont match those, then no purpose in going forward with a BigY.. So in that case, a Y37 is a prudent $ move.


    You can also get haplogroups inexpensively via AncestryDNA coupled with Cladefinder (free). Or by testing at 23andMe or LivingDNA which are economical, particularly when on sale. If your prospect gets a compatible haplogroup there, then proceed at FTDNA
    Last edited by mabrams; 23 January 2023, 06:08 PM.


    • #3
      Originally posted by james1994 View Post
      In particular, just to check:
      • Whether a Y-37 match is sufficient to prove a recent shared ancestor? (I'm pretty sure on the answer to this one but I'm so new to genetic genealogy I thought it worth asking the question anyway.)
      • How accurately a Y-37 match can predict how many generations back the MRCA is?
      • Whether Big Y-700 is significantly more accurate at predicting how many generations back an ancestor is shared, when compared to Y-111 or Y-37? Assuming so, how accurate is it possible to be?
      Sorry for what are probably quite obvious questions! I have the best of intentions to properly educate myself on genetic genealogy and to start a surname project in due course - so hopefully in a month or two I'll be able to answer these questions myself! But given relatively lengthy turnaround times for tests, I'm keen to get the ball rolling in advance of having a chance to sit down and swot up.
      Basic Y-DNA testing is limited for predicting with accuracy just how many generations back you are related. But if you have some evidence that you may share a recent ancestor, why not try an autosomal test first? As mabrams mentions, you could test at 23andMe to get the Y-DNA haplogroup to a certain degree, plus get a relationship estimate. These autosomal tests go on sale often at the various companies during the year. If you and your potential relative do such a test, and show as related autosomally to any degree, even distant, and share the same YDNA haplogroup designation (links to 23andMe info), then you could do further Y-DNA testing at FTDNA (Big Y-700 would be best) to find out more. If you do an autosomal test at a company other than FTDNA, you can transfer your raw data file from 23andMe, Ancestry or MyHeritage (but apparently not from LivingDNA) to FTDNA to get accounts here, and then order further tests. Keep in mind that any further tests at FTDNA when ordered from an account based on an autosomal transfer, would need to supply a new DNA sample, since only the autosomal data is transferred, not the DNA.

      Since you have already done a Y-111 test, if you haven't already joined a Y-DNA project for your predicted haplogroup, surname, or geographical group, you should do so. The administrators might be able to guide you for other testing.

      mabrams pointed out that a Y-37 test will tell you if you are in the same basic haplogroup, so you may want to just go that route before any other Y-DNA testing. As STR tests, Y-37, Y-67, or Y-111 (Y-111 included in the Big Y-700, as mabrams wrote), have a feature called the FTDNA TiP Calculator, which FTDNA currently describes as:
      The Time Predictor, or FTDNATiP™, is a program that incorporates Y-chromosome DNA STR mutation rates to increase the power and precision of estimates of Time to Most Recent Common Ancestor (TMRCA).
      TiP is more of a probability distribution than an exact number of generations, as it says on the linked page, and also says it gets more refined with more STRs tested (say, Y-67 or Y-111 vs Y-37). You could ask your prospective relative to do a Y-111 test and see if he matches you

      The other kind of Y-DNA test, for SNPs, will give you a haplogroup result, but eventually they will add up, because if you want to narrow down your branch, you will need to do more SNP testing. That's why the Big Y-700 is advised; to save all the whittling down. It is an SNP discovery test; if you and your prospective relative were to take a Big Y-700 test, you may even help form a new twig on your haplogroup's branch.

      A Big Y-700 test would reveal if you and your potential relative are in the same branch/twig on the Y-DNA haplotree. So if you go another route vs. doing Big Y-700 first, and you find either from a 23andMe Y-DNA haplogroup result, or an STR test at FTDNA that you and your prospective relative actually are in the same basic group, then investing in the Big Y-700 for both of you will show if you are in the same later branch or not.

      Here are some links for pages from an old FTDNA FAQ, which have information I couldn't find in the current Learning Center (maybe I just missed it, though):
      Look for the questions that are like "If two men share a surname, how should the genetic distance at (#) Y-chromosome STR markers be interpreted?," but there are other topics that may be of interest to you in the list. Use the menu at the top of the linked page for other categories for Y-DNA and others.

      Last edited by KATM; 23 January 2023, 08:05 PM.


      • #4

        Earlier this week I was reviewing some Y tests in a project I co-admin.

        Two men with the same surname, and a perfect Y67 match. Both are R1b or R-M269

        Yet one is from the R-P312 branch (from his BigY test) and the other is from the R-U106 branch. That is incompatible. They are not a match after all, despite 67/67 and same surnames.

        The U106 tester is from an older test. Maybe that test got the wrong result, and I will look into it, but I dont expect that to be the problem.

        In general my experience with Y37 is that they are reliable, and the situation above is atypical. But as I said in my original post, "The result 'looks like' a match or not, but you can't be sure."


        • #5
          Thanks mabrams and KATM for your incredibly helpful insights - very much appreciated!

          KATM - thanks for the suggestion of trying an autosomal test first. My understanding is that autosomal testing is only really helpful for identifying matches where the MCRA is 5 or 6 generations back, and I can rule this out through documentary evidence. So would it still be valuable to do an autosomal test?

          I should have clarified in my initial post - but documentary evidence proves that me and my potential relative don't share a common ancestor within the past 9 generations. However, we do share a similar surname, so if we do share a recent common ancestor, I would expect that ancestor to be within the British surname period - i.e. for our MRCA to have lived sometime between c.1100 CE and c. 1700 CE. This would, by my rough calculations, be in the region of between 40 and 10 generations up the line.

          Assuming we are a match, one of the main things I'd like to be able to do is estimate where in that 600 year/30 generation window our MRCA is placed. I know this will all come down to probabilities, but does anyone know of any data, formula, calculator, study, etc. that would help me in understanding the different probability outputs that I could expect from comparing Y-37 results vs Big-Y 700 results?

          On a slightly different note, I was interested to see you both refer to Y-111 as being 'included' in the Y-700. I thought this was odd, as my understanding was that Y-700 tested for about 700 STRs. Going back to look on the website, I note that while the Y-700 test does indeed test '589+ bonus STR markers' that these are 'not used for matching purposes'. Does anyone know why FTDNA wouldn't use these additional markers for matching purposes, in the event that there is a match between two people who have both taken a Y-700 test? Would the similarities/differences in the additional 589 markers still be taken into account by the FTDNATiP tool?

          Again, apologies if any of my questions result from my misunderstanding; and thanks, once again, for your help!


          • #6
            I hesitate to throw in my two cents, since the answers have been so good already, but I will anyway.

            I agree with the idea that you should find out if it's worth spending as much money as the Big Y-700 will cost before actually doing it. So, a Y-37 is probably the best way to start. However, based on what you've written thus far, James, it sounds like you and your potential Y-DNA relative are both going to have to go for the Big Y-700, that is, if you match at Y-37. Your relationship may not be close enough in time for an autosomal test to show, although one never knows. My best-ever, breakthrough Y-DNA match started out as a measly 9 cM match at Ancestry, but that was miracle territory: our Y-DNA MRCA was born about 1750.

            FTDNA does report how far off your Big Y-tested matches are on the additional STRs, but you won't be all that interested in that, because the SNPs tell a much more accurate story. STRs are really yesterday's news, but STR testing is still around because it is an economical way for people to get their feet in the door. The future is in SNP testing.

            I am attaching a couple of things that attempt to show how closely various STR matches might be related. Hope they're helpful.

            STR matches and relatedness_genetic distance_dna.jpg

            Y-DNA_111 y dna str marker_genetic distance and range of generations_David Vance.jpg


            • #7
              I hope this post doesn't overlap what Stevo wrote - I left my post written and forgot to post it for too long. Stevo has done a great job with his points and charts. I'll still post mine, for what it's worth:

              james1994 Apologies, that was my mistake, and poor reading comprehension! You did say in your first post, in the first couple of sentences, that there is no common ancestor within the surname period. So with that now brought to my attention, it's probably unlikely that an autosomal test will be of use to you. I have three examples for the farthest back generation-wise to MRCA that I've seen in autosomal matches with kits that I manage:
              • 7 generations back to the common ancestor for the match, and 5 generations back for the person whose kit I manage. They are 4th cousins, twice removed, and the common ancestor was born maybe 1770-1780s in Ireland.
              • 7 generations back to the common ancestor for the match, and 7 generations back for me. We are 6th cousins this way; our common ancestral Maltese 5th great-grandparents were born in the early 1700s (based upon their marriage in 1729, as neither the match nor I have their birth dates).
              • But I am also related to this same Maltese match as a 9th cousin via a second Maltese line, where the MRCA couple is 11 generations up for the match, and 10 generations for me! This ancestral couple was married by 1605. So it goes with an endogamous population.
              A caveat: for the two latter (Maltese) examples, the match has done the research beyond my confirmed 3rd great-grandfather, which I trust due to the match's known personal connections to those who keep the records in Malta, from which he based his information. Although I trust his research, I do need to confirm the early generations to be truly sure.

              Regarding Big Y-700, when we say the Y-111 is included: On the Help Center page "A Comparison of Our Y-DNA Tests" (in what used to be called the Learning Center, not sure when that name was changed), it explains that
              The Big-Y 700 test includes both STR and SNP results and provides matches for both. In addition to the STR markers included in the Y-111, it provides a minimum of 589 additional markers, giving you a total of 700 STRs, though only the first 111 are used for matching. These additional markers can be useful for determining mutations unique to a lineage, clan, or patriarch.
              So they use those first 111 STRs for Y-111 matching.

              I manage a Big Y-700 kit which has had one update in its terminal SNP in the last year or two, moving from the previous SNP which was estimated at 608 CE, to the current SNP estimated as 978 CE. I expect that eventually, as more people do the Big Y-700, more men will match other private variants for that kit and the terminal SNP will change again, hopefully to get within a genealogical timespan.

              I'm not aware of any formulas or calculators, etc. that would help with the probabilities, but from what I wrote above, I think you can see that Y-37 and Y-111, as STR tests, get you an estimated haplogroup. Doing higher STR level tests don't change the estimated haplogroup, but SNP testing will. Comparing STR testing to the Big Y-700: Big Y-700 tests both STRs and SNPs, but importantly also discovers new variants. When two or more men have the same variant, they form a new branch. Big Y-700 will refine your terminal SNP as more men test and others match your private variants. Your updated terminal SNPs will advance in time (to more recent centuries/decades) as this happens. STR tests can't do that.​

              For lowest cost, you could have your potential relative do an STR test at FTDNA, such as the Y-37, and you then would be able to see if he matches you at that level. If so, he would then share your estimated haplogroup. If the possible relative does a Y-37 and you find he is in the same estimated haplogroup as you, he and you could also do SNP testing to narrow down the branch. That would be a longer path, though, which is why we say do the Big Y-700 and be done with it. Keep an eye out for sales (Father's Day, end-of-year holiday sales).

              But I would say that it would be best to get advice from a project manager for further tests. Join a project for your (uncommon) surname, if one exists, or, since you have an estimated haplogroup from your Y-111 test, join a haplogroup project for that main branch. Some haplogroups have STR packs, to test for sub-branches below the main haplogroup. The only glitch there is that you can then be in a cycle of getting more tests to get further out in the branches of your haplogroup.

              To get what you want, to see where in the 600 year/30 generation window your MRCA would be, you would both need to do the Big Y-700, although it might not put you both in that timespan. Big Y-700 is much more expensive, so unless you are pretty confident you are related via that ancestor, I would hesitate to skip right to Big Y. But, if you either compare first with STR testing, or jump right to Big Y-700, what Big Y-700 will do is:
              • tell you whether this potential relative and you share a private variant, which will place both of you in a new terminal SNP, or
              • tell you that he has a nearby SNP, but not one that shares the same variant defining your SNP (meaning perhaps that you or one of your ancestors had one or more mutations since the common ancestor), or
              • tell you that he is has a different terminal SNP than yours, and is even further away, or not related
              You can take your estimated haplogroup you got from your Y-111 test, and check it on FTDNA's Y-Haplogroup tree. Whatever it is, you will be able to see many branches and twigs below it, and one of those twigs is where you will be if/when you do either SNP testing or the Big Y-700 test.

              BTW - I am not any kind of Y-DNA expert, but I am just relating what I have learned from reading about it, and viewing webinars and videos on Y-DNA and other types of DNA. I believe what I've written above is true, and hope if I made any errors that someone will chip in and correct or refine my comments.​


              • #8
                Ancestry, MyHeritage, and 23andMe all include limited Y-DNA SNP testing. The FamilyFinder test also could support this but FTDNA has opted not to implement it. This is unfortunate as it could be a big plus for encouraging more advanced tests. It would probably hurt their SNP pack market though.

                Even if there is no autosomal match, these tests would likely provide a more accurate answer than any predicted by Y-37 (also less expensive). A matched haplogroup could be enough to go ahead with the Big Y-700 test.

                23andMe includes about twice the number of SNPs as the others and also reports the result without having to use an external extraction tool. They use an old haplotree, so the reported group may not be quite correct but the error would be the same for both tests if there is a match.
                Last edited by JeffZ; 30 January 2023, 11:52 AM.


                • #9
                  JeffZ, what you wrote is true, and would work to get an idea of the Y-DNA haplogroup of the possible relative. But as james1994 stated in his original post, the possible relative has agreed to take a Y-DNA test. He may, or may not, be willing to do an autosomal test at a different company, then if results are promising for the haplogroup, test again at FTDNA to delve further with Big Y-700 or other Y-DNA testing. It depends on how willing the proposed tester is to do more tests at more companies. Even if he transfers his file from 23andMe or another company to FTDNA, that transfer is just the raw autosomal data, not the actual DNA sample. He'd have to give another sample to FTDNA so they could perform tests on it.

                  james1994 - If he only wants to give one sample, then the Y-37 test at FTDNA would seem to be the best solution, as then his sample is already at FTDNA and further testing can proceed. Just MAKE SURE that he does the swabbing/scraping very well, so the sample is of high quality for further testing. I speak from experience about that - I manage one kit where the tester's sample was insufficient to proceed to Big Y-700; it had sufficient DNA for the initial Family Finder and Y-67 tests, and an upgrade to Y-111, but no more, even with two vials. I kick myself for not having the Big Y done vs. the Y-111 at the time. Big Y-700 usually requires a good, preferably fresh sample. I was not present to observe how well he swabbed, and had to rely on other, more disinterested parties to supervise him getting the sample and then mailing it in. His sons do not want to do a DNA test, so I am stuck at Y-111 for this person, with his estimated Y haplogroup. A project manager predicted a lower branch, but it's nowhere near what Big Y could do. So the moral is to swab well, and follow the directions!

                  Currently, the regular price for Y-37 is $119, so it is indeed more expensive than an autosomal test, even when those are not on sale (as they seem to be more often than not!). The Big Y-700 is regularly $449, so is $330 more expensive than the Y-37. I haven't saved information about sale pricing as I did in the past, but during a sale the Big Y-700 might be $50 off, so about $399. It might be lower at Father's Day or even winter holiday sales. Looking at some prices from at least 2016-2019, the Y-37 used to be regularly priced at $169, so has come down a full $50 in price since then. For the Black Friday sale in 2022, FTDNA sold the Y-37 for $79, and the Big Y-700 for $379. Father's Day is not far off, and "DNA Day" is in April. Keep an eye out for sales.
                  Last edited by KATM; 30 January 2023, 02:18 PM.