Announcement

Collapse
No announcement yet.

Basic question - I am ashamed that I do not know...

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Basic question - I am ashamed that I do not know...

    but I am here to learn.
    In a project with over 20 men in one ancestral grouping, I was pleased to help sponsor a claimed descendant from the estimated eldest son of a root ancestor, and that he did 111 markers.

    What I do not understand is: he falls into the 105/111 RELATED group compared against average based on the text at FTDNA Learning Center for interpreting genetic distance for surname shares at 111 Y-chromosome STRs, but 3 of his mismatches at are at DYS439 ("fast"), DYS389I and DYS389II, disturbing because they fall within the crucial first 12, yielding a matching cohort haplogroup but...had he tested only 12 or 25 markers, it seems he would be NOT RELATED. He shows no FTDNA kit matches with any of his surname/ancestor cohort until 37 markers.

    In my own ancestor/surname cohort, we have no mismatches at all except for one sampler at DYS389I and DYS389II and one at DYS458. We only have one man in each group who tested only to 12, and they are basically models for each subgrouping. I have been arguing for years with folks that take 12-marker tests to reunions because they are inexpensive: I have always said that a mismatch on 2 or 3 markers suggests a non-shared ancestor regardless of surname use, while a perfect match as 12 proves little, since a common ancestor even with the common surname does not mean that the common male is identifiable as an individual. In other words, I think that the only value of 12 is to get a haplogroup, but is fairly useless for genealogy.

    Is this a non-issue based on my own misinterpretation of the text at FTDNA Learning Center for interpreting genetic distance for surname shares at 12 Y-chromosome STRs? No other male in his subgrouping shows more than 1 mismatch among the first 12.

  • #2
    Originally posted by clintonslayton76 View Post
    I have been arguing for years with folks that take 12-marker tests to reunions because they are inexpensive: I have always said that a mismatch on 2 or 3 markers suggests a non-shared ancestor regardless of surname use ....
    No. It happens. We have brothers (autosomally) / half brothers (on paper) who have a gen distance = 3 at 111 - and two of those mismatches are in the first panel of 12.

    They do not match at 12. They have different surnames. They are labeled as matches at all other STR levels and at "big Y"

    So while 12 is great for at least a rough haplogroup, and so a clue as to what to do next, it may not only bring up false matches, but also miss real matches.
    Last edited by abuelita; 8th March 2019, 09:41 AM.

    Comment


    • #3
      Originally posted by clintonslayton76 View Post
      I have been arguing for years with folks that take 12-marker tests to reunions because they are inexpensive: I have always said that a mismatch on 2 or 3 markers suggests a non-shared ancestor regardless of surname use, while a perfect match as 12 proves little, since a common ancestor even with the common surname does not mean that the common male is identifiable as an individual. In other words, I think that the only value of 12 is to get a haplogroup, but is fairly useless for genealogy.
      I think Haplogroup is the key word here. If the two men are of the same basic Haplogroup they MAY BE related. If the two men are not of the same basic Haplogroup they ARE NOT related. Of course this applies to the direct male line only.

      Comment


      • #4
        Thank you both for these replies, and I am reinforced in my view that 12 marker tests should probably be dropped by vendors, and then perhaps the cost of more useful Y tests can come down. I like the idea of taking test kits to reunions if funding supports it, but not when it obtains something so equivocal saying they may be related or not, as Mr Barrett says, which is hard to explain to someone whose haplo does not match at 12 and as Abuelita points out misses a real match if only tested at that level.That suggests that there are existing male line descendants who cannot be accessed because they only tested at 12. Is this a reasonable conclusion? Are we saying that had this man tested at only 12, we would never have seen a match?

        The tracing of the various male lines is the entire reason for the project, similar to Clan Donald and Maurice Gleeson's projects.With so many variants of Slaton/Sladen/Slatten, we are trying to corral the theoretical patrilineal groups in this way, since atDNA is much broader and shallower and the surname variant is almost never seen (so far) among the results of "shares."

        As someone pointed out in another context (and I think it was Mr Barrett) another vendor gives a Y haplotype with the lower cost of the autosomal offerings, but not the DYS values. Just knowing the haplogroup is an encouragement, but for those with only 12 markers tests or haplos derived from them, it seems questionable to feel strongly about for the male line descent without a strong genealogical trail, and worse, some might be out there an unknown to me!

        Comment


        • #5
          I am GD2 with my father by Y37. For genealogical use, Y37 is the minimum you should want, Y67 being better, with Y111 or better being ideal, for the reasons your own experience demonstrated. (I'm GD3 with my father on STR markers once the BigY500 STRs are considered)

          In the case of my own BigY, my closest matches(besides my father) differed on only 8 SNPs with the Y500 test. However, we never match on Y12/25/37 or the Y67 test. They didn't test beyond Y67 and FTDNA didn't upgrade them to Y111. Going on the STR scoring from their data available through a mutual group project, our GD was high enough at Y67 that even if they did do a Y111 test, we still wouldn't match as we were beyond GD11 by the conclusion of the Y67 panels. Granted our MRCA is estimated at over 1,000 years ago all the same, but failing to match on STRs doesn't necessarily mean they're going to fail to match on SNPs.

          Y12 through Y37 remains useful for population studies on the cheap, and for "Ancient Ancestry" purposes. But for those trying to do genealogy with Y-DNA, they need to be very careful of what they're doing with STR test results.

          Comment


          • #6
            Worth stating that the text mentioned from the FTDNA Learning Center is just an estimation. I would not go by those numbers. In one surname project, I have seen that two kits can still share a common ancestor in genealogical time frame even though they have a GD of 11 and do not appear on the Y111 match lists. I have not seen any false matches in that project. For the most part those that match at Y37 will still show up in Y111 with a few exceptions (in this surname project). One of which has two multi step mutations in the Y12 panel. This caused them to have no matches from Y12 to Y37. Once they went for Y67, they had matches.

            The Y12 is still of some use. FTDNA has no need to remove it. In fact, a regular customer would not even see it as an option as it is not listed in the main site. There was this case where a project member did not match the YDNA of the claimed paper trail ancestor. I suggested he test a distant cousin to try to narrow down his NPE. He didn't know any so he went with a first cousin. He used a Y12 for this test. The cousin did not match him at all, even the haplogroup was different. Obviously a red flag, I told him to test a different first cousin. Another Y12 used. It matched the first cousin he tested. He was the odd man out. If he had used Y37, it would have cost him more for the same outcome. So Y12 is still good for ruling testing candidates out.

            Comment


            • #7
              The_Contemplator
              Which is where I think FTDNA has struck a decent balance on this, although making sure the Project Admins understand things could probably use a little more education out-reach.

              "Normal Users" are going to be hard pressed to get their hands on a Y12 or Y25 test kit, so long as Project Admins and/or other "helpful types" don't give it to them on a silver platter. Basically at this stage, if you're ordering a new Y12 or Y25 kit from FTDNA its assumed you know what their pros and cons are, and that you have a reason for pursuing that particular option over Y37 or higher.

              Comment


              • #8
                I correspond with (and am a member of) the Guild of One Name Studies, and I continue to see postings about buying the less-expensive smaller Y tests in bulk for taking to reunions (mainly in the UK). But there are plenty of Americans who read the blogs there.
                This is not the thrust of my question, but I wish to point out that at least one haplo admin wrote to me that he advises that men only test at 12 for the haplo and then do SNPs.
                This is completely opposite of my advice to potential members of my study, which is that 67 markers is the minimum for a truly useful test for our purposes of corralling what I interpret as unrelated male lines. My goals are obviously different to that person's. Again, thanks very much for all replies.

                Comment


                • #9
                  @clintonslayton76, the perceived discrepancy comes from the margin of error being very high with only 12 markers. Then decreasing with 25, 37, and 67 markers.

                  My guess is that starting with Y-DNA111 the current model that uses predictions from mutation rates gives too narrow intervals for relationships (from - to) given true randomness of mutations.****

                  We had seen that randomness with people reporting in the forum about sibling or father-son pairs having STR differences at Y-DNA67.


                  Mr. W.


                  ****
                  A similar process involving light is described in physics books. When there is lots of light (many photons), very nice and easy equations can be used, and they give us very good results. When the light intensity drops to the level that we are observing individual photons arriving randomly only from time to time, then the very different (and complex) mathematical apparatus needs to be involved. And still over a very long period of time flow of individual photons agrees with the nice and easy equations.

                  Either time has to be very long or many elements have to be concerned in order to be able to use predictions (equations) that assume a continuous process. STR mutations or photons...

                  There is a point when the equations that assume continuum stop giving results that flawlessly agree with observations. FTDNA does acknowledge that probabilities are involved.


                  Does that mean the population studies should move from Y-DNA12 to Y-DNA25? Maybe. However, if the sample (number of men) is large enough, some correct conclusions about populations can be drawn even with only 12 markers.

                  Comment


                  • #10
                    The other point to be raised is going to a Family Reunion for my Grandparents, Y12 test kits shouldn't come up with much in the way of unusual results. You might have one or two people turn out to be GD1 on Y12, but statistically speaking it should take a pretty significant sample size to even achieve that. Of course, statistics being what they are, and random being random, the very first person you test could be GD2 at Y12. (Which underscores getting atDNA testing as well, so you don't create a scandal where there wasn't one)

                    But as the the distance to a common ancestor moves further back in generations, the more likely it becomes that multiple mutation events will present themselves, and the testing strategy needs to adjust accordingly.

                    Comment


                    • #11
                      Thanks, everyone. My take-home message from this is that the mismatch ratio for a 111-marker test among men who took from 12- to 111-markers do not effect the idea of two men with high probabilty of sharing a common ancestor, despite the fact that one man's 12-marker results do not "match" that of a 111-marker sampler when limited to under 37-marker levels of matching. They share the same surname and haplogroup prediction.

                      Some of our results are transfers or invited members; those that I have encouraged or sponsored have always been at the 67- and above marker level.

                      Seeing this man with "colorized" mismatches within the first 12 DYSes troubled me, because we have no other claimed descendant from the presumed eldest son of the common ancestor mooted for these men from a genealogy paper-trail. In time, I can hope that we get at least one more claimed eldest son descendant to test if those same mutation events occur.

                      Comment


                      • #12
                        Pretty much, from what I've seen when I went looking for mutation rates, the Y-12 markers aren't that stable all things considered. Their predicted mutation rates are inline with most of the other STR markers that are being checked. They just happen to be the easier ones to test, and the odds of any one specific STR mutating are pretty low. So generally speaking Y-12 should match with 0 mutations.

                        But with Statistics being what it is, and random mutations being random, somebody is going to be the (un)lucky guy who achieves the Genetic equivalent of winning the PowerBall jackpot.

                        Much like my getting hit with a ploymorphic marker mutation on both the Y25 level and again at Y37 against my own father.
                        Last edited by bartarl260; 12th April 2019, 12:21 AM.

                        Comment

                        Working...
                        X