Announcement

Collapse
No announcement yet.

Same haplotype, different SNP

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

  • Same haplotype, different SNP

    How is it possible that two persons with identical Y-STR haplotype, both doing SNP testing and belong to two very different haplogroups?

    Noaide

  • #2
    First, keep in mind that this happens at 12 markers but not at 25 markers. I don't think we have ever seen two people with the same 25-marker haplotype but different haplogroups.

    At 12 markers, we have now seen spectacular examples of convergence. In other words, purely due to the random 'wandering' of mutations, patrilineal lines in different haplogroups actually re-converge at the same point. We have seen this for

    G and I
    N and O

    Notice in both cases that the converging haplogroups are rather close together in the haplogroup tree. N and O were actually long ago a combined NO; and G and I are both offshoots of F (although I is actually even closer to J, with whom it was once combined as IJ).


    I agree, though, that it is disappointing for a newcomer to purchase a 12-marker test in the expectation that it will predict his haplogroup, and then be told that his 12-marker haplotype has been found in two different haplogroups. He is then advised to pay $65 for a so-called backbone test, which I consider a waste of money in comparison with the purchase of additional markers (which will not only identify the haplogroup but also increase the matching power for purposes of genealogy and deep ancestry).
    Last edited by lgmayka; 1 July 2006, 06:53 AM.

    Comment


    • #3
      FTDNA could not predict my haplogroup after 37 marker test so I felt I had to order the Y Hap Backbone test.

      Comment


      • #4
        Originally posted by lgmayka

        I agree, though, that it is disappointing for a newcomer to purchase a 12-marker test in the expectation that it will predict his haplogroup, and then be told that his 12-marker haplotype has been found in two different haplogroups. He is then advised to pay $65 for a so-called backbone test, which I consider a waste of money in comparison with the purchase of additional markers (which will not only identify the haplogroup but also increase the matching power for purposes of genealogy and deep ancestry).
        The Genographic project informed me that they couldn't determine my haplogroup based on STRs so they had to do an SNP test that will delay the results few weeks. However, they apparently only tested that I'm M170 and belong to the haplogroup I. I later ordered a deep-SNP test from FTDNA that confirmed I'm I1a. But I do have some 10/12 matches with some J2s.

        Comment


        • #5
          Originally posted by davis1946
          FTDNA could not predict my haplogroup after 37 marker test so I felt I had to order the Y Hap Backbone test.
          This is extremely unusual. Did you upload your data to Ysearch? If so, can you tell us the username? I'd like to see what you have that is so unusual.

          To be fair, keep in mind that FTDNA, apparently as a matter of policy or computer programming, only makes haplogroup predictions based on the first 12 markers. You can use all 37 markers to make a prediction yourself using Whit Athey's predictor:

          https://home.comcast.net/~whitathey/predictorinstr.htm

          Or simply by looking at your nearest matches in Ysearch.

          Unfortunately, FTDNA does not 'recognize' these unofficial predictions. Hence, if Athey's predictor says you are R1b but FTDNA is unconvinced, I think you cannot directly order FTDNA's deepSNP test for R1b. You know it's appropriate for you, but FTDNA is not convinced without the results of the backbone test.
          Last edited by lgmayka; 1 July 2006, 01:28 PM.

          Comment


          • #6
            Originally posted by Eki
            However, they apparently only tested that I'm M170 and belong to the haplogroup I. I later ordered a deep-SNP test from FTDNA that confirmed I'm I1a.
            The deepSNP test is useful--I have ordered one for my uncle. But did you consider increasing your markers to 37 or at least 25? Most people would have done this first, before the deepSNP.

            Comment


            • #7
              Originally posted by lgmayka
              This is extremely unusual. Did you upload your data to Ysearch? If so, can you tell us the username? I'd like to see what you have that is so unusual.

              To be fair, keep in mind that FTDNA, apparently as a matter of policy or computer programming, only makes haplogroup predictions based on the first 12 markers. You can use all 37 markers to make a prediction yourself using Whit Athey's predictor:

              https://home.comcast.net/~whitathey/predictorinstr.htm

              Or simply by looking at your nearest matches in Ysearch.

              Unfortunately, FTDNA does not 'recognize' these unofficial predictions. Hence, if Athey's predictor says you are R1b but FTDNA is unconvinced, I think you cannot directly order FTDNA's deepSNP test for R1b. You know it's appropriate for you, but FTDNA is not convinced without the results of the backbone test.
              Last year I started out with 37 markers and was predicted as R1b by FTDNA. However, I had doubts about that. I did plug my marker values into Whit Athey's predictor and came up with R1b-47/Q-39 in the first round and R1b-27/Q-20 in the second round - hardly conclusive. So I doubted their prediction and, since FTDNA was not offering much SNP testing then, I tested through Ethnoancestry. I found out that I was indeed R1b, in fact R1b1c9*, since I tested S21+, S26- and S29-.

              If you look at my entry on the haplotype table for the R1b project, at http://www.familytreedna.com/(aa1r2n...r1b/index.aspx, you will see that FTDNA is still not confident enough to predict my haplogroup publicly, since they have not SNP-tested me yet. (I see no reason to pay for an FTDNA R1b deep-clade test, since I already know more from Ethnoancestry than FTDNA's tests will presently tell me.) On my personal page they still do predict me as R1b, but I do have a few matches with SNP-tested Q's in their database, at 3- and 4-step mutations. I don't have any exact match or 1-step muation match with SNP-tested R1b's in their database.

              The bottom line is that I have an unusual R1b haplotype that does not match many other R1b's. Anyone who has an unusual haplotype, even at 37 markers, will be hard to predict haplogroup without a SNP test.

              Mike Maddi

              Comment


              • #8
                Originally posted by MMaddi
                I did plug my marker values into Whit Athey's predictor and came up with R1b-47/Q-39 in the first round and R1b-27/Q-20 in the second round - hardly conclusive. So I doubted their prediction and, since FTDNA was not offering much SNP testing then, I tested through Ethnoancestry.
                Those Athey results would indeed be a bit disconcerting. But on Ysearch, all of your nearest matches are R1b.

                I do not second-guess your desire to be certain of your haplogroup. I have already done exactly the same thing for my uncle's Hg G. My primary point is that if a customer can only afford one of them, additional markers give much more information for one's money than the haplogroup backbone test. At 37 or even 25 markers, the customer can use the combination of Athey and Ysearch to get a very strong indication of haplogroup; and the additional markers then make further genealogy or ancestry investigation possible.

                Comment


                • #9
                  Haplogroup/No Prediction

                  Igmayka, my YSearch ID #4TWBD. I'd love for you to take a look at the markers. Thanks.

                  There are three others in the same group (not all registered with YSearch) who are in the same boat.

                  FTDNA has finally determined through YHap testing that my male cousin is R1b but they are having to do further SNP testing because the sample turned out to be M343+and P25-. Due date April 19.

                  Comment


                  • #10
                    Originally posted by lgmayka
                    The deepSNP test is useful--I have ordered one for my uncle. But did you consider increasing your markers to 37 or at least 25? Most people would have done this first, before the deepSNP.
                    Yes, I did 25 markers before the deepSNP, but I1a was in the haplogroup predictor only about 50% and I1b wasn't that much behind, so I wanted to be sure and ordered the deepSNP.

                    Comment


                    • #11
                      Originally posted by davis1946
                      Igmayka, my YSearch ID #4TWBD. I'd love for you to take a look at the markers.
                      ...
                      FTDNA has finally determined through YHap testing that my male cousin is R1b but they are having to do further SNP testing because the sample turned out to be M343+and P25-.
                      As you presumably know, you have a rather close relative named Davis (9B8UM) on Ysearch, as well as another Davis (FSU68) who might be a relative, and a Maddox (7X7B4) whose family probably split off from yours shortly before surnames came into use in Wales. Other near-matches are pretty distant, but all are of R1b haplogroup (or Unknown).

                      But your cousin's P25 is registering negative, and since R1b* is extremely rare among Europeans and P25 is quirky, FTDNA is presumably testing M269 et al. I wish that the genetics literature were more straightforward about P25. In other words, if P25 has really switched back (i.e., undone itself) sometime in R1b1 history, that would be a new clade. But no one treats it that way, so apparently the problem is with the way most companies test for it. That is, the test is not totally reliable, and instead fails on some subjects who really are P25+. For Europeans, at least, the simple remedy is to test for M269.

                      Comment


                      • #12
                        [QUOTE=lgmayka]As you presumably know, you have a rather close relative named Davis (9B8UM) on Ysearch, as well as another Davis (FSU68) who might be a relative, and a Maddox (7X7B4) whose family probably split off from yours shortly before surnames came into use in Wales. Other near-matches are pretty distant, but all are of R1b haplogroup (or Unknown).
                        QUOTE]

                        Thanks for looking at my results. The Davis 9B8UM you mentioned is rather exciting. He is from a branch of the family I did not know about. And, we have yet to establish a common ancestor (pre 1750). But, I was not aware of Maddox 7X7B4. I appreciate that information and will contact him. I do not yet know who our immigrant ancestor was or what country he was from. Do you think it is pretty certain that he would have originated from Wales? And, could we be R1b although as you said that is rare in Europe.

                        Comment


                        • #13
                          Originally posted by davis1946
                          But, I was not aware of Maddox 7X7B4. I appreciate that information and will contact him. I do not yet know who our immigrant ancestor was or what country he was from. Do you think it is pretty certain that he would have originated from Wales? And, could we be R1b although as you said that is rare in Europe.
                          Since you didn't know about Maddox 7X7B4, I should probably explain how I found him. I clicked on Search For Genetic Matches with your username, and on the resulting page I selected
                          ---
                          Show users that tested at least 37 of the markers that I did.

                          Allow maximum genetic distance of 1 per marker compared above 25 markers. In other words, I am requesting all entries that are no more than 12 away at 37 markers.

                          Do not limit search by last name.

                          All Haplogroups

                          All Regions
                          ---

                          The resulting list contains no haplogroup other than R1b or Unknown, so it's pretty safe to assume that you are somewhere in R1b (even if your cousin had not been tested).

                          You can see that Davis-9B8UM is only 1 away from you at 37 markers, so the common ancestor should be almost within reach. Maddox-7X7B4 is 7 away at 37 markers. Typically, FTDNA estimates such a distance as roughly equivalent to about 600 years--meaning that Davis and Maddox have a common patrilineal ancestor dating from shortly before the common adoption of surnames in Britain. Note that this is only a probabilistic estimate, and some experts consider FTDNA's reckoning to be considerably "optimistic"--a more pessimistic calculation would be 1000 years!

                          My guess of Wales is entirely based on the Maddox entry claiming Welsh descent. You would have to ask that entry's contact person for any more specific information. Unfortunately, the Maddox entry is almost your only clue as to country of origin, because other entries are so genetically distant and varied as to be of dubious value.

                          I think you misunderstood my comment about R1b*. What I said was that R1b* (note the asterisk) is rare in Europe. In other words, any European who is R1b is almost certainly R1b1 also. In case you don't understand the notation, the asterisk is like an end-mark meaning that the person is not a member of any further subclades. Once again, almost all Europeans who are R1b are also members of its subclade, R1b1. R1b1 is extremely common in Western Europe.

                          Let's put it this way. When your cousin gets some more SNP results, post them to this forum and I or someone else can interpret them for you if you wish.
                          Last edited by lgmayka; 1 July 2006, 09:01 PM.

                          Comment

                          Working...
                          X