Announcement

Collapse
No announcement yet.

R1b1b2e haplotype w/ R1b1b2h1 SNP?

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

  • R1b1b2e haplotype w/ R1b1b2h1 SNP?

    Anyone care to comment on the following?

    As I believe I may have mentioned in another post, I am investigating some unusual Y results reported for myself and others of my surname.

    One particularly strange phenomenum that I am curious about is an individual w/ haplotype strongly suggestive of R1b1b2e subclade BUT reported SNP of R1b1b2h1.

    I wish I had more details to report, but the Y Search account holders don't appear too eager to respond to inquiry. Go figure.

    Anyhow, these reported results initially struck me as ridiculous because I believed that:

    #1. R1b1b2e haplotype was considered VERY distinctive.

    #2. There is no particularly tight association between R1b1b2h1 SNP and any specific haplotype, but that most variants are pretty far from R1b1b2e haplotype.

    #3. It is NOT possible for a single individual to fit within BOTH the 'b2e AND 'b2h1 subclades.

    But maybe 1 or more of these ideas is wrong?

    I note that of the 7 entries in YSearch for folks w/ R1b1b2h1 haplogroup, 2 of them have strongly R1b1b2e haplotypes.

    I share a surname with one of the individuals--not exclusively associated with any specific origin, but is usually considered English.

    The other case noted in YSearch has an old Cineal Eoghain surname, so I have little doubt that R1b1b2e haplotype is basically correct.

    Anyone have any insights as to the probability that R1b1b2h1 SNP assignments for these fellows are correct?

    Thanks much.

    Jack

  • #2
    Originally posted by Clochaire
    Anyone have any insights as to the probability that R1b1b2h1 SNP assignments for these fellows are correct?
    Thanks much.
    Jack
    Unfortunately, Ysearch does have various entries with the wrong haplogroup specified. The haplogroup field is just a pull-down menu and can be easily changed by hand by the account owner. So it's possible for someone to select the wrong haplogroup when they've been SNP-tested as another haplogroup, and it's even possible for someone who has never been SNP-tested to enter a haplogroup.

    In other words, if a haplogroup designation in Ysearch doesn't look right, then it may very well be a user-entered error.

    Unfortunately, unless the user in question responds to your email, all you can do is speculate about why the seemingly incorrect haplogroup is listed.
    efgen
    Webinar Coordinator
    Last edited by efgen; 8 July 2008, 10:44 PM.

    Comment


    • #3
      Originally posted by efgen
      Unfortunately, unless the user in question responds to your email, all you can do is speculate about why the seemingly incorrect haplogroup is listed.
      Exactly why I'm wondering if the 3 premises I outlined in the original post are reasonable. I know that conclusions can only be weak and provisional, but I'm learning anyhow.

      For example, it occurs to me that haplogroup R1b1b2 in general is defined by multiple SNP's. Then, in theory, at least, isn't it possible that we will ultimately discover that the SNP's that supposedly exclusively define distinct subclades are in fact not so exclusive? Maybe there are individuals out there that have SNP's that characterize BOTH R1b1b2e AND R1b1b2h1, but they just haven't been identified yet?

      Cloud-talk to be sure, and maybe the answer would be quite clear to me if I had a deeper understanding of the nature of SNP's or the ISOGG vetting process. But I personally do not know at this point, and would like to challenge my understanding.

      Comment


      • #4
        Tour of R1b1b2h Project

        Just thinking out-loud here.

        But looking at R1b1b2h project results, it looks like there are a number of individuals in this hg who share some of (what appeared to me) the stranger DYS values reported for me: 26/14/18 at 447/437/448.

        Maybe this fellow I discussed above, with my surname, and R1b1b2e haplotype but with odd w/ self-reported R1b1b2h1 does have some significance with regard to my discordant results discussed elsewhere.

        Lots of gaps in info, though. 1.) I guess I tested neg. for M126. 2.) I never tested for S28/U152. Given problems encountered to date, I have real reservations about throwing more $$ into this pit. 3.) This fellow seems to have declined to respond to an email inquiry. Don't blame him. Most sane people in a fix like I describe would have written off DNA testing altogether long ago.

        Most people are likely to have heavy assumption that lab results are largely correct, and these phenom are totally unrelated. Maybe that view will ultimately be correct. My gut here is that these phenom' are related, somehow. One way or the other something strange is going on with these result. I figure:

        1. Either I have very rare pattern of natural mutations in my own profile, and these crypto-R1b1b2e fellows are of no immediate significance to me at all. That would still leave the very weird coincidence that we tested at the same time, and that so many guys of this largely-S.E. England surname match dna of my suspected PE-guy, an O'Donnell.

        OR

        2. We all have a surprisingly large # of mis-reported loci and/or very rare pattern of natural mutations, but we all stem from the same origin and should have the same true haplotype. Another curious phenomenom may be relevant here: 2 of these guys claim relation to a somewhat famous 18th century figure who happens to have lived in the same county as my paper-trail ancestors. I tested negative for M126, which runs counter to this theory, but the data entry error scenario efgen pointed out is quite likely. Maybe it should really be R1b1b2h. Won't know for sure without that fellow's input.

        OR


        3. Our results are somehow switched. But there is a gi-normous # of factors running counter to this: 1.) I tested neg for M222. Well, I was reported to be neg M222.; 2.) despite unexplained descrepancies reported for me and discussed elsewhere, this scenario would entail a mind-bogglingly massive volume of actual mis-reports which would be un-nervingly consistent accross multiple labs.


        I guess that unless I am able to implement some sort of test plan these scenarios can be nothing more than speculation. Smart money would probably be on #1. Innocent coincidences do happen, but the O'Donnell coincidence thing startled me, esp. given unexplained discordances discussed elsewhere. Not quite done thinking about that idea.


        Jack

        Comment


        • #5
          Originally posted by Clochaire
          . . . I believed that:

          #1. R1b1b2e haplotype was considered VERY distinctive.

          #2. There is no particularly tight association between R1b1b2h1 SNP and any specific haplotype, but that most variants are pretty far from R1b1b2e haplotype.

          #3. It is NOT possible for a single individual to fit within BOTH the 'b2e AND 'b2h1 subclades.

          But maybe 1 or more of these ideas is wrong?
          I think ideas #1 & #2 may not be quite right. After doing some more touring around, including Kerchner's R1b1b2 and subclades projects, I've seen quite a few individuals w/ vaguely-R1b1b2e haplotypes and R1b1b2h (or equivalent R1b1c10) SNP reported.

          So, perhaps I should not be too suspicious of the original report of someone with R1b1b2e haplotype and R1b1b2h SNP. It seems as if these subclades may even be related. Lots of talk of R1b1b2h representing La Tene Celts. Maybe R1b1b2h is ancestral to R1b1b2e?

          Jack

          Comment


          • #6
            Originally posted by Clochaire
            I think ideas #1 & #2 may not be quite right. After doing some more touring around, including Kerchner's R1b1b2 and subclades projects, I've seen quite a few individuals w/ vaguely-R1b1b2e haplotypes and R1b1b2h (or equivalent R1b1c10) SNP reported.

            So, perhaps I should not be too suspicious of the original report of someone with R1b1b2e haplotype and R1b1b2h SNP. It seems as if these subclades may even be related. Lots of talk of R1b1b2h representing La Tene Celts. Maybe R1b1b2h is ancestral to R1b1b2e?

            Jack
            It is the case that R1b1b2e and R1b1b2h are both S116+. In other words, they share a common ancestor sometime between the beginning of R1b1b2* (M269) and the birth of their own SNP-defined lineages.

            I think you've been giving too much credit to the idea that you can tell an R1b1b2 haplotype's subclade by some distinctive marker values. The closest you can come to being accurate by that method is DYS492=13 strongly indicating R1b1b2g (U106). It's well-known that R1b1bh has no distinct pattern that anyone has been able to detect, although there are some (not all) with DYS492=14, while most have DYS492=12, the R1b1b2 modal value.

            Comment


            • #7
              Originally posted by MMaddi
              I think you've been giving too much credit to the idea that you can tell an R1b1b2 haplotype's subclade by some distinctive marker values...
              Mike,

              Recent observations do seem to bear out this thought.

              But I think the idea that there is such a close correspondence for R1b1b2e has been sold pretty hard. That's certainly the impression I was left w/ after reading M222 materials.

              But think on this: One of the R1b1b2h1 fellows that I mentioned above matches the modal R1b1b2e haplotypes to an eerie degree on 67 loci--385b:13; 392:14; 481:25, etc., etc. And he's got a wicked ol' Cineal Eoghain surname.

              In fact, I could site a couple other similar examples, with old Ui Neill surnames, rendered less shocking only by lower resolution reports--only 25 markers.

              So now I wonder 2 things:

              #1. Does SNP mean anything at all?

              #2. If I should understand SNPs to really be as stable as advertised, doesn't this suggest a simply staggering level of convergence? What would such a high level of convergence suggest about mutation rates over the long-term? I take this to indicate that current models may be WAY understated.

              Jack

              Comment


              • #8
                Originally posted by Clochaire
                Mike,

                Recent observations do seem to bear out this thought.

                But I think the idea that there is such a close correspondence for R1b1b2e has been sold pretty hard. That's certainly the impression I was left w/ after reading M222 materials.

                But think on this: One of the R1b1b2h1 fellows that I mentioned above matches the modal R1b1b2e haplotypes to an eerie degree on 67 loci--385b:13; 392:14; 481:25, etc., etc. And he's got a wicked ol' Cineal Eoghain surname.

                In fact, I could site a couple other similar examples, with old Ui Neill surnames, rendered less shocking only by lower resolution reports--only 25 markers.

                So now I wonder 2 things:

                #1. Does SNP mean anything at all?

                #2. If I should understand SNPs to really be as stable as advertised, doesn't this suggest a simply staggering level of convergence? What would such a high level of convergence suggest about mutation rates over the long-term? I take this to indicate that current models may be WAY understated.

                Jack
                What level of convergence are you talking about at the 67 marker level, in terms of genetic distance? That's as important as matching on three specific markers you've mentioned as generally indicating M222+. (Although I do admit that if they have surnames that would seem to strengthen an identification with M222, that is suggestive.)

                For instance, I am a member of a "null 425" cluster of R1b-U106. My closest neighbor in terms of GD is 14, at 67 markers. Many other members of this cluster are have a GD of 20+ in comparision to me. Yet, when I do a search on ysearch for matches with a GD of 25 from me, I get a significant number of people with GD from me of less than 20, who are not cluster members. (It's easy to tell who's not a cluster member, since they won't have a "null 425.")

                What I'm saying is that it's possible to have a GD of 20 or slightly less and be comparing men in two different subclades. I don't know of a specific case, but I bet there are cases where two men have a GD of even 10-15, at 67 markers, and are in different subclades.

                So what is the GD between the man who's M222+ and the other who's supposedly U152+?

                Comment


                • #9
                  I was hasty in my original response when I said we could only speculate without input from those whose haplotypes are in question. I just did some research on this.

                  I looked up the 7 supposed R1b1b2h1 guys in Ysearch, then searched for them in FTDNA projects to see if they show SNP results. I found 5 of them in their respective surname projects, and not a single one of them had been SNP tested, at least at FTDNA. If they had been SNP tested by another company, FTDNA's project websites wouldn't show the SNP result.

                  One of the two that I didn't check in the surname projects was tested by Genographic, doesn't have his haplotype displayed, and he wrote in the comments section: "The exact group is HAPLOGROUP R1B (M343)." Sounds to me like someone who hasn't done any further testing beyond his initial Genographic test, and Genographic never tests for subclades.

                  The second guy who I didn't check in a surname project wrote in the comments section that he's R1b1c7 -- as you know, that's previous nomenclature, and is the same as R1b1b2e in the current nomenclature.

                  So based on all this, it's quite apparent to me that the R1b1b2h1 entries in Ysearch are completely erroneous, due to data entry errors, and are not worth any further stress, analysis or confusion.

                  Elise
                  efgen
                  Webinar Coordinator
                  Last edited by efgen; 11 July 2008, 10:35 PM.

                  Comment


                  • #10
                    Originally posted by efgen
                    I just did some research on this...

                    So based on all this, it's quite apparent to me that the R1b1b2h1 entries in Ysearch are completely erroneous, due to data entry errors, and are not worth any further stress, analysis or confusion.

                    Elise
                    Thanks, Elise. I'm certain that there is an element of error in many Y Search profiles, and that conclusions drawn from them can only be tentative in absence of other evidence.

                    However, it can also be the case that Y Search profiles may be more accurate than the projects--after all there is the direct haplotype download feature from FTDNA (less manual intervention).

                    In my own case, project administrators have described me as "R1bic" (sic), and have not posted my SNP results nor indicated that hg has been tested by font colour change to green. Awful nice folks, those administrators, and my undifferentiated clade led me to give up trying to correct them after a few tries. Maybe this was lazy of me, but it didn't seem worth additional effort.

                    So while I agree that there is a possibility of error, I wouldn't conclude that it is account holder just yet. Or that SNP tests haven't been done.

                    Jack
                    P.S. 'Cause I'm replying on blacberry, I can't access YSearch functions to answer Mike's question just yet. It is a good one, though--when I get a chance I wanna run a YCompare on this 67-'bh1 dude vs. Niall modal and b2e members of O'Doherty project.

                    Jack

                    Comment


                    • #11
                      Y Compare results

                      Results:

                      Magee YCompare analysis of ZAKV8, a gentleman named Campbell whose 67-locus haplotype seems to strongly indicate R1b1b2e, but ysearch profile states subclade of R1b1b2h1 (tested). All caveats apply, as Efgen has pointed out, about accuracy of this attribution. Mr. Campbell has declined to respond to an email sent attempting to clarify SNP status.

                      TMRCA, in years, for Mr. Campbell is:
                      570 vs. overall modal
                      510 vs. M5UKQ
                      His result is at least as close to the overal modal as 3 of the 26 O'Dohertys. One of these 3 (i.e., XWYM7) appears to claim very good documentation going back to one of the principal branches of the O'Doherty lords of Inis Eoghain.

                      As a sidebar, M5UKQ is only 180 years from the overall modal. I don't have direct knowledge of details regarding construction of that modal, but there certainly are a large number of individuals with non Ui Neill surnames participating in the R1b1b2e project. This may have implications for consideration of level of convergence of unrelated lines, as touched on in Discussion section.

                      Methodology:

                      Comparing through Magee Y Compare utility ZAKV8 with 26 individuals surnamed O'Doherty or variants and M5UKQ (M222 modal account owned by R1b1b2e project chief, David Wilson). O'Doherty was used as proxy for alternate for Niall modal haplotype, in order to control (to some extent) for possibility that MUK5Q may more broadly represent M222 rather than descendants of Niall specifically. Quite a few individuals of Lowland Scots or even completely unaffiliated English origins appear within the R1b1b2e project.

                      O'Doherty is one of the better documented lines descending from Niall, and in my opinion, one of the least likely to comprehend a large number of unrelated R1b1b2e lines which would dilute the resulting modal haplotype. Although from a different branch of the Northern Ui Neill (i.e., O'Dohertys were descended from Niall's son Conall Gulban, rather than Eoghan), my guess is that that any resulting differences in haplotype between these various branches may be diluted by the long span of time since the inception of these surnames. I think O'Doherty name dates from about 1000 A.D., MacCathmhaoil may date from shortly thereafter.

                      O'Doherty seems to be a surname so closely associated with its traditional homeland of Donegal that I consider it LESS likely that it would attract multiple, unrelated R1b1b2e origins as compared to other, more geographically diverse surnames such as Campbell.


                      Discussion:

                      ZAKV8's published genealogical info states earliest known origin in Ireland, but speculates ultimate origin in Scotland. Could be, but I doubt it. Campbell is well-known as an anglicization of the old Cineal Eoghain surname MacCathmhaoil, chieftains of the territory of Cineal Fhearadhaigh, in south of present day county Tyrone. Results of YCompare analysis discussed above, with TMRCA well within the approx. 1,000 year history of the relevant surnames seem to bear out my hypothesis.

                      The relatively short TMRCA, well within the history of the relevant surnames, amd the possible inclusion of many haplotypes of non-Irish origin in development of MU5KQ, and ZAKV8's stated SNP status of R1b1b2h1 seems strongly suggestive of either or both of the following:

                      1.) Data entry with regard to ZAKV8's SNP designation;
                      2.) A GINORMOUS amount of convergence from unrelated folks. What might that suggest about accuracy or possible significant understatement of published mutation rates?

                      Conclusion:

                      Who knows? Efgen's probably right that someone just feebed out when entering data into the haplogroup field of YSearch profile ZAKV8. But given the 30% R1b1b2e haplotypes listed with R1b1b2h1 I've observed in YSearch, and the unlikely nature of the data entry error (i.e., how confused do you have to be to choose an option 6 subclades away from the correct option?), conclusion can only be tentative.

                      Jack

                      Comment


                      • #12
                        Alexander Surname

                        Forgive me for hijacking this thread. I just received the Y-DNA67 results of my maternal Alexander uncle. He tested R1b1b2e M-222. I'm trying to ascertain which branch he/I may be descended from.

                        You seem so well-versed on surname origins & and I was hoping you might know where the surname Alexander originated?

                        Thank you,
                        Regina
                        RegRiv
                        FTDNA Customer
                        Last edited by RegRiv; 15 July 2008, 10:47 PM.

                        Comment


                        • #13
                          Connacht, Ulster and Scottish Highland haplotypes

                          Originally posted by RegRiv
                          Forgive me for hijacking this thread. I just received the Y-DNA67 results of my maternal Alexander uncle. He tested R1b1b2e M-222. I'm trying to ascertain which branch he/I may be descended from.

                          You seem so well-versed on surname origins & and I was hoping you might know where the surname Alexander originated?

                          Thank you,
                          Regina
                          Regina,

                          I should state that my area of interest has been genealogies and haplotypes of old families of Connacht, Ulster and Scottish Highlands, rather than R1b1b2e specifically. I don't really know too much about SNP's specifically. Although the Niall Modal haplotype is strongly associated with M222 SNP, I don't want to take that relationship for granted. You may consider it significant to know for a fact whether this Alexander has tested positive for M222 SNP or merely predicts that he would based on haplotype.

                          With that caveat in mind, I'll share my thoughts about the Alexander surname and R1b1b2e haplotype.

                          I am aware of only 2 "Gaelic" surnames made into Alexander in English. Neither of them would be associated with R1b1b2e.

                          The first are the MacAlastair branch of the Scottish clan MacDonald. They would be assoicated with a R1a modal haplotype -- so extremely different, especially from an SNP point of view, that most would consider it to be totally UNrelated to R1b1b2e. These families are currently believed to be, ultimately, of Viking origin -- hence the quote marks surrounding my description of Gaelic surnames above.

                          The other are the MacGiollaLaisir family of County Fermanagh in Ireland whom, I believe, much rarer. The only information I have about that family is from O'Hart's Irish Pedigrees, where they are described as a branch of the Maguires. However, the Maguires have a distinctly non-R1b1b2e modal haplotype. And I would take O'Hart with a grain of salt -- he seemed to offer an elaborate pedigree and coat of arms for every surname he encountered. This is dubious when one considers that the continental military model giving rise to the use of coats of arms never made much headway in Gaelic Ireland and Scotland. My sense is that true coats of arms were used only by those "Gaelic" families of deep Norman, Welsh or English origins, or whose unusual regional prominence required some presence in London, Edinburgh or other centers of English or Lowland Scottish political power (i.e., O'Neills, Maguires, etc.).

                          That said, the extreme economic and military upheavals in Ireland and Scotland's history, and the traditions of patronage framed by exaggerated claims of kinshp are well-known. It's entirely conceivable that an individual from one of the classic Gaelic R1b1b2e lineages (i.e., families of the Ui Neill, Ui Bruin or Ui Fiachrach tribes) could have accepted patronage of a better-off individual named Alexander at some time in the remote past. The actual relationship may have been of a landlord-tennant nature or maybe even blood kinship, but in the maternal rather than in the paternal line.

                          I think that of those Alexanders claiming a connection to Ireland, most are from Co. Donegal -- one of the big R1b1b2e hotspots.

                          I am told that the surname Alexander is actually much more commonly of Lowland Scottish origin rather of the MacAlastair or MacGiollaLaisir families mentioned above.

                          I won't claim to be an expert on R1b1b2e outside of the old Gaelic families, but I believe that David Wilson of the M222 project has noted an "R1b1b2e hotspot" in the Western Scottish Lowlands -- Ayr, Galloway region. This seems to make some sense given the deep cultural and economic ties between that area and the Northern Ui Neill homeland of Co. Donegal that probably go back to pre-Roman times. Niall Naoighiallach had to come from somewhere, and it makes sense to me that the Lowland R1b1b2e Alexanders may have shared a common male ancestor with Niall some 1,000's of years ago.

                          Here is a link to M222 website.

                          http://www.familytreedna.com/public/R1b1c7


                          And that's pretty much all I know about that.

                          Jack

                          Comment


                          • #14
                            Thank you

                            Thank you, Jack! That was so very enlightening. Especially explaining the different regions. And I was suspicious that a name change may have occurred, and I now see that is indeed plausible. (So far, I am not very closely matching any Alexander's on the FTDNA project.) Your note offers clues on where to look for further info though.

                            Regarding more ancient ancestry, our family's autosomal DNA indicate heavy Norway & Iceland hits, as compared to Irish or Scots.

                            This is all fascinating & confusing at the same time, for someone like me with absolutely no science or history background. So, I'm especially appreciative of your taking the time to enlighten me.

                            Thanks again,
                            Regina

                            Comment


                            • #15
                              Pattern is emerging

                              Originally posted by Clochaire
                              #1...looking at R1b1b2h project results, it looks like there are a number of individuals in this hg who share some of (what appeared to me) the stranger DYS values reported for me: 26/14/18 at 447/437/448.

                              #2. Maybe this fellow I discussed above, with my surname, and R1b1b2e haplotype but with odd w/ self-reported R1b1b2h1 does have some significance with regard to my discordant results discussed elsewhere.

                              Jack
                              Thinking out-loud again, but recent developments show an emerging pattern of a WIDER pool of discordant results.

                              #1. Come to find out that some of the near matches, sharing weird 447/437/448 motif, on my personal page have tested positive as R1b1b2h. I would hypotheize that motif may represent indicator of an English East Midlands cluster for R1b1b2h.

                              #2. This crypto-R1b1b2e w/ R1b1b2h SNP fellow does indeed have some significance for interpretation of my own results.

                              I have received a response from him. It does address some aspects of his own patriline with relevance for assessing his possible relationship w/ other supposed R1b1b2e persons in our surname project.

                              But he seemed almost evasive with regards to specific SNP status. I'll have to be diplomatic in follow up. My impression is that he is upset about apparent discrepancy between R1b1b2e prediction and R1b1b2h SNP.

                              He seems to have some gaps in paper trail (surprise). But the consistent weight of evidence currently available suggests that his family migrated from N.Central Kentucky, to Marion Co., M.O. And adjacent co's of I.L., and finally to Texas.

                              This is uncannily similar to the pattern for 3 other supposed R1b1b2e participants, who provide additional information of earliest known ancestor in colonial Stafford Co., Virginia.

                              However, of these 4 supposed R1b1b2e individuals, Magee's YDNA Compare utility suggests that only 2 have a common ancestor in last 600 years. Does this make sense considering relative rarity of R1b1b2e in English population? 2 pct occurence is quote I've heard.

                              Still have to work on this in order to tie a bow on it. But it's starting to seem that some critical hypotheses may be more likely than earlier thought.

                              Comment

                              Working...
                              X