Loading ...
Sorry, an error occurred while loading the content.
 

Re: GFather/Father/Son DNA?

Expand Messages
  • Debra Dunbar Nowell
    Paul, Getting your father and son tested should be determined on what kind of information you re looking for and the amount of money your willing to spend to
    Message 1 of 27 , Jun 1 4:34 AM
      Paul,

      Getting your father and son tested should be determined on what kind
      of information you're looking for and the amount of money your willing
      to spend to find out. As David says, it is possible that all three of
      you would match exactly or reveal an "oops" event. However, I have
      two very good examples of the benefit of getting them tested.

      Terry Barton, founder of WorldFamilies has tested both and found that
      both he and his father started a mutation. Terry is proud to say that
      his son in not a "mutant". All three have tested 105 markers each.

      Within my own Dunbar Project, we have a father/son team. The son has
      a mutation different from his father on DYS 390 from 24 to 25. This
      is one of the very slow moving markers which they say only mutates
      every 400 to 500 years. (I hope I have the # years right!)

      Debra



      >
      >
      > I am thinking of having my father and son sequenced for 67 markers.
      Is this worthwhile? I guess it would help in the analysis of
      mutations if there were any? Presumably they would be the same as mine.
      >
      > Paul Roche
      >
      >
      >
    • Lyn McMullen
      Debra, that is a very important piece of information, and could have implications for many of the project assumptions. Where exactly does your information
      Message 2 of 27 , Jun 2 10:00 PM

        Debra, that is a very important piece of information, and could have implications for many of the project assumptions.  Where exactly does your information about the average years for DYS 390 (400-500) come from?  Lyn

      • David Weston
        Lyn, You ll find an academic paper with a commonly accepted derivation of the mutation rates for FTDNA s 37-marker panel here:
        Message 3 of 27 , Jun 3 3:30 AM
          Lyn,
           
          You'll find an academic paper with a commonly accepted derivation of the mutation rates for FTDNA's 37-marker panel here:
           
           
          In this derivation DYS390 has an average mutation rate of about 0.00311 or 1 in 321 generations (~9,000 yrs).  This compares with the slowest marker in the 37-marker panel DYS426, which is given an average rate of 0.00009 (1 in 11,111 generations), and the fastest CDY, which is 0.03531 (1 in 28 generations).  It is important to note that these are entirely random events.  You could find the same marker mutating two generations in a row and never see it mutate again for 1,000.  All the rates say is that average occurence of such events over infinite number of generations.  If you took 321 father/son pairs, odds are one of them would have a difference of 1 on DYS390.  Or, you could pick another set of 321 pairs and find 10 with difference of 1 at this marker.  The derivation of mutation rates are based on averaging of many 100s of mutations.
           
          Individual marker mutations rates are much less important than the overall combined mutation rate for the haplotype.  That is what matters most, for example, is the chance of 1 mutation on 37-markers not which marker actually mutates.  All algoriths for calculation Time to Most Recent Common Ancestor work on this basis.
           
          Cheers, David.


          From: R1b1c_U106-S21@yahoogroups.com [mailto:R1b1c_U106-S21@yahoogroups.com] On Behalf Of Lyn McMullen
          Sent: June 3, 2008 2:01 AM
          To: R1b1c_U106-S21@yahoogroups.com
          Subject: [R1b1c_U106-S21] Re: GFather/Father/Son DNA?

          Debra, that is a very important piece of information, and could have implications for many of the project assumptions.  Where exactly does your information about the average years for DYS 390 (400-500) come from?  Lyn

        • Debra Dunbar Nowell
          I was afraid someone would ask that question. Went digging through my paperwork and found the information posted at my FTDNA Group Administrator Page under
          Message 4 of 27 , Jun 3 6:39 AM
            I was afraid someone would ask that question.  Went digging through my paperwork and found the information posted at my FTDNA Group Administrator Page under "Gap Interpretation".  Here is the info:

            8. I hear about `fast moving markers'…what does that mean?

            The current estimate by the Anthropological community of the mutation rate of Y chromosomal markers used for genealogy is .002, which means that 1 mutation is expected to occur 1 time, per marker, every 500 generations. While this number may be accurate for unrelated males within a population it appears to understate the actual mutation rate when comparisons are made from within a family. Therefore we highlight markers which appear to be more volatile in red, as you can see when you click the GENERATE page from within your GAP. The volatility rate of these markers hasn't been established. We believe that a standard rate of change, across the entire panel, is not likely either.

            If you have 2 people who match, exactly, except on a single marker and that marker is one that is highlighted in RED within your surname group then the current estimate of distance between 2 people are related is probably overstated, and they are more closely related than could be assumed by the distance suggested by a standard single marker deviation.

            The father/son team, in the Dunbar Project, have both SNP'd.  The son is currently #106 in the R1b1c-U106 Haplogroup Project.  We are awaiting his father's SNP results and he will join this group once the test comes back from the lab.  Their surname is Cockburn. 

            Hope this helps, Debra


            --- In R1b1c_U106-S21@yahoogroups.com, Lyn McMullen <ldminc1@...> wrote:
            >
            > Debra, that is a very important piece of information, and could have
            > implications for many of the project assumptions. Where exactly does your
            > information about the average years for DYS 390 (400-500) come from? Lyn
            >
          • Raymond Wing
            Back in July 2007, a member of my family [Wing] received SNP results from EthnoAncestry indicating he was S21+. He was able to join the R1b-S21 Project based
            Message 5 of 27 , Jun 6 2:33 PM
              Back in July 2007, a member of my family [Wing] received SNP results from EthnoAncestry indicating he was S21+. He was able to join the R1b-S21 Project based on these results.

              This same family member had recently had a SNP test completed by Family Tree DNA. The results (which were initially expected back on May 5th) came through today and indicated he was U106- (U106 being identical to S21).

              As the Genealogist for the family association (as well as the Y-DNA Project administrator), I contacted both companies, explaining this discrepancy. FT DNA indicated they have recently had several cases of this happening, and agreed to re-test this individual's SNP (free of charge, by hand). I have also contacted EthnoAncestry to see if they would also be willing to retest the DNA sample.

              I wanted to post this notice on this group to let people know that there still appears to be some issue somewhere regarding the testing of S21/U106. 

              Using an on-line Haplotype predictor, (http://members.bex.net/jtcullen515/haplotest.htm)
              the family's 67 markers show a fairly high probability (65%) of being R1b-S21, with the other predictions being R1b-Frisian3 (24%) and R1b-Frisian (10%).  It was my understanding (and I could be wrong) that these Frisian R1bs are also R1b-S21. (If so, then this haplotype predictor is 99% certain that the family is R1b-S21)

              Raymond T. Wing
            • David Weston
              Raymond, Well, this is an unfortunate turn of events. I am aware of several discrepancies between EA and FTDNA SNP results for U106/S21. Two that have been
              Message 6 of 27 , Jun 6 3:25 PM
                Raymond,
                 
                Well, this is an unfortunate turn of events.  I am aware of several discrepancies between EA and FTDNA SNP results for U106/S21.  Two that have been resolved came out in favour of FTDNA.  I have no grounds (yet) on which to doubt FTDNA's SNP test result.
                 
                The WING haplotype with DYS492 = 13 does appear to be U106+ on face value.  However, although it would be unusual for a U106- result with DYS492 = 13 it is not unheard of.  I know of at least four (assuming the SNP test results are correct).  It will be very interesting indeed to see what FTDNA comes back with on the recheck.  Please let me know directly as soon as you find out.
                 
                I should add that with the exception of DYS492=13, U106+ status cannot be reliably predicted.  I have no faith what so ever in the reliabilit of any attempts to make such a prediction.
                 
                David Weston.
                 


                From: R1b1c_U106-S21@yahoogroups.com [mailto:R1b1c_U106-S21@yahoogroups.com] On Behalf Of Raymond Wing
                Sent: June 6, 2008 6:34 PM
                To: R1b1c_U106-S21@yahoogroups.com
                Subject: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results

                Back in July 2007, a member of my family [Wing] received SNP results from EthnoAncestry indicating he was S21+. He was able to join the R1b-S21 Project based on these results.

                This same family member had recently had a SNP test completed by Family Tree DNA. The results (which were initially expected back on May 5th) came through today and indicated he was U106- (U106 being identical to S21).

                As the Genealogist for the family association (as well as the Y-DNA Project administrator) , I contacted both companies, explaining this discrepancy. FT DNA indicated they have recently had several cases of this happening, and agreed to re-test this individual's SNP (free of charge, by hand). I have also contacted EthnoAncestry to see if they would also be willing to retest the DNA sample.

                I wanted to post this notice on this group to let people know that there still appears to be some issue somewhere regarding the testing of S21/U106. 

                Using an on-line Haplotype predictor, (http://members. bex.net/jtcullen 515/haplotest. htm)
                the family's 67 markers show a fairly high probability (65%) of being R1b-S21, with the other predictions being R1b-Frisian3 (24%) and R1b-Frisian (10%).  It was my understanding (and I could be wrong) that these Frisian R1bs are also R1b-S21. (If so, then this haplotype predictor is 99% certain that the family is R1b-S21)

                Raymond T. Wing

              • Justin Loe
                For the record, I ve been tested for U106/S21 with both Ethnoancestry and FTDNA, with the same result: S21+. Perhaps FTDNA is running into some quality control
                Message 7 of 27 , Jun 6 3:48 PM
                  For the record, I've been tested for U106/S21 with both Ethnoancestry and FTDNA, with the same result: S21+. Perhaps FTDNA is running into some quality control problems with their SNP test, we shall see. Ysearch = HUGNS.


                  To: R1b1c_U106-S21@yahoogroups.com
                  From: westons@...
                  Date: Fri, 6 Jun 2008 19:25:31 -0300
                  Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results


                  Raymond,
                   
                  Well, this is an unfortunate turn of events.  I am aware of several discrepancies between EA and FTDNA SNP results for U106/S21.  Two that have been resolved came out in favour of FTDNA.  I have no grounds (yet) on which to doubt FTDNA's SNP test result.
                   
                  The WING haplotype with DYS492 = 13 does appear to be U106+ on face value.  However, although it would be unusual for a U106- result with DYS492 = 13 it is not unheard of.  I know of at least four (assuming the SNP test results are correct).  It will be very interesting indeed to see what FTDNA comes back with on the recheck.  Please let me know directly as soon as you find out.
                   
                  I should add that with the exception of DYS492=13, U106+ status cannot be reliably predicted.  I have no faith what so ever in the reliabilit of any attempts to make such a prediction.
                   
                  David Weston.
                   


                  From: R1b1c_U106-S21@ yahoogroups. com [mailto:R1b1c_ U106-S21@ yahoogroups. com] On Behalf Of Raymond Wing
                  Sent: June 6, 2008 6:34 PM
                  To: R1b1c_U106-S21@ yahoogroups. com
                  Subject: [R1b1c_U106- S21] Discrepency between EthnoAncestry and Family Tree DNA results


                  Back in July 2007, a member of my family [Wing] received SNP results from EthnoAncestry indicating he was S21+. He was able to join the R1b-S21 Project based on these results.

                  This same family member had recently had a SNP test completed by Family Tree DNA. The results (which were initially expected back on May 5th) came through today and indicated he was U106- (U106 being identical to S21).

                  As the Genealogist for the family association (as well as the Y-DNA Project administrator) , I contacted both companies, explaining this discrepancy. FT DNA indicated they have recently had several cases of this happening, and agreed to re-test this individual's SNP (free of charge, by hand). I have also contacted EthnoAncestry to see if they would also be willing to retest the DNA sample.

                  I wanted to post this notice on this group to let people know that there still appears to be some issue somewhere regarding the testing of S21/U106. 

                  Using an on-line Haplotype predictor, (http://members. bex.net/jtcullen 515/haplotest. htm)
                  the family's 67 markers show a fairly high probability (65%) of being R1b-S21, with the other predictions being R1b-Frisian3 (24%) and R1b-Frisian (10%).  It was my understanding (and I could be wrong) that these Frisian R1bs are also R1b-S21. (If so, then this haplotype predictor is 99% certain that the family is R1b-S21)

                  Raymond T. Wing
                • Thomas Gull
                  David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13? I ve seen some in other major haplogroups with DYS 492=13 but so far none
                  Message 8 of 27 , Jun 6 5:38 PM
                    David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13?  I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before.  Thks / Tom


                    To: R1b1c_U106-S21@yahoogroups.com
                    From: westons@...
                    Date: Fri, 6 Jun 2008 19:25:31 -0300
                    Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results

                    Raymond,
                     
                    Well, this is an unfortunate turn of events.  I am aware of several discrepancies between EA and FTDNA SNP results for U106/S21.  Two that have been resolved came out in favour of FTDNA.  I have no grounds (yet) on which to doubt FTDNA's SNP test result.
                     
                    The WING haplotype with DYS492 = 13 does appear to be U106+ on face value.  However, although it would be unusual for a U106- result with DYS492 = 13 it is not unheard of.  I know of at least four (assuming the SNP test results are correct).  It will be very interesting indeed to see what FTDNA comes back with on the recheck.  Please let me know directly as soon as you find out.
                     
                    I should add that with the exception of DYS492=13, U106+ status cannot be reliably predicted.  I have no faith what so ever in the reliabilit of any attempts to make such a prediction.
                     
                    David Weston.
                     


                    From: R1b1c_U106-S21@ yahoogroups. com [mailto:R1b1c_ U106-S21@ yahoogroups. com] On Behalf Of Raymond Wing
                    Sent: June 6, 2008 6:34 PM
                    To: R1b1c_U106-S21@ yahoogroups. com
                    Subject: [R1b1c_U106- S21] Discrepency between EthnoAncestry and Family Tree DNA results

                    Back in July 2007, a member of my family [Wing] received SNP results from EthnoAncestry indicating he was S21+. He was able to join the R1b-S21 Project based on these results.

                    This same family member had recently had a SNP test completed by Family Tree DNA. The results (which were initially expected back on May 5th) came through today and indicated he was U106- (U106 being identical to S21).

                    As the Genealogist for the family association (as well as the Y-DNA Project administrator) , I contacted both companies, explaining this discrepancy. FT DNA indicated they have recently had several cases of this happening, and agreed to re-test this individual's SNP (free of charge, by hand). I have also contacted EthnoAncestry to see if they would also be willing to retest the DNA sample.

                    I wanted to post this notice on this group to let people know that there still appears to be some issue somewhere regarding the testing of S21/U106. 

                    Using an on-line Haplotype predictor, (http://members. bex.net/jtcullen 515/haplotest. htm)
                    the family's 67 markers show a fairly high probability (65%) of being R1b-S21, with the other predictions being R1b-Frisian3 (24%) and R1b-Frisian (10%).  It was my understanding (and I could be wrong) that these Frisian R1bs are also R1b-S21. (If so, then this haplotype predictor is 99% certain that the family is R1b-S21)

                    Raymond T. Wing


                  • David Weston
                    Tom, YSearch IDs: - from David Faux s U152/S28+ project: 8VSEP
                    Message 9 of 27 , Jun 6 7:26 PM
                      Tom,
                       
                      YSearch IDs:
                       
                       
                       
                      - from John McEwan's R1b SNP Summary Page, both R1b-M269*:
                       
                       
                      There are bound to be others that have not been SNP tested.  I anticipate that we will find on order of 2-5% of DYS492 = 13 haplotypes will be U106- once we are able to identify the R1b-M269* haplotypes in the FTDNA tables and more new Deep-R SNP test results come in.   The four above are from a collection of 238 67-marker U106/S21- haplotypes or 1.6%.  I recommend you not include non-SNP tested haplotypes in your analysis for this reason.  You risk throwing off any conclusions you make about U106+ otherwise.
                       
                      Cheers, David.


                      ________________________________

                      From: R1b1c_U106-S21@yahoogroups.com [mailto:R1b1c_U106-S21@yahoogroups.com] On Behalf Of Thomas Gull
                      Sent: June 6, 2008 9:39 PM
                      To: r1b1c_u106-s21@yahoogroups.com
                      Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results



                      David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13?  I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before.  Thks / Tom
                    • David Weston
                      We have expanded the scope of the research group to include confirmed U198/S29+ and L1/S26+ (Null 439) haplotypes in the individual allele statistics and
                      Message 10 of 27 , Jun 8 5:01 PM
                        We have expanded the scope of the research group to include confirmed U198/S29+ and L1/S26+ (Null 439) haplotypes in the individual allele statistics and European geographic distribution from the S29 Y-DNA Project and Null 439 Project..  Invidividual participation in the research group will still be limited to those confirmed U106/S21+, U198/S29-, L1/S26- (non-Null 439).   U198/S29+ or L1/S26+ (Null 439) individuals are encouraged to add their results to the S29 Y-DNA Project and Null 439 Project, respectively. 

                        David Weston
                        Group Administrator
                        R1b-U106/S21+ Research Group


                        From: R1b1c_U106-S21@yahoogroups.com [mailto:R1b1c_U106-S21@yahoogroups.com] On Behalf Of germyn
                        Sent: May 30, 2008 2:30 PM
                        To: R1b1c_U106-S21@yahoogroups.com
                        Subject: [R1b1c_U106-S21] Why exclude subclades?

                        Quoting from the goals of the U106 project site at FTDNA:
                        "This project was started to create a repository for YDNA haplotypes
                        belonging to the haplogroup R1b-U106/S21 (formerly R1b1c9*), i.e.
                        confirmed U106+, and L1- (non-null 439) and U198-."

                        Why are the subclades excluded? They too are U106+ and I would have
                        expected their geography etc to be as much a part of the U106+ picture
                        as those having no subclades. By not counting the S29+ in the north is
                        there not a risk of creating the appearance of more U106 data points
                        in the south? If new subclades are discovered, will those entries in
                        the parent clade data be removed? Even though the numbers are still
                        greater than those of the subclades, it seems like the data will only
                        reflect the statistics of the "left-overs" .

                      • Raymond Wing
                        Tom, I have checked the new rs34276300+ (S116) project at FT DNA http://www.familytreedna.com/public/atlantic-r1b1c According to their Y Results, out of 129
                        Message 11 of 27 , Jun 9 2:55 AM
                          Tom,

                          I have checked the new rs34276300+ (S116) project at FT DNA http://www.familytreedna.com/public/atlantic-r1b1c

                          According to their Y Results, out of 129 members currently in the project, 4 of them have DYS492=13.

                          It does seem that some folk who are R1b and have DYS492=13 are not U106/S21+.

                          Ray


                          --- On Fri, 6/6/08, Thomas Gull <tmrgull@...> wrote:
                          From: Thomas Gull <tmrgull@...>
                          Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results
                          To: r1b1c_u106-s21@yahoogroups.com
                          Date: Friday, June 6, 2008, 8:38 PM

                          David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13?  I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before.  Thks / Tom

                        • bob_bootle
                          Ray, I was advised ( by the late Leo Little ) that about 95% of S21 are DYS492=13. Your 4 in 129 is 3.1%, so you might expect to get a few more soon. Regards
                          Message 12 of 27 , Jun 9 4:22 AM
                            Ray,

                            I was advised ( by the late Leo Little ) that about 95% of S21 are
                            DYS492=13.

                            Your 4 in 129 is 3.1%, so you might expect to get a few more soon.

                            Regards

                            Bob Bootle



                            --- In R1b1c_U106-S21@yahoogroups.com, Raymond Wing
                            <wing_genealogist@...> wrote:
                            >
                            > Tom,
                            >
                            > I have checked the new rs34276300+ (S116) project at FT DNA
                            http://www.familytreedna.com/public/atlantic-r1b1c
                            >
                            > According to their Y Results, out of 129 members currently in the
                            project, 4 of them have DYS492=13.
                            >
                            > It does seem that some folk who are R1b and have DYS492=13 are not
                            U106/S21+.
                            >
                            > Ray
                            >
                            >
                            > --- On Fri, 6/6/08, Thomas Gull tmrgull@... wrote:
                            > From: Thomas Gull tmrgull@...
                            > Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and
                            Family Tree DNA results
                            > To: r1b1c_u106-s21@yahoogroups.com
                            > Date: Friday, June 6, 2008, 8:38 PM
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            > David, are you saying you have IDs for some U106- R1b individuals
                            with DYS 492=13?  I've seen some in other major haplogroups with
                            DYS 492=13 but so far none in R1b. If their IDs are public,
                            could you send me a list? I'd like to make sure I didn't
                            miscategorize them if I have seen them before.  Thks / Tom
                            >
                          • Thomas Gull
                            Excellent, thanks. The picture is never completely simple, is it? / Tom To: R1b1c_U106-S21@yahoogroups.comFrom: wing_genealogist@yahoo.comDate: Mon, 9 Jun
                            Message 13 of 27 , Jun 9 2:00 PM
                              Excellent, thanks.  The picture is never completely simple, is it?  / Tom


                              To: R1b1c_U106-S21@yahoogroups.com
                              From: wing_genealogist@...
                              Date: Mon, 9 Jun 2008 02:55:28 -0700
                              Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results

                              Tom,

                              I have checked the new rs34276300+ (S116) project at FT DNA http://www.familytr eedna.com/ public/atlantic- r1b1c

                              According to their Y Results, out of 129 members currently in the project, 4 of them have DYS492=13.

                              It does seem that some folk who are R1b and have DYS492=13 are not U106/S21+.

                              Ray


                              --- On Fri, 6/6/08, Thomas Gull <tmrgull@msn. com> wrote:
                              From: Thomas Gull <tmrgull@msn. com>
                              Subject: RE: [R1b1c_U106- S21] Discrepency between EthnoAncestry and Family Tree DNA results
                              To: r1b1c_u106-s21@ yahoogroups. com
                              Date: Friday, June 6, 2008, 8:38 PM

                              David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13?  I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before.  Thks / Tom

                            • Thomas Gull
                              Bob, I think U106 is all S116-, isn t it? That is, if I have that correct and these four are S116+, it s the first time DYS 492=13 has been pointed out within
                              Message 14 of 27 , Jun 9 2:04 PM
                                Bob, I think U106 is all S116-, isn't it?  That is, if I have that correct and these four are S116+, it's the first time DYS 492=13 has been pointed out within R1b but not in U106.  I've seen it in some I and Q haplotypes before, but not in R1bs.  But it's not like I expected that to hold up forever...
                                 
                                So anyway, I think both Ray and David are finding the R1bs who are DYS 492=13 but who aren't U106+.  Did I interpret Ray's message incorrectly? / Tom


                                To: R1b1c_U106-S21@yahoogroups.com
                                From: r.bootle@...
                                Date: Mon, 9 Jun 2008 11:22:01 +0000
                                Subject: [R1b1c_U106-S21] Re: Discrepency between EthnoAncestry and Family Tree DNA results

                                Ray,

                                I was advised ( by the late Leo Little ) that about 95% of S21 are
                                DYS492=13.

                                Your 4 in 129 is 3.1%, so you might expect to get a few more soon.

                                Regards

                                Bob Bootle

                                --- In R1b1c_U106-S21@ yahoogroups. com, Raymond Wing
                                <wing_genealogist@ ...> wrote:
                                >
                                > Tom,
                                >
                                > I have checked the new rs34276300+ (S116) project at FT DNA
                                http://www.familytr eedna.com/ public/atlantic- r1b1c
                                >
                                > According to their Y Results, out of 129 members currently in the
                                project, 4 of them have DYS492=13.
                                >
                                > It does seem that some folk who are R1b and have DYS492=13 are not
                                U106/S21+.
                                >
                                > Ray
                                >
                                >
                                > --- On Fri, 6/6/08, Thomas Gull tmrgull@... wrote:
                                > From: Thomas Gull tmrgull@...
                                > Subject: RE: [R1b1c_U106- S21] Discrepency between EthnoAncestry and
                                Family Tree DNA results
                                > To: r1b1c_u106-s21@ yahoogroups. com
                                > Date: Friday, June 6, 2008, 8:38 PM
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                >
                                > David, are you saying you have IDs for some U106- R1b individuals
                                with DYS 492=13?&nbsp; I've seen some in other major haplogroups with
                                DYS 492=13 but so far none in R1b.&nbsp;If their IDs are public,
                                could you send me a list? I'd like to make sure I didn't
                                miscategorize them if I have seen them before.&nbsp; Thks / Tom
                                >


                              • Thomas Gull
                                Thanks, David. If the results are only off by 5% or so, I d still look at DYS 492=13 as a strong indicator of possibilities, but not as a lock, as you suggest.
                                Message 15 of 27 , Jun 9 2:06 PM
                                  Thanks, David. If the results are only off by 5% or so, I'd still look at DYS 492=13 as a strong indicator of possibilities, but not as a lock, as you suggest.  I actually didn't expect it to be this useful at the outset, even.  / Tom


                                  To: R1b1c_U106-S21@yahoogroups.com
                                  From: westons@...
                                  Date: Fri, 6 Jun 2008 23:26:14 -0300
                                  Subject: [R1b1c_U106-S21] DYS492 = 13, U106-

                                  Tom,
                                   
                                  YSearch IDs:
                                   
                                   
                                   
                                  - from John McEwan's R1b SNP Summary Page, both R1b-M269*:
                                   
                                   
                                  There are bound to be others that have not been SNP tested.  I anticipate that we will find on order of 2-5% of DYS492 = 13 haplotypes will be U106- once we are able to identify the R1b-M269* haplotypes in the FTDNA tables and more new Deep-R SNP test results come in.   The four above are from a collection of 238 67-marker U106/S21- haplotypes or 1.6%.  I recommend you not include non-SNP tested haplotypes in your analysis for this reason.  You risk throwing off any conclusions you make about U106+ otherwise.
                                   
                                  Cheers, David.


                                  ____________ _________ _________ __

                                  From: R1b1c_U106-S21@ yahoogroups. com [mailto:R1b1c_ U106-S21@ yahoogroups. com] On Behalf Of Thomas Gull
                                  Sent: June 6, 2008 9:39 PM
                                  To: r1b1c_u106-s21@ yahoogroups. com
                                  Subject: RE: [R1b1c_U106- S21] Discrepency between EthnoAncestry and Family Tree DNA results



                                  David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13?  I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before.  Thks / Tom


                                • Raymond Wing
                                  Tom, You understood my message correctly.  I realized S116+ meant those individuals were NOT U106/S21+.  It does appear that DYS492=13 is quite rare
                                  Message 16 of 27 , Jun 9 2:59 PM
                                    Tom,

                                    You understood my message correctly.  I realized S116+ meant those individuals were NOT U106/S21+. 

                                    It does appear that DYS492=13 is quite rare for the other R1b subclades. The percentages I am seeing are under 5% of the population.

                                    It'll be interesting (for me anyways) to learn whether or not I am U106/S21 + or -.  [I wasn't personally tested, but a cousin of mine was. Nothing really would be learned by getting SNP tested as well.]

                                    Ray


                                    --- On Mon, 6/9/08, Thomas Gull <tmrgull@...> wrote:
                                    From: Thomas Gull <tmrgull@...>
                                    Subject: RE: [R1b1c_U106-S21] Re: Discrepency between EthnoAncestry and Family Tree DNA results
                                    To: r1b1c_u106-s21@yahoogroups.com
                                    Date: Monday, June 9, 2008, 5:04 PM

                                    Bob, I think U106 is all S116-, isn't it?  That is, if I have that correct and these four are S116+, it's the first time DYS 492=13 has been pointed out within R1b but not in U106.  I've seen it in some I and Q haplotypes before, but not in R1bs.  But it's not like I expected that to hold up forever...
                                     
                                    So anyway, I think both Ray and David are finding the R1bs who are DYS 492=13 but who aren't U106+.  Did I interpret Ray's message incorrectly? / Tom

                                  • Thomas Gull
                                    Back from vacation, David, and catching up on loose ends. I m glad I asked for the details on DYS 492=13 exceptions. I had 8VSEP down as U106+ from an old
                                    Message 17 of 27 , Jun 13 10:27 AM
                                      Back from vacation, David, and catching up on loose ends.

                                      I'm glad I asked for the details on DYS 492=13 exceptions. I had 8VSEP down as U106+ from an old project spreadsheet. He shows as U152+ now across the board. There's one who went the other way - 6KPR2 is shown as U106+ in Ysearch now.

                                      ZGFUS is the same as 49846 in the S116+ project.

                                      So between your message and Ray's, we have the following as DYS 492=13 but not U106+ in R1b:
                                      8VSEP = U152+
                                      Q6V6Y = U152+
                                      ZGFUS = 49846 = S116+
                                      N17254 = S116+
                                      110656 = same ancestor as N47554 = S116+

                                      I built a small table showing the DYS 492 values in three projects (U106+, S116+, David Faux's U152+). These represent known quantities where an SNP was tested and DYS 492 was tested, so it seems like if there is any predictive value to this marker, it should potentially show up in this table. It has been 34 years since my last statistics course, so I'm not sure what measure best expresses any underlying relation. I was thinking the Pearson's R and came up with a value of around -.676 for that if I set it up for "U106" and "not-U106". Ken, if you get the time, can you maybe comment on the table? To me, 13 repeats looks highly predictive for U106+ and 12 looks highly predictive for the S116+ branch.

                                      DYS 492 U106+ U152+ S116+
                                      11 0 3 2
                                      12 3 73 125
                                      13 160 2 3
                                      14 6 15 2


                                      / Tom

                                      ______________________
                                      i'm EMAILING FOR THE GREATER GOOD
                                      http://im.live.com/Messenger/IM/Home/?source=EML_WLHM_GreaterGood




                                      ________________________________

                                      To: R1b1c_U106-S21@yahoogroups.com
                                      From: westons@...
                                      Date: Fri, 6 Jun 2008 23:26:14 -0300
                                      Subject: [R1b1c_U106-S21] DYS492 = 13, U106-







                                      Tom,



                                      YSearch IDs:



                                      - from David Faux's U152/S28+ project:



                                      8VSEP

                                      Q6V6Y



                                      - from John McEwan's R1b SNP Summary Page, both R1b-M269*:



                                      6KPR2

                                      ZGFUS



                                      There are bound to be others that have not been SNP tested. I anticipate that we will find on order of 2-5% of DYS492 = 13 haplotypes will be U106- once we are able to identify the R1b-M269* haplotypes in the FTDNA tables and more new Deep-R SNP test results come in. The four above are from a collection of 238 67-marker U106/S21- haplotypes or 1.6%. I recommend you not include non-SNP tested haplotypes in your analysis for this reason. You risk throwing off any conclusions you make about U106+ otherwise.



                                      Cheers, David.



                                      ________________________________

                                      From: R1b1c_U106-S21@yahoogroups.com [mailto:R1b1c_U106-S21@yahoogroups.com] On Behalf Of Thomas Gull
                                      Sent: June 6, 2008 9:39 PM
                                      To: r1b1c_u106-s21@yahoogroups.com
                                      Subject: RE: [R1b1c_U106-S21] Discrepency between EthnoAncestry and Family Tree DNA results



                                      David, are you saying you have IDs for some U106- R1b individuals with DYS 492=13? I've seen some in other major haplogroups with DYS 492=13 but so far none in R1b. If their IDs are public, could you send me a list? I'd like to make sure I didn't miscategorize them if I have seen them before. Thks / Tom
                                    • Raymond Wing
                                      Does anyone know if those few individuals whose DYS492 value doesn t match the usual SNP Modal (DYS492=13 for U106/S21+; DYS492=12 or 14 for U106/S21-) have
                                      Message 18 of 27 , Jun 13 2:36 PM
                                        Does anyone know if those few individuals whose DYS492 value doesn't match the "usual" SNP Modal (DYS492=13 for U106/S21+; DYS492=12 or 14 for U106/S21-) have had their SNP test rechecked by their original testing company? 

                                        (In other words, has anyone whose DYS492=13 but SNP U106/S21- or DYS492=12 or 14 but SNP U106/S21+ had their SNP test rechecked)

                                        I don't doubt there are some outliers whose DYS492 doesn't match their SNP. However, it is also clear there have been a SMALL number of cases where persons (like my cousin) have received conflicting results from two DNA testing companies regarding SNP U106/S21. 

                                        I would imagine the DNA testing companies would not be the most open to discussing their error rates (which I sincerely hope/believe is quite low). Given the thousands of tests performed, such errors are probably inevitable.

                                        Raymond T. Wing
                                        Genealogist, Wing Family of America, Inc. [WFA]
                                        www.wingfamily.org_._,___
                                      • adam bradford
                                        I am DYS492=14 and also S21 positive according to both FTDNA and EthnoAncestry. Some (perhaps most) of the S21/492=14 people are going to prove to part of a
                                        Message 19 of 27 , Jun 13 2:41 PM
                                          I am DYS492=14 and also S21 positive according to both FTDNA and EthnoAncestry.

                                          Some (perhaps most) of the S21/492=14 people are going to prove to part of a little subcluster of the Frisian cluster that is also characterized by 464d=19. 


                                          On Fri, Jun 13, 2008 at 5:36 PM, Raymond Wing <wing_genealogist@...> wrote:

                                          Does anyone know if those few individuals whose DYS492 value doesn't match the "usual" SNP Modal (DYS492=13 for U106/S21+; DYS492=12 or 14 for U106/S21-) have had their SNP test rechecked by their original testing company? 

                                          (In other words, has anyone whose DYS492=13 but SNP U106/S21- or DYS492=12 or 14 but SNP U106/S21+ had their SNP test rechecked)

                                          I don't doubt there are some outliers whose DYS492 doesn't match their SNP. However, it is also clear there have been a SMALL number of cases where persons (like my cousin) have received conflicting results from two DNA testing companies regarding SNP U106/S21. 

                                          I would imagine the DNA testing companies would not be the most open to discussing their error rates (which I sincerely hope/believe is quite low). Given the thousands of tests performed, such errors are probably inevitable.

                                          Raymond T. Wing
                                          Genealogist, Wing Family of America, Inc. [WFA]
                                          www.wingfamily.org_._,___

                                        • Thomas Gull
                                          Adam, per Ken Nordtvedt s quote just below, 464d would have 18 repeats as a mark of inclusion in the R1bSTR22 Frisian cluster... Are you maybe thinking the
                                          Message 20 of 27 , Jun 14 3:08 PM
                                            Adam, per Ken Nordtvedt's quote just below, 464d would have 18 repeats as a mark of inclusion in the R1bSTR22 "Frisian" cluster... Are you maybe thinking the larger set of STRs with DYS390 = 23?

                                            Ken in a recent email to me: "Do you mean the clade with 24 at 447, 10 at H4, 18 at 464d, 23 at 390, 11 at 391? That's what I consider the Frisian clade...This clade has 13 at 487 and is S21+ --- although certainly not all of S21+ by a long shot."

                                            On a related topic in this thread, I listed some raw numbers earlier for DYS 492 = 13 and its potential value as a predictive value for U106+ and family (S29, etc). Here's a quote from a later message I sent to three people on that topic:

                                            "I guess one way to look at it is that if you adopted a simple 'DYS 492 value of 13 predicts U106+, all other 492s don't' hypothesis, then in this sample... [U106 project, U152 David Faux project, S116+ project]

                                            You would predict 165 U106+ results and 229 non-U106+ results.
                                            Your predictions would be correct in 380 cases, falsely positive in 5 cases, and falsely negative in 9 cases.
                                            False positive would be a prediction someone was U106+ when they were S116+ or U152+.
                                            False negative would be a prediction someone was not U106+ when in fact they were.

                                            So the overall error rate in both directions would be 14/394 or about 3.55% on this sample. If I've stated that correctly, it's not far off from a projected 5% error rate mentioned by David [Weston] or others earlier.

                                            I added 16 SNP-untested R1b DYS 492=13 pins to a U106+ map to see if this set of data seemed to track other U106s and might help get at geographic trends. So it sounds like an expected error out of that sample would be 0-2 pins, maybe.

                                            My gut feel is that I'll be very impressed if we find any other marker or combination of markers with a higher predictive value for U106+ than this one. I wish I could call the stock market this well! / Tom"

                                            ______________________
                                            i'm EMAILING FOR THE GREATER GOOD
                                            http://im.live.com/Messenger/IM/Home/?source=EML_WLHM_GreaterGood

                                            > ________________________________
                                            >
                                            > To: R1b1c_U106-S21@yahoogroups.com
                                            > From: adam.bradford@...
                                            > Date: Fri, 13 Jun 2008 17:41:14 -0400
                                            > Subject: Re: [R1b1c_U106-S21] DYS492 = 13, U106-
                                            >
                                            > I am DYS492=14 and also S21 positive according to both FTDNA and EthnoAncestry.
                                            >
                                            > Some (perhaps most) of the S21/492=14 people are going to prove to part of a little subcluster of the Frisian cluster that is also characterized by 464d=19.
                                          • adam bradford
                                            Hi Tom, Yes, Frisian is 464d=18. But there is a further subcluster of Frisian that is like Frisian in most respects (390=23, 447=24, etc.) but which is
                                            Message 21 of 27 , Jun 15 8:46 AM
                                              Hi Tom,

                                              Yes, Frisian is 464d=18.  But there is a further subcluster of Frisian that is like Frisian in most respects (390=23, 447=24, etc.) but which is defined by 492=14.  There is also a high correlation between 494=14 and 464d=19 within this group, 19 presumably being a step away from the Frisian.  There were some messages about this on the GENEALOGY-DNA list last month.  And I believe someone else on this list also noticed it. 

                                              - Adam

                                              On Sat, Jun 14, 2008 at 6:08 PM, Thomas Gull <tmrgull@...> wrote:


                                              Adam, per Ken Nordtvedt's quote just below, 464d would have 18 repeats as a mark of inclusion in the R1bSTR22 "Frisian" cluster... Are you maybe thinking the larger set of STRs with DYS390 = 23?

                                              Ken in a recent email to me: "Do you mean the clade with 24 at 447, 10 at H4, 18 at 464d, 23 at 390, 11 at 391? That's what I consider the Frisian clade...This clade has 13 at 487 and is S21+ --- although certainly not all of S21+ by a long shot."

                                              On a related topic in this thread, I listed some raw numbers earlier for DYS 492 = 13 and its potential value as a predictive value for U106+ and family (S29, etc). Here's a quote from a later message I sent to three people on that topic:

                                              "I guess one way to look at it is that if you adopted a simple 'DYS 492 value of 13 predicts U106+, all other 492s don't' hypothesis, then in this sample... [U106 project, U152 David Faux project, S116+ project]

                                              You would predict 165 U106+ results and 229 non-U106+ results.
                                              Your predictions would be correct in 380 cases, falsely positive in 5 cases, and falsely negative in 9 cases.
                                              False positive would be a prediction someone was U106+ when they were S116+ or U152+.
                                              False negative would be a prediction someone was not U106+ when in fact they were.

                                              So the overall error rate in both directions would be 14/394 or about 3.55% on this sample. If I've stated that correctly, it's not far off from a projected 5% error rate mentioned by David [Weston] or others earlier.

                                              I added 16 SNP-untested R1b DYS 492=13 pins to a U106+ map to see if this set of data seemed to track other U106s and might help get at geographic trends. So it sounds like an expected error out of that sample would be 0-2 pins, maybe.

                                              My gut feel is that I'll be very impressed if we find any other marker or combination of markers with a higher predictive value for U106+ than this one. I wish I could call the stock market this well! / Tom"



                                              ______________________
                                              i'm EMAILING FOR THE GREATER GOOD
                                              http://im.live.com/Messenger/IM/Home/?source=EML_WLHM_GreaterGood

                                              > ________________________________
                                              >
                                              > To: R1b1c_U106-S21@yahoogroups.com
                                              > From: adam.bradford@...
                                              > Date: Fri, 13 Jun 2008 17:41:14 -0400
                                              > Subject: Re: [R1b1c_U106-S21] DYS492 = 13, U106-

                                              >
                                              > I am DYS492=14 and also S21 positive according to both FTDNA and EthnoAncestry.
                                              >
                                              > Some (perhaps most) of the S21/492=14 people are going to prove to part of a little subcluster of the Frisian cluster that is also characterized by 464d=19.


                                            Your message has been successfully submitted and would be delivered to recipients shortly.