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

Re: So is there any set of Alleles / STR values with strong prob of S21+

Expand Messages
  • phelpscan
    Partial answer to my own question below (does DYS492 = 13 show up outside of R1b1c?). I ve taken a sweep through six geographic projects and one haplogroup
    Message 1 of 5 , May 1, 2008
    • 0 Attachment
      Partial answer to my own question below (does DYS492 = 13 show up
      outside of R1b1c?).

      I've taken a sweep through six geographic projects and one haplogroup
      project looking for DYS492=13. Those are:

      French-Swiss project
      French-Canadian founders project
      Iberian project
      Germany project
      Danish Demes project
      Scandinavian project
      Kerchner's R1b and Subclades YDNA Haplogroup Project (in progress)

      From the geographic projects:

      54 DYS492 = 13 entries (technically 57 because there are two father-son
      pairs and one half-brother pair)
      53 = some flavor of R1b
      1 = Q
      3 = I1DC

      SNP tested
      10 R1b1c9
      1 R1b1c9a
      1 R1b1c10 (yes, Fitzbag N2125 in French-Canadian proj with DYS492=13)

      From a partial sweep through Kerchner's R1b and Subclades YDNA
      Haplogroup Project, so far every tested R1b1c9 has a 13 at DYS492, and
      the tested ones total about half of the entries with 13. I haven't seen
      another subclade show up with that value, but I've still got about half
      the tested SNPS to check.

      So it looks like a DYS492=13 search should be restricted to R1b to get
      rid of a few outliers (unless those I1C/Q tests were wrong). Within
      R1b, the correlation between DYS 492 = 13 and R1b1c9 in these samples
      seems quite high as earlier feedback to me indicated was the case.

      / Tom

      --- In R1b1c_U106-S21@yahoogroups.com, Thomas Gull <tmrgull@...> wrote:
      > Thanks, Mike. It's easy to lose track of a generalized conclusion
      people have reached in the middle of a mass of threads. I see from a
      quick scan of the Germany project that only 20 or so people do have
      measurements at DYS492, so it's definitely not going to add lots of
      datapoints from that project.
      > Do you know any kind of confidence band for the "...13, the person is
      probably U106+..." hypothesis? Is it qualified by saying that
      statement is true only in context (that is, only when looking at
      R1b)? / Tom
    • gregkirokh
      I read this the other day. I originally thought my markers past 25 did not match modals because of mutation rates and an early American heritage. I think the
      Message 2 of 5 , May 2, 2008
      • 0 Attachment
        I read this the other day. I originally thought my markers past 25 did
        not match modals because of mutation rates and an early American
        heritage. I think the time from for 1-25 marker modals to be different
        than 1-37 marker modals. Modals beyond the ones which have the
        fundamental six, DYS 390 - 3, 388, 19, I find interesting, but I do
        not fully understand their application. Markers DYS 456, CDYa, and
        CDYb confuse haplogroup predictors when I tried them. One even thought
        I was hg X which is crazy for AMC.

        I am same at all the critical S21 markers:
        DYS390 ---> 24 to 23; 492 ----> 12 to 13; 576 ----> 18 to 17
        I differ from the stated S21 modal at six markers:
        385b 14 ---> 15 only one between markers 1 - 25
        460 11 ---> 10
        456 16 ---> 18
        607 15 ---> 16
        CDYa 36 ---> 39
        CDYb 37 ---> 40

        --- In R1b1c_U106-S21@yahoogroups.com, Michael Maddi <mtmaddi@...> wrote:
        > The short answer to your question is there is only one
        > marker that has any value in acting as an indicator
        > for who is probably U106+ and who is not. That's
        > DYS492 - if it's 13, the person is probably U106+, if
        > it's 12 he is probably U106-. I believe there a couple
        > of people who've tested U106+ who have DYS492=14, but
        > by and large you would look for those with DYS492=13.
        > And it's not foolproof and only those with 67 marker
        > will have a value for DYS492.
        > There just is not a lot of difference between the
        > modal for U106 haplotypes and those of general
        > R1b-M269. Here's a comparison of the two modals -
        > There are differences in modal values for just 3
        > markers - DYS492, DYS390 and DYS576. Everyone knows
        > about the famous "Frisian" DYS390=23, which is the
        > modal for U106, but barely so. Look at this table put
        > together by Leo Little, who's done a lot of good work
        > on U106 -
        > This shows the frequency of STR counts at each of the
        > 67 markers, for those with DYS492=12 and DYS492=13.
        > You can see that there is some sort of correlation
        > between DYS492=13 and DYS390=23. Only 7% of those who
        > have DYS492=12 have DYS390=23, while that climbs to
        > 56% for those who have DYS492=13.
        > But this is only slightly more than half of the
        > sample. If you were to take DYS390=23 as the indicator
        > of U106+ men, you'd be leaving out a lot of them.
        > Still, Leo created another table which used DYS390=23
        > as a key marker in analyzing R1b-M269 in the YHRD
        > database. You can see that table at -
        > It shows that the areas of highest percentage of
        > R1b-M269 haploytpes with DYS390=23 occur in northern
        > and central European nations, from Belgium to Denmark,
        > south in Austria and the Czech Republic. But this
        > rough estimate of the percentage of U106 in
        > populations is only getting a bit more than half,
        > given what I described above. We don't know the
        > distribution of the rest of U106 throughout Europe for
        > those who have DYS390=24, since that's the modal value
        > for general R1b-M269. With the short haplotypes in
        > YHRD, DYS492 is not tested.
        > Mike Maddi
        > --- phelpscan <tmrgull@...> wrote:
        > > Say I wanted to look at the Germany DNA project R1b
        > > data to find (a)
        > > people with decent ancestral locations listed and
        > > (b) not yet SNP-
        > > tested for S21+ but (c) high probability of being so
        > > because of their
        > > values for a specified set of alleles. The allele
        > > distributions for the
        > > tested people in the U106/S21+ project are online.
        > > Are we at or near a
        > > point yet where some set of allele values does seem
        > > predictive for
        > > S21+? Maybe using a combo of 10-20 of them or
        > > something? Of the two
        > > predictive errors (person is S21 but not identified
        > > or person isn't S21
        > > but is identified as such), I'd only care about
        > > minimizing the false
        > > positives. That is, tiny false positives and sizable
        > > false negatives
        > > still might enlarge our geographically-detailed
        > > sample size
        > > considerably without dropping our overall confidence
        > > levels
        > > enormously. / Tom Gull
        > >
        > >
        > Be a better friend, newshound, and
        > know-it-all with Yahoo! Mobile. Try it now.
      Your message has been successfully submitted and would be delivered to recipients shortly.