Results 1 to 8 of 8

Thread: FTDNA's Z18 haplotree (and SNP Pack?) goofed up

  1. #1
    Gold Class Member
    Posts
    2,648
    Sex

    FTDNA's Z18 haplotree (and SNP Pack?) goofed up

    Kit N88045 of Poland received his Big Y results this week. He was already known to be Z18+ ; the Big Y found that he was more specifically DF95+ . FTDNA reclassified him as R-DF95. So far, so good.

    Today, I notice that FTDNA has re-reclassified him as R-CTS12023, on the basis of CTS12023+ CTS2206+ CTS3624+ . FTDNA's haplotree depicts CTS12023 as a sibling of Z17, and DF95 as a child of Z17. Thus, in FTDNA's tree, CTS12023 is incompatible with DF95. FTDNA's classification algorithm has now apparently chosen to prefer CTS12023 over DF95.

    What makes this particularly surprising is that FTDNA recently introduced a Z18 SNP Pack, presumably based on input from the R-Z18 project. Didn't that project know that FTDNA's newly published haplotree for Z18 was wrong?

    EDIT: To be fair, I must point out that the R-U106 project doesn't appear to have its haplotree entirely correct either. It lists CTS12023 as a subclade of DF95, but then lists S19116 as a sibling of CTS12023 even though the one member of this S19116 subclade has tested CTS12023+ .
    Last edited by lgmayka; 05-14-2015 at 03:31 PM.

  2. #2
    Registered Users
    Posts
    613
    Sex
    Location
    Waterville, ME
    Ethnicity
    Great Migration Colonists
    Nationality
    American
    mtDNA (M)
    J1c2g (FMS)
    mtDNA (P)
    T2b3 (23andMe)
    Y-DNA (M)
    I-P109 (23andMe)
    Y-DNA (P)
    R1b-U106 (S10415)

    United States Gadsden England Scotland Ireland Wales
    Quote Originally Posted by lgmayka View Post
    ... [portion clipped for brevity]...

    EDIT: To be fair, I must point out that the R-U106 project doesn't appear to have its haplotree entirely correct either. It lists CTS12023 as a subclade of DF95, but then lists S19116 as a sibling of CTS12023 even though the one member of this S19116 subclade has tested CTS12023+ .
    Unfortunately, due to an unannounced "behind the scenes" change by FTDNA, coupled with an OS update to the Apple iPad, individuals using iPads (including the U106 admin, Charles Moore, as well as at least one co-admin) have not been able to scroll through any of the GAP pages (including the page which allows them to edit the STR results list) for several weeks. Due to this, it is impossible for them to make any changes to the project. Those admins affected have had a number of conversations with staff at FTDNA about this issue, but to date, no solution has occurred.

    Our U106 Tree https://app.box.com/s/afqsrrnvv2d51msqcz2o continues to be updated and lists CTS12023 and S19116 (as well as several other SNPs) as equivalent to DF95.

    Given the continued problem between the iPads and the GAP pages, the affected members are in the process of getting PCs so the pages can once again be updated.

  3. The Following User Says Thank You to Wing Genealogist For This Useful Post:

     lgmayka (05-15-2015)

  4. #3
    Gold Class Member
    Posts
    2,648
    Sex

    Quote Originally Posted by Wing Genealogist View Post
    Our U106 Tree https://app.box.com/s/afqsrrnvv2d51msqcz2o continues to be updated and lists CTS12023 and S19116 (as well as several other SNPs) as equivalent to DF95.

    Given the continued problem between the iPads and the GAP pages, the affected members are in the process of getting PCs so the pages can once again be updated.
    Thank you! Your tree is extremely helpful. I now see that my project member tested positive at a more specific level:
    16524131
    18917851
    21921677
    23097741

  5. #4
    Registered Users
    Posts
    62
    Sex
    Location
    Amsterdam
    mtDNA (M)
    H
    Y-DNA (P)
    R-Z18/L257+

    Netherlands
    Quote Originally Posted by lgmayka View Post
    Kit N88045 of Poland received his Big Y results this week. He was already known to be Z18+ ; the Big Y found that he was more specifically DF95+ . FTDNA reclassified him as R-DF95. So far, so good.

    Today, I notice that FTDNA has re-reclassified him as R-CTS12023, on the basis of CTS12023+ CTS2206+ CTS3624+ . FTDNA's haplotree depicts CTS12023 as a sibling of Z17, and DF95 as a child of Z17. Thus, in FTDNA's tree, CTS12023 is incompatible with DF95. FTDNA's classification algorithm has now apparently chosen to prefer CTS12023 over DF95.

    What makes this particularly surprising is that FTDNA recently introduced a Z18 SNP Pack, presumably based on input from the R-Z18 project. Didn't that project know that FTDNA's newly published haplotree for Z18 was wrong?

    EDIT: To be fair, I must point out that the R-U106 project doesn't appear to have its haplotree entirely correct either. It lists CTS12023 as a subclade of DF95, but then lists S19116 as a sibling of CTS12023 even though the one member of this S19116 subclade has tested CTS12023+ .
    I'm a member of the administration team of project R-Z18. From your post it's not exactly clear to me where and what we did wrong in your opinion. We are well aware of CTS12023 and consider it to be equivalent to DF95 (at the moment of writing). It is specified as such as part of the Y-DNA Tree R-Z18 on our web site www.L257.org, and has been specified as such for months. We have explicitly ignored any statements as to any other position of CTS12023 (whether or not equivalent to Z17) by anybody because we consider such statements to be in conflict with available testing data (and we prefer test data over trees).

    FT-DNA is trying to define their own tree and continuously changing the tree they're showing on the web. This was one of the bigger discussions we had with them designing the R-Z18 Panel. The R-Z18 Y-Tree is based on our own database and derived using our own software tools. The R-Z18 Panel is based on that tree. Changes to (or possibly even errors in) the Y-DNA Tree by FT-DNA have not influenced the design of the panel.

    So: what exactly is your comment and what do you think the adminstration of the R-Z18 Project did wrong ?? If you think there are errors in our tree (see www.L257.org), then feel free to help us and specify your improvements together with references to the test data that support your case).
    Last edited by Peter M; 05-15-2015 at 05:30 PM.

  6. The Following User Says Thank You to Peter M For This Useful Post:

     lgmayka (05-15-2015)

  7. #5
    Gold Class Member
    Posts
    2,648
    Sex

    Quote Originally Posted by Peter M View Post
    Changes to (or possibly even errors in) the Y-DNA Tree by FT-DNA have not influenced the design of the panel.
    Just so I understand: You mean that
    - FTDNA accepted your advice as to the design of the Z18 panel itself, but then
    - Threw away your recommended tree and constructed their own incorrect Z18 haplotree?

    If true, that is unexpected and inexplicable behavior on FTDNA's part.

  8. #6
    Registered Users
    Posts
    62
    Sex
    Location
    Amsterdam
    mtDNA (M)
    H
    Y-DNA (P)
    R-Z18/L257+

    Netherlands
    Quote Originally Posted by lgmayka View Post
    Just so I understand: You mean that
    - FTDNA accepted your advice as to the design of the Z18 panel itself, but then
    - Threw away your recommended tree and constructed their own incorrect Z18 haplotree?

    If true, that is unexpected and inexplicable behavior on FTDNA's part.
    What happened is far more complex, so I will simplify things a bit - if a project wants advice in designing an FT-DNA panel, feel free to ask, we have a lot more to say about it. $:-)

    There are two important but seperate issues here: (1) the panel and (2) the tree. FT-DNA accepted the SNPs we proposed for inclusion in the panel, but they have their own tree which includes only part of those SNPs. The R-Z18 Project has its own tree with most of the SNPs in the panel (more than in the FT-DNA tree). I've never compared our own tree to the FT-DNA tree, because: (1) we are working on R-Z18 on a daily basis and I expect our tree to be better than the FT-DNA tree; (2) the FT-DNA tree tends to change often and I would have to repeat my comparison regularly; not something I look forward to doing. Apart from that, it's unclear to me what to do with any discrepancies found, as I don't think FT-DNA would change their tree if I asked them to (we have experience with this).

    Personally, I don't see this as inexplicable behaviour. FT-DNA appears to be changing their policy slightly, in that they seem to be getting more Y-Tree centric (possibly as a result of the success, in terms of numbers sold, of Big-Y). But to do that they need a stable tree and they don't have one at the moment. They are currently building up expertise (and experience) in building and maintaining trees (or at least that's my impression). And then there seem to be some issues with their IT. As my job title is Information Systems Architect, that's my domain and I'm very interested in what they are doing. They appear to be working on things in the IT area as well. So, I tend to understand the process they're going through (and I don't think any other testing company has the IT support available either). And then use of state of the art tools is the weakest point of the administration of nearly all FT-DNA projects - with R-Z18 and the most notable exception, I guess. $:-) I guess we need to be forgiving and give FT-DNA a chance to do what they are doing.

    In general: a panel is a set of SNPs considered/expected to be in a haplogroup and thus downstream of the key SNP defining the haplogroup (e.g. Z18). This set of SNPs is a superset of the SNPs in the Y-Tree of the haplogroup. A SNP could well be in a panel without it's position in the tree being known exactly. In general, I would expect more SNPs in the panel than in the tree (of the same haplogroup).
    Last edited by Peter M; 05-16-2015 at 12:52 AM.

  9. #7
    Gold Class Member
    Posts
    341
    Sex
    Location
    Ohio
    Ethnicity
    German
    Nationality
    Galactic Empire
    mtDNA (M)
    H5
    Y-DNA (P)
    R-Z324, Z5055, L188+

    Quote Originally Posted by lgmayka View Post
    Just so I understand: You mean that
    - FTDNA accepted your advice as to the design of the Z18 panel itself, but then
    - Threw away your recommended tree and constructed their own incorrect Z18 haplotree?

    If true, that is unexpected and inexplicable behavior on FTDNA's part.
    I think there is an underlying contractual requirement that FTDNA had/has around the display of the Geno 2 based haplotree. When that requirement is removed then we should hopefully see them move towards the ISOGG structure to update and clean up current issues present in what they are providing to their customers. I have found it interesting that they could add in SNPs from panels but not correct items based upon Big-Y results or what haplogroup projects have determined from other test results.

    The big refresh of the U106 region for ISOGG is nearing completion since we feel comfortable that most major branches are stable and defined. Well at least until we get some additional FGC Elite results. We do not have high hopes that FTDNA will update their system any time soon to incorporate the changes presentthe pending submission.

  10. #8
    Member
    Posts
    32
    Sex
    Location
    Montana, USA
    Ethnicity
    Caucasian
    Nationality
    American/mutt
    Y-DNA (P)
    R Z15,downstreamU106

    Oh so there is a problem using iPads on ftdna, I thought it was just me, took me an hour scrolling down a French DNA site blindly because page wouldn't move!

Similar Threads

  1. Replies: 6
    Last Post: 03-10-2015, 05:37 PM
  2. FTDNA New YDNA Haplotree
    By Táltos in forum FTDNA
    Replies: 185
    Last Post: 10-09-2014, 01:24 AM
  3. Replies: 0
    Last Post: 04-26-2014, 11:52 AM
  4. Replies: 5
    Last Post: 04-25-2014, 02:47 AM
  5. New FTDNA haplotree, finally?
    By lgmayka in forum Other
    Replies: 7
    Last Post: 11-10-2013, 07:50 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •