Page 157 of 162 FirstFirst ... 57107147155156157158159 ... LastLast
Results 1,561 to 1,570 of 1620

Thread: Big Y-700

  1. #1561
    Registered Users
    Posts
    3
    Sex

    Mike, thank you for showing an interest in A1487. I consider the branch highlights something that can be improved by FTDNA.

    One of the frustrations we had under A1487 is that we knew there was a close genetic match between one Fitzpatrick project member and an unknown person. We can see from the block tree that the pair share the same terminal haplogroup, but the match does not appear in the list of ‘BigY matches’. The shared ancestry is ca. 1100-1200 AD, so after the formation of surnames in Ireland, and within the ‘genealogical timeframe’, as you term it.

    And this is where FTDNA referring to ‘BigY matches’ as such misleads – they aren’t BigY matches at all. Rather, they are matches that meet the STR-matching threshold in testers who have also taken a BigY. This approach to haplotype matching misses some matches because, at least given our experience under A1487, there can be some crazy off modal STR-signatures. We have now ‘discovered’ the ‘missing match’ and he is 10/12, 20/25, 26/37, etc – so you get the picture.

    Changing the 12-marker threshold for matching from one step to two steps would have resolved our issue – but I imagine doing this across the board would place too much demand on a system that is already sluggish. So, how difficult can it be for FTDNA to base ‘BigY matches’ on SNPs instead of STRs?


    Mike Fitzpatrick PhD
    Fitzpatrick DNA Project

  2. The Following User Says Thank You to Mike Magill For This Useful Post:

     IanFitzpatrick (07-22-2021)

  3. #1562
    Registered Users
    Posts
    4,402
    Sex
    Y-DNA (P)
    R1b
    mtDNA (M)
    H

    Quote Originally Posted by Mike Magill View Post
    Mike, thank you for showing an interest in A1487. I consider the branch highlights something that can be improved by FTDNA.

    One of the frustrations we had under A1487 is that we knew there was a close genetic match between one Fitzpatrick project member and an unknown person. We can see from the block tree that the pair share the same terminal haplogroup, but the match does not appear in the list of ‘BigY matches’. The shared ancestry is ca. 1100-1200 AD, so after the formation of surnames in Ireland, and within the ‘genealogical timeframe’, as you term it.

    And this is where FTDNA referring to ‘BigY matches’ as such misleads – they aren’t BigY matches at all. Rather, they are matches that meet the STR-matching threshold in testers who have also taken a BigY. This approach to haplotype matching misses some matches because, at least given our experience under A1487, there can be some crazy off modal STR-signatures. We have now ‘discovered’ the ‘missing match’ and he is 10/12, 20/25, 26/37, etc – so you get the picture.

    Changing the 12-marker threshold for matching from one step to two steps would have resolved our issue – but I imagine doing this across the board would place too much demand on a system that is already sluggish. So, how difficult can it be for FTDNA to base ‘BigY matches’ on SNPs instead of STRs?


    Mike Fitzpatrick PhD
    Fitzpatrick DNA Project
    You have the R-A1487 MRCA at 1100 AD? Ian had “ca. 900AD” but 1100 makes sense as far movements related to the Normans.

    I personally think Y12 and Y25 matching are almost useless.
    In the past I have requested that FTDNA open the spigot on Y37 STR matching from GD=4 to GD=5. Robert Casey, who posts here, disagrees with me but there is an Owston family study that show good matches at Y67 show up that are missing at Y37.

    Something still seems off here. SNPs can happen in fits and starts but on both sides of a deep split within A1487?
    Last edited by TigerMW; 07-22-2021 at 09:26 PM.

  4. #1563
    Registered Users
    Posts
    105
    Sex
    Ethnicity
    Irish/Scot/English
    Y-DNA (P)
    FGC11134/ZZ44/BY9002
    mtDNA (M)
    H4a1a1

    Ireland Scotland England
    Here is a good Example of how the ďsurname lineĒ between our results and other surname blocks lines up perfectly

    On the right is a huge Maxwell group under FGC5494 and the on the left is A1506 on the Big Tree. which BTW is still the go to tree for me, it is everything the Block tree is not.

    See, deflecting by a couple of hundred years between what I said and Mike said doesnít change the facts, at least we are not saying surname blocks could have originated 1700 years ago, we all know this isnít possible. You need to read the paper.

    Yes, something is off, the old thinking of 70 years/SNP doesnít apply across the board, the DNA information uncoupled from history is useless, itís just a giant mix of surname jumbles. Do the work and you can make sense of it and it all fits quite well

    90FBA0F5-7EB6-4B7F-B5CA-B1A9BD939399.jpeg

  5. #1564
    Registered Users
    Posts
    3
    Sex

    No, I don’t have A1487 at 1100-1200 AD – I agree with Ian. I was referring to a branch under A1487.

    On my part I have recently re-considered the usefulness of 12- and 25-marker matches. And that’s because the SNP filter is a wonderful addition – sure, it was possible to manually filter before, but the upgrade makes it easier.

    As for something being off under A1487 –it is a really interesting and unusual group, well at least A1487>A1499 it is. You only have to look at DYS390 and YCAiia, which (before the nature of these particular mutations was understood) used to be referred to as ‘the double-triple’ – yes, the first and probably the last time you will ever see DNA explained using basketball terminology.

  6. #1565
    Registered Users
    Posts
    105
    Sex
    Ethnicity
    Irish/Scot/English
    Y-DNA (P)
    FGC11134/ZZ44/BY9002
    mtDNA (M)
    H4a1a1

    Ireland Scotland England
    The balance sheet on both sides of the A1487 branching works out very well, the tree was built from both top down and bottom up.

    Also, if you could see the over 150 STR results charted out by SNP’s you would see how STR’S will send you off on wild goose chases, too many outliers when you deal with a large group. They are a guide to place a kit generally, let’s say with the last 1000 years, but the fine branching below this point is impossible to determine without SNP’s and history so building the matching algorithm on STR’S is a not a good way to do it.
    Last edited by IanFitzpatrick; 07-22-2021 at 10:03 PM.

  7. #1566
    Registered Users
    Posts
    4,402
    Sex
    Y-DNA (P)
    R1b
    mtDNA (M)
    H

    Quote Originally Posted by Magill
    No, I donít have A1487 at 1100-1200 AD Ė I agree with Ian. I was referring to a branch under A1487.
    Iain has been referring to R-A1487 and the six men with R-A1487 that are not Big Y tested. That's what I followed up on and responded to. Here is what Ian said,
    Quote Originally Posted by IanFitzpatrick
    over 50 BigY tests on men under the A1487 block and they all share a common ancestor of ca. 900AD. This line has a line of Fitzpatrick who descend from the 1st Baron of Ossory so the written records are quite good.
    I emboldened the 900 AD date for the Ian's proposed R-A1487 common ancestor.

    I ask again, how is the 900AD MRCA date calculated for R-A1487?

    Quote Originally Posted by IanFitzpatrick View Post
    The balance sheet on both sides of the A1487 branching works out very well, the tree was built from both top down and bottom up.

    Also, if you could see the over 150 STR results charted out by SNPís you would see how STRíS will send you off on wild goose chases, too many outliers when you deal with a large group. They are a guide to place a kit generally, letís say with the last 1000 years, but the fine branching below this point is impossible to determine without SNPís and history so building the matching algorithm on STRíS is a not a good way to do it.
    We can set the STRs aside for the moment, but we should not forget them or arbitrarily throw away data.

  8. #1567
    Registered Users
    Posts
    105
    Sex
    Ethnicity
    Irish/Scot/English
    Y-DNA (P)
    FGC11134/ZZ44/BY9002
    mtDNA (M)
    H4a1a1

    Ireland Scotland England
    I ask again, how is the 900AD MRCA date calculated for R-A1487?
    If you read the paper I linked and look at the tree I posted in post #1559 your question has been answered already

  9. #1568
    Registered Users
    Posts
    4,402
    Sex
    Y-DNA (P)
    R1b
    mtDNA (M)
    H

    Quote Originally Posted by IanFitzpatrick View Post
    If you read the paper I linked and look at the tree I posted in post #1559 your question has been answered already
    The narrative of the paper doesn’t mention A1487. I do see it in the appendix/diagram haplotree. It looks like paper is mainly discussion of which lineages rightfully descend of a particular Fitzpatrick patriarch. I could see how this could be a sensitive issue.

    Ian or Mike, how many of the Big Y tests on the R-BY1499 deep split of R-A1487 are Big Y/Y500 versus Big Y700?

    That may make a difference on mutation rates.
    Last edited by TigerMW; 07-23-2021 at 09:53 PM.

  10. #1569
    Registered Users
    Posts
    105
    Sex
    Ethnicity
    Irish/Scot/English
    Y-DNA (P)
    FGC11134/ZZ44/BY9002
    mtDNA (M)
    H4a1a1

    Ireland Scotland England
    Quote Originally Posted by TigerMW View Post
    The narrative of the paper doesn’t mention A1487. I do see it in the appendix/diagram haplotree. It looks like paper is mainly discussion of which lineages rightfully descend of a particular Fitzpatrick patriarch. I could see how this could be a sensitive issue.

    Ian or Mike, how many of the Big Y tests on the R-BY1499 deep split of R-A1487 are Big Y/Y500 versus Big Y700?

    That may make a difference on mutation rates.

    Yes we only have A1506 in the paper but you can count the SNP’s up to A1499, as I said in my previous post it’s 4 SNP’s up

    They are almost exclusively BigY700 and yes that does make a difference, there may be a couple of Y500 test but they are not relevant, they are under the Fitzpatrick block which has huge number of men that share a 1480 AD common ancestor

    There are also a number of very closely related men that tested and this “Wrings” out quite a number of SNP’s that normally would not be named.

    There are 5 men who have 4 SNP’s under a shared common ancestor at 1820 so throw out the 70years/SNP, we are seeing the 70 years shrink to below 50 years on other groups we have also heavily tested.

    Here is another one, the BY12234 block goes back to origins of surname times and is 22 SNPs deep when you add the privates…so about 40-45 years/SNP to make that work, again, the 70 years doesn’t work.

    http://www.ytree.net/DisplayTree.php?blockID=1084
    Last edited by IanFitzpatrick; 07-23-2021 at 11:32 PM.

  11. #1570
    Registered Users
    Posts
    4,402
    Sex
    Y-DNA (P)
    R1b
    mtDNA (M)
    H

    I agree with people who want FTDNA's matching thresholds loosened up a bit. As I mentioned before I formally requested change the matching limit at Y37 from GD=4 to GD=5. This would also help on the Big Y Block Tree on Big Y matches.

    This is confusing at first, but let me provide a little history on Big Y matching. Originally, Big Y matching only had one criteria, a 30 variant mismatch threshold. If there were more than 30 mismatches of variants between any two Big Y testers they were not considered. Less were considered matches. FTDNA never disclosed their methodology for coming up with the number 30. It appears to be correlated somewhat with their average STR matching thresholds, which amount to 10-11% diversity (Genetic Distance =4 at Y37, =7 at Y67, 10 at Y111. They appear to tighten things up at Y111 versus the lower matching levels. I can't figure that out unless they have a higher confidence level of matching.)

    Back then, there were concerns expressed, such as Ian's, about the restrictiveness of the Big Y variant matching threshold. As a result, FTDNA added a Y37 matching threshold for the Big Y Block Tree.

    As displayed by tester names on the Block Tree there are two thresholds for Big Y matching. Their "either or" so if you meet either threshold you are match and the user name is displayed. In other words, you don't have to pass through both gates to be a match, just either gate. For a testers name to appear on the Block Tree you still have to have Big Y, but even if you don't meet the 30 variant mismatch criteria, your name will also be display if you are on the Y37 Y-DNA (STR) matching display. It makes sense in that there is no violation of privacy if you are already seen to the other person on Y37.

    I think the Big Tree is great and I recommend its use for those willing to submit their data to a third party. I have so Alex Williamson has my data and has done a great job. In fact, FTDNA works with Alex and even adapted the Big Tree formatting for its Big Y Block Tree. We've also seen the credit on the bottom of the display. The Big Tree will never have as many individuals as the Block Tree and Alex acknowledges that, but there is no reason not to look at both.

    I view the Big Tree the same way I view YFull. It's just another portal that you can display your Big Y results publicly. It's your option. In the case of the Big Tree it is free. Of course another portal you can choose to display your data is the FTDNA project system. Many projects are public, particularly the large haplogroup projects. It's also free, just like the Big Tree.
    Last edited by TigerMW; 07-24-2021 at 02:01 AM.

  12. The Following 3 Users Say Thank You to TigerMW For This Useful Post:

     Dewsloth (07-24-2021),  MitchellSince1893 (07-24-2021),  Webb (07-24-2021)

Page 157 of 162 FirstFirst ... 57107147155156157158159 ... LastLast

Posting Permissions

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