Page 72 of 72 FirstFirst ... 2262707172
Results 711 to 713 of 713

Thread: Big Y changes

  1. #711
    Gold Member Class
    Galactic Empire
    R-Z324, Z5055, L188+

    Quote Originally Posted by Mikewww View Post
    The Hg38 conversion project is working!

    It is clear that FTDNA has re-evaluated SNP results and branching across the haplotree as a part of this project.

    Hg38 has enabled them to be more confident in variant calls. This has uncovered some new variants, but more importantly has increased FTDNA's aggressiveness in calling more variants as reliably derived and culling out marginal calls. At the same time, we've seen an acceleration of Big Y testers.

    I just ran through the R1b-M269>L23>L51>L151>P312>Z290(and L21) haplotree and compared it what I had archived on October 9th, right before the Hg38 conversion started.

    The number of distinct branches in L21 increased over the last four months by 9.1%, from 2,034 to 2,220.

    The total number of SNPs on the L21 tree increased by a whopping 24.7%, from 7,909 to 9,859. L21 is closing in on 10,000.

    (for reference) DNAexplained article on Hg38 and the new functions in Big Y:

    P.S. Nothing is perfect nor ever will be save one, but a lot of progress has been made without upcharges or subscriptions for existing Big Y testers. It's been a long time coming, that's for sure.
    Re-do this. You are comparing result list of "A" samples against a list of "A+B" samples. Of course you will have more SNPs since you have more BigY results present in the data set. You are making an improper association that the difference is due to the hg38 comparison. Bad science.

  2. #712
    Registered Users

    Quote Originally Posted by Cofgene View Post
    Re-do this.
    No, I didn't get to my second post which I've already made on several forums.
    Notice how I emphasized the word "project".

  3. #713
    Registered Users

    I posted this is on the Big Y FB forum within minutes after the original FB post. All of this I did before posting here.

    I also posted the same three reasons on the L21 group yesterday.

    "There are three reason that I can see for the growth. One is all of the passionate Big Y testers. More are testing than ever before. I think this is the primary driver for the increase in the number of branches.

    However, the increased aggressiveness in FTDNA's variant calls and their re-evaluation, branch by branch, of the haplotree under the Hg38 reference has also been helpful.

    So the second reason is that FTDNA says the Hg38 reference is allowing them to make more confident calls. I noticed "Y" and "FGC" SNPs added as phylogenetic equivalents to many branches already defined by a series of "BY" or "Z" SNPs. At the same time I've even seen new "BY" SNPs on the same branches. You can tell this by number assigned. FTDNA now has BY SNPs up into the BY35500's. There have been branches and SNPs removed too, I presume for quality reasons.

    A third reason is that FTDNA appears is driving the tree themselves. They appear to have rolled through the tree, branch by branch manually. They use BAM data to do this so the quality of the tree should be good. Their analysis is pure Hg38 now so they don't have to mix apples and oranges of Hg19 and Hg38 calls.

    Please note. Any manual process can miss things. I always look at my own branches myself. I don't want anything to fall through the cracks. This forum is a good place to ask questions about how to do this. The first thing to do is probably maintain patience.

    As an old software developer I should add that automated processes can miss things too, it's just that when they do the results are often wide spread and easy to notice."

    It is definitely legitimate to consider #3 (FTDNA driving the tree more themselves) as part of the project because it is. They ran through branches to clean things up as Hg38 results came.

    I am not studying this but after going through every single branch on the R1b-L21 tree, my observation that the increase in total # of SNPs is mostly related to FTDNA being more aggressive due to their confidence in Hg38.
    "The total number of SNPs on the L21 tree increased by a whopping 24.7%, from 7,909 to 9,859. L21 is closing in on 10,000."

    Many branches that were BY series named had "Y" and "FGC" SNPs recently added. At the same time, a very recent "BY" SNP or two may be added. This is a sign of going beyond YFull's and FGC's aggressiveness because of the confidence in Hg38.

    The increase in branches is probably mostly driven by new Big Y sales, which were all Hg38 based. If you like to argue terminology you can argue new results shouldn't be part of the project but this all mixed together and these are Hg38 orders.
    "The number of distinct branches in L21 increased over the last four months by 9.1%, from 2,034 to 2,220."

    After thinking about it, I probably could have called this the "Big Y2 Project" but not many people call it that.

    My intent is always full disclosure... I already did the three reasons on other forums. I was just slower here as I do attend to other duties, liking updating L21 haplotypes now and then R1b in general. Thanks for keeping me on my toes.

    This has been a period of tremendous progress for Big Y. I never really liked this characterization by a strong ISOGG advocate on another forum, but maybe he is converting me. He always says "Big Y is the gift that keeps on giving".

    This year we can expect Big Y testers to get the extra Y STRs. My understanding is they will be free. FTDNA calls this the Y500 project, but I think it is all a part of the same thing. It is Big Y2 but Big Y1 folks get it too so it's really just a massive conversion project.

    I think there is a chance we will get age estimates this year too. That's quite speculative, though.
    Last edited by Mikewww; 02-13-2018 at 03:08 PM.

Page 72 of 72 FirstFirst ... 2262707172

Posting Permissions

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