No announcement yet.

Can anyone make sense of this?

  • Filter
  • Time
  • Show
Clear All
new posts

  • Can anyone make sense of this?

    I did a test for my father in law who is from a small village in the mountains in Bulgaria. I found a match to a man in Florida who has mostly Spanish ancestry. I then did a triangulation and got a handful of people who matched both. I wrote to them all and got two responses. The only conclusion I can make is that I have four people who show a match with each other but don't seem to have any regions in common let alone individuals. Can anyone explain? The system says five and a bit generations back. I am guessing that is wrong and that the matches of 10 cM and 11 cM are just chance ... GEDmatch® Genesis

    People who match both kits, or 1 of 2 kits

    Kit 1: HK9849916 (Raicho Velkov)
    Kit 2: WU3674237 (Federico Hurtado Gisbert)

    Matches both kits
    Match Name HK9849916 WU3674237 Generations
    Shared Largest Gen Shared Largest Gen
    M203037 Kivanc Kalar 23.8 13.9 4.6 13.1 13.1 5.0 0.4
    T912945 *H. A. Stuart 12.3 12.3 5.1 10.5 10.5 5.2 0.1
    QY4253501 *Tiny A's Great Uncle 11.6 11.6 5.1 13.0 13.0 5.1 0.1
    Z958622 C. J. 11.5 11.5 5.1 13.6 13.6 5.0 0.1
    M331169 Ginnie Lee 11.4 11.4 5.2 11.0 11.0 5.2 0.0
    A215379 Virginia McMahon 10.7 10.7 5.2 11.0 11.0 5.2 0.0
    YD6733294 *Kemal 10.4 10.4 5.2 12.0 12.0 5.1 0.1
    M083104 David Haseltine 10.3 10.3 5.2 10.8 10.8 5.2 0.0
    LW4340393 *DaleD 10.1 10.1 5.2 11.1 11.1 5.2 0.1
    M926688 *DaleD 10.1 10.1 5.2 10.9 10.9 5.2 0.1
    A342209 *Jen 10.1 10.1 5.2 11.2 11.2 5.2 0.1
    Response from Z958622
    "I am German. My parents have mostly German Ancestry, as much as I know actually (I researchrd about 1550 ancestors actually) . There are some of my mothers ancestors with polish ancestry. But that is 5 generations from my mother. Perhaps there are some common slavic ancestors centuries ago to your Bulgarian matches. My fathers ancestors are mostly from Hessen. 350years ago there are a lot of wallon immigrants (today Belgium). But no east European or slavic ancestry."

    Response from YD6733294
    "I am in Houston USA, but originally from Pakistan. My grandparents were from north India. I also have a little DNA from the UK as my paternal grandmother's mom was British."

  • #2
    When I use Genesis, I always try to do a one-to-one to verify that the grouping is worth looking at. I tried your father-in-law below, and in my opinion anything under 10 cM makes your chance of finding something meaningful very low. Paying for the Tier 1 tools may reduce the amount of one-to-one compares you would need to do.
    GEDmatch® Genesis Autosomal One-to-one Comparison - V1.0

    Software Version Mar 26 2019 01:00:34
    Comparing Kit HK9849916 (Raicho Velkov) [FTDNA] and WU3674237 (Federico Hurtado Gisbert) [23andMe]

    Segment threshold size will be adjusted dynamically between 200 and 400 SNPs
    Minimum segment cM to be included in total = 7.0 cM
    Mismatch-bunching Limit will be adjusted dynamically to 60 percent of the segment threshold size for any given segment.

    Chr B37 Start Pos'n B37 End Pos'n Centimorgans (cM) SNPs
    22 18,533,434 22,599,216 8.9 229
    Largest segment = 8.9 cM

    Total Half-Match segments (HIR) = 8.9 cM (0.248 Pct)
    Estimated number of generations to MRCA = 6.4

    1 shared segments found for this comparison.

    152938 SNPs used for this comparison.

    52.423 Pct SNPs are full identical

    Comparison took 0.042 seconds.
    CPU time used: 0.015 cpu seconds.

    Ver: Mar 26 2019 01:00:34


    • #3
      Why does the system so precisely say 5.1 or 5.2 generations when it might not even be meaningful?


      • #4
        mde, Just so you know: for privacy reasons you should not post information on your matches that includes their email addresses or names. See the section in this Forums FAQ under Communications, where it says:
        Community members shall not:
        • Disclose the personal information of others, such as name and or e-mail addresses of your matches, without their written consent.
        Also, unless you obtained their permission, it's not good to post the contents of emails from your matches. You've shown the kit number and then the email address for them in your table. For both in this case, the email address contains their names.


        • #5
          Okay. Fair enough. I will delete it. But as the emails are up there on Gedmatch for anyone to see it seems to be in the public domain so it didn't occur to me that it would not be allowed.
          Last edited by mde; 18th April 2019, 05:16 AM.


          • #6
            Actually it seems it is only possible to edit the last post you have made in a thread ...


            • #7
              Originally posted by mde View Post
              Okay. Fair enough. I will delete it. But as the emails are up there on Gedmatch for anyone to see it seems to be in the public domain so it didn't occur to me that it would not be allowed.
              You only see the email addresses of matches. We don't have access to the full database. You are showing email addresses and IDs we wouldn't have normal access to. You might want to check the terms of use for GEDmatch and this forum.


              • #8
                Why are you still looking for Sephardic Jewish ancestry from Spain in a person where it doesn't exist? The cited Bulgarian person is not a real genetic match to the cited Spanish-descended person; they "match" on an especially notorious Excess IBD Region of chromosome 22 and share only 229 SNPs despite a length of 8.9 cM, which is way lower than any trustable ratio.

                Meanwhile, those British+Indian and German+Polish "matches-in-common" are just as dubious because their ethnic backgrounds don't intersect with the Balkans nor with Iberia within the autosomal timeframe of ~15 generations.