No announcement yet.

Calculating relationship based on shared/total CM

  • Filter
  • Time
  • Show
Clear All
new posts

  • Calculating relationship based on shared/total CM

    I have a question in regards to estimating relationship. I will explain the background ...

    I have had mine and my brother's DNA tested here, 23andMe, and Ancestry. Recently I was contacted by a connection at 23andme, and after sharing genomes we learned that she and I shared a substantial chunk of DNA on Chr 9 in the segment range of 25.6-70.3, and we are estimated as 3rd to 6th cousins there, with .42% shared DNA. She gave me the surname of her Polish grandmother, believing that to be through whom she connected with me.

    So I went to and looked up her SURNAME-A in Poland, because I knew they had Catholic Church Records indexed there. Sure enough, her "SURNAME-A" showed up in Boleslaw, the village that my 2nd great grandmother came from. Her grandmother later confirmed that is indeed where her ancestors came from.

    Then I remembered that I had another distant cousin here at FTDNA who had a "SURNAME-B" ancestor that traced back to the same town. I compared the DNA and the Chromosome and Segment matches overlapped. Here, THIS cousin is listed as 2nd to 4th, with a shared CM of 95.95 and longest block 41.97. But here I see a little more data … in addition to this sizeable overlap on Chr 9, we have a total of 11 segment matches on Chr 1, 2, 3, 7, 9, 10, and 12 in places, from 500-8545 SNPS.

    So ... after tracing back through to a possible connection in the indexes at Boleslaw, I did find a connection to SURNAME-A, where a man of MY ancestor's surname married a woman that belonged to SURNAME-A. If this woman's father is our common ancestor (the indexes don't go back far enough to ascertain), then the paper trail would point to a common 5th great grandparent. That would make us 6th cousins.

    So now it's looking to me like the 23andMe relationship might be closer to what records potentially show ... is it possible for a supposed .42% shared DNA relative to be a 6th cousin? Because I can't work out yet, tree-wise, how we can all end up being 2nd to 4th cousins through the SAME ancestor, and DNA points to a common ancestor somewhere between the three of us.
    Last edited by DeeTyler; 3rd February 2013, 07:52 AM.

  • #2
    Some people can have an inversion near the centromere on chromosome 9 (pericentric inversion) that can cause larger segments to persist longer than usual. But elsewhere on chromosome 9 larger segments tend to persist for longer periods of time too because there is DNA associated with the immune system there.


    • #3
      That appears to be where it occurs ... about half and half.

      If you could explain this to me in "newbie" language I'd be most appreciative. A lot of this is way over my head. I'm just thrilled to find this degree of a connection, though!


      • #4
        I strongly recommend that you upload your autosomal and X chromosome raw data to This will do two things. It will allow you to compare your genome to those of customers from several testing services. It will also allow us to use the analytical tools on the site to get a better handle on your data than we can by reading your description of them. That said, based on your description of your data, I estimate that you are separated by about 4.2 generations from the Most Recent Common Ancestor (MRCA) of you and your match. In other words, you and your match are probably 3rd cousins or 3rd cousins once removed.


        • #5
          Thank you MFWare, I've added my Gedmatch info to my signature.

          In some respects, this makes sense ... that takes me to a generation where many questions lie in the paper trail.
          Last edited by DeeTyler; 3rd February 2013, 12:08 PM.


          • #6
            The numbers suggest you should be genealogically closer to your FF match than your RF match, and your FF match does not include X Chr. data.

            Why not suggest your FF and RF matches both add their data to Gedmatch as well.