Announcement

Collapse
No announcement yet.

cM differences on Ancestry and FTDNA

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • cM differences on Ancestry and FTDNA

    I'm trying to place a match at a possible cousin level and I have two different amounts of DNA shared through Ancestry's test and FTDNA. Ancestry says we share 52cM across 4 segments and predicts a 4th cousin relationship. FTDNA gives a shared amount of 91cM in 15 segments for the same match with me, and predicts a 3rd.

    At FTDNA, the largest segment shared between us is on the 18th chromosome at 40.72cM, with 2 other segments at 10.94cM and 15.26cM. The rest of the 15 segments range from 1.32 to 2.65cMs. So what do you think, 3rd or 4th cousin?

    Thanks

  • #2
    Why worry? I wouldn't trust completely either estimate. You know that you have some sort of relationship that is not close, but also is not extremely distant. I'd get busy with the paper genealogy and also try to find other kits to triangulate with your two matching kits.

    Jack

    Comment


    • #3
      Thanks, and I agree that the distance is kind of vague, but the paper genealogy is even more vague. I have a fairly good tree and my match basically has none. And so many of the shared matches are equally empty.

      Comment


      • #4
        Originally posted by keigh View Post
        Thanks, and I agree that the distance is kind of vague, but the paper genealogy is even more vague. I have a fairly good tree and my match basically has none. And so many of the shared matches are equally empty.
        I'd look for places where the lines from the two trees intersect in the space-time continuum except you don't have another tree to compare to. Maybe you can help your match get pass a brickwall. Also you may have to start thinking about some not so usual possibilities. I have come across a lot of made-up genealogies, running off and starting a new identity, and other sorts of things.

        I say frequently about DNA analysis is that I have found out a lot of things which I was not supposed to know.

        Just out of curiosity, are you doing your comparison on GEDmatch, or are both of you with the same vendor?

        Good luck!

        Jack

        Comment


        • #5
          I concur with Jack Wyatt but all things considered, 40 cM seems awfully large for a 4th cousin.

          Not speaking statistically, just from hunch mode.

          52 cM of phased DNA at AncestryDNA or 91 at FTDNA is way above expected for a 4th cousin. AncestryDNA is overly conservative.

          I would expect a third cousin, but consider 4th as an option.

          I do have a couple of distant cousins who have 5 to 7 times the amount of predicted DNA so anything can happen.

          Comment


          • #6
            We're both did FTDNA and then did the Ancestry test. So I haven't suggested going over to GEDmatch. I'm trying to go through "in common with" to chromosome matching here. And using the shared cousins at Ancestry.

            Hopefully, we'll end up with an idea of our link. Of, course on my dad's side, it could be link, upon link, upon link. My mom's paper trail seems to spread out more.

            Comment


            • #7
              If you use ADSA, you can instantly see who is matching you on Chr 18 during the 40 cM. And eliminate those ICWs who are matching you on another chromosome.

              And I would also maximize the use of the tools at GedMatch. Matrix, Tier One tools, XDNA??,

              Comment


              • #8
                Originally posted by keigh View Post
                We're both did FTDNA and then did the Ancestry test. So I haven't suggested going over to GEDmatch.
                Why not? They have tools you won't find anywhere else and they also except data from several other testing companies including 23andMe.

                Comment


                • #9
                  By all means, transfer the data to GEDmatch! You will be able to examine matches in all combinations and from all directions and control the matching parameters. You will be able to identify any other people who happen to match on the same chromosome segment (triangulation groups).

                  There is always a chance that a distant cousin will share a long segment, and it is always possible that a less distant cousin won't share anywhere near the expected amount. I have a couple third cousins who don't share any segments over 10 cM but who match strongly with all the other third cousins. I have some fourth cousins who match, some who don't. I even have a fifth cousin who shares a 24 cM segment. The estimates that are usually cited are generally theoretical averages or averages based on self-reported samples that could be biased in various ways. There is every reason to believe that the actual distribution of matching segment lengths for relationships of third cousin and beyond is very broad (i.e., extremely large standard deviation and 95% confidence limits).

                  Comment


                  • #10
                    I think I've finally got him pegged as a 4th cousin using the "In Common With" and chromosome browser here at FTDNA and the "Shared Cousin Matches" at Ancestry.

                    There seem to be 2 names that keep showing up in both groups. And they are attached to my 3rd great grandparents along with my 3rd great grandparents siblings who also married into the other family. And, locations, helped enormously. They all got married in one state and then moved together (it seems) to another state. The family that marries together, moves together, stays together.

                    Thanks for all the advice.

                    Comment


                    • #11
                      Examples of segments shared with four confirmed third cousins on FtDNA:
                      R - 15 segments - longest 40cM, total 120cM, est 2-4th
                      M - 16 segments - longest 22cM, total 86cM, est 2-4th
                      H - 17 segments - longest 12cM, total 55cM, est 4th-remote
                      J - 13 segments - longest 10cM, total 39cM, est 4th-remote

                      N - 2c1r - 12 segments - longest 41cM, total 88cM, est 2-4th
                      A - 2c2r - 18 segments (child of above - there seems to be some paternal DNA matching as well)- longest 22cM, total 80cM, est 2-4th

                      3c1r (niece and nephew of H) - not a match on FtDNA nor Gedmatch - one-to-one comparison on Gedmatch shows a 5cM segment in common (so not much at all).

                      Half3C1R - 8 segments, longest 14cM, total 36cM, est 3rd-5th
                      Half4C (daughter of above) - 12 segments, longest 8cM, total 32cM, est 5th-remote.

                      In-common-with shows that by FtDNA criteria N matches 3 of her 4 3C1R above - i.e. R, H and J, but not M.

                      N also matches the Half3C1R (so her half 4th cousin), but not the half fourth cousin's daughter.
                      Last edited by ltd-jean-pull; 13 October 2016, 07:07 PM.

                      Comment

                      Working...
                      X