Announcement

Collapse
No announcement yet.

identical chromosome segments

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

  • identical chromosome segments

    I'm wondering what conclusions, if any, I might draw from seeing three or more people who match identical chromosome segments with me--that is, same length, same starting point, and same ending point. For example…
    • on chromosome 14, three people match me on the same segment of 10.96 cM
    • on chromosome 16, four people match me on the same segment of 10.06 cM

    As far as I know, within each of these two groups, no one is related to anyone else, and I have no discernible relationship with any of them. Family Finder lists my relationship to each person as "4th cousin-remote cousin."

    FTDNA's FAQ for "Understanding Results: Family Finder" tells me that overlapping segments may or may not indicate a relationship (answer #51) and that identical start/stop points are no more meaningful than mere overlap (answer #58).

    The most skeptical conclusion would be that I can't draw any confident conclusions from overlapping chromosome segments, much less identical ones.

    What do you think?

  • #2
    I'm not an expert, and I am barely even an amateur, but have you used FTDNA's matrix option to detect whether these matches may be related to each other? Or are you assuming so because you can't find connections in their family trees or surname lists?

    My understanding - and please, someone correct me if I am wrong - that an 11 cM matching segment has the probability of being 99% IBD (identical by descent, or "inherited") and 1% IBS (identical by state, or "random"). That would suggest that these matches, although distant, received these segments IBD. And thus, I think that there's a greater likelihood that these matches are also related to each other even if you aren't sure what the connection is quite yet.

    Comment


    • #3
      Miss Blue's suggestion (compare the four people in the matrix option) is a good one. Then you can see if any of the three or four match each other. If so, ask them to look in their chromosome browser and see if they match another of your matches on the same segment(s) that you wrote about.

      Or upload your raw data to gedmatch.com and ask them to do the same. Then you can compare them to each other without asking them to do it, as long as your know their gedcom kit numbers. This approach has the benefit that you may find shared segments among your matches which didn't meet the criteria of FTDNA's algorithm to qualify as a match.

      The thing to understand is that each chromosome is a pair, one from the mother and the other from the father. It's possible that you match one person on the chromosome your mother contributed and another person on the chromosome your father contributed and the segments, by chance, look like they're overlapping on the same chromosome. This is why you have to check with the matches to see if they match each other on the same chromosome, same location as they're matching you.

      Comment


      • #4
        next steps

        Thanks -- yes, I did use the matrix, which I understand to provide another avenue to the results also available through the "in common with" (aka triangulation) feature, and the people who match me with identical chromosome segments do indeed show up as mutual matches.

        More particularly, I'm wondering whether the precisely identical matches support any conclusions regarding the validity of the apparent relationships--the FTDNA FAQ indicates that they don't.

        In part, my question starts with my observation that the vast majority of those FTDNA identifies as matches have no apparent paper connection with me, so I'm trying to figure out ways to get the results into sharper focus, as it were.

        Comment


        • #5
          Originally posted by stennor View Post
          Thanks -- yes, I did use the matrix, which I understand to provide another avenue to the results also available through the "in common with" (aka triangulation) feature, and the people who match me with identical chromosome segments do indeed show up as mutual matches.

          More particularly, I'm wondering whether the precisely identical matches support any conclusions regarding the validity of the apparent relationships--the FTDNA FAQ indicates that they don't.

          In part, my question starts with my observation that the vast majority of those FTDNA identifies as matches have no apparent paper connection with me, so I'm trying to figure out ways to get the results into sharper focus, as it were.
          Unless you know (through having the matches check their chromosome browser or by uploading to gedmatch so that you can do the research for them) that the shared segments which seem to be the same are in fact the same, you can't be sure that all the matches are sharing the same segment. That's because of the possibility of the shared segments being on the opposite chromosome of the pair, as I posted earlier.

          It is a good sign that some of the matches match each other. However, that could mean that they are matching each other in another line in their tree, different from the one in which they're matching you. That's especially the case if you and/or some of them have either colonial American or Ashkenazi lines. There's a lot of intermarriage in those two groups, so there could be multiple common ancestors for the matches you're posting about, not just one common ancestor in one line.

          Comment


          • #6
            I my experience, I have only one incident where overlapping segments of came from the same ancestor. I have proven in at least 10 cases that overlapping segments came from different ancestors.

            The only time I use overlapping now is to contact each person and ask them if they are very closely related i.e. sibling, parent, most of the time you can tell because they have the same surname, but with females you can not always tell by surname.

            Once you find out there relationship you can figure out many things.

            At 23andMe you can group closely related people together, which greatly helps your analysis and find your common ancestor.

            FTDNA needs to do this also.

            Comment


            • #7
              Ericyle: Today, too many matches, too little time, to extol the virtues of FTDNA, but put simply, have you tried their chromosome browser and matrix? They are awesome tools. AND, at FTDNA, you do not have to contact anyone to share genomes, to cajole, play the waiting game, and hope they respond to you in a timely manner, or at all. What a burden!

              If you have not, please try the chromo browser and matrix at FTDNA.
              Last edited by Biblioteque; 2nd February 2014, 09:36 AM. Reason: add

              Comment


              • #8
                Originally posted by marietta View Post
                Ericyle: Today, too many matches, too little time, to extol the virtues of FTDNA, but put simply, have you tried their chromosome browser and matrix? They are awesome tools. AND, at FTDNA, you do not have to contact anyone to share genomes, to cajole, play the waiting game, and hope they respond to you in a timely manner, or at all. What a burden!

                If you have not, please try the chromo browser and matrix at FTDNA.
                I love ftDNA, but the fact is this is one thing 23andMe got right. Very useful, though severely limited by uncooperative matches and no import capability...

                The matrix would work even better if it gave you some qualitative info when you clicked on the check mark so you could figure where the match was and if it even aligned with you... [or gave you their kit numbers so you could find them on GedMatch...]

                Bob

                Comment


                • #9
                  The tools we have now don't seem to provide verification that 3 people who are a 3-way match, even on the same segment, do in fact all 3 share a common (half-identical) sequence. See if you agree with the following thought experiment:

                  Consider 3 kits, called A, B, and C. Let's say they turn out to match, by using the chromosome browser, in all possible combinations, on a 25 cM segment of chromosome 1. Do all 3 share a common ancestor?

                  What we don't know about these three, is that within this particular segment, A has DNA sequences we will call Red and Blue (Red came from one parent, Blue from the other). B has sequences Red and Green. C has sequences Blue and Green. The matching algorithms then show that A and B match one another on this segment, because they both have the Red sequence. A and C match because they both have the Blue sequence. B and C match because they both have the Green sequence.

                  In other words, A is related to (or shares a long DNA sequence with) both B and C, but not through a common ancestor. The relationship between B and C is not a relationship shared with A, and not even the chromosome browser or the matrix tool will reveal that this is the case. Rather, one parent of A is related to B, the other parent of A is related to C, and B and C are related to one another. How often does this happen? I have no idea, because I don't have the tools that would allow me to detect this situation!

                  I managed to find one tool on the internet (Dave Pike's utilities) that will compare 3 kits on a particular segment and identify the locations where all 3 share a common sequence. With some further development, a tool like this would get us to "real" triangulation, so that a 3-way match really does mean that all 3 share a common sequence, and therefore almost certainly a common ancestor.

                  I would be interested to know if the reason we don't have this tool already, is (1) nobody thought of it; (2) it's too difficult computationally (requires too much computing power); or (3) it somehow raises a privacy issue (but if so, exactly what is the issue?).

                  Comment


                  • #10
                    Stennor, I have many examples where a group of seemingly unrelated matches (in some cases 12 ppl) all share a matching segment on the same chromosome with identical start and end points, but total cMs is only around 2cM. It's tempting to think that these matching segments are identical by descent. If they are, I suspect the MRCA is in the very, very distant past

                    Comment


                    • #11
                      Ericyle: Today, too many matches, too little time, to extol the virtues of FTDNA, but put simply, have you tried their chromosome browser and matrix? They are awesome tools. AND, at FTDNA, you do not have to contact anyone to share genomes, to cajole, play the waiting game, and hope they respond to you in a timely manner, or at all. What a burden!

                      If you have not, please try the chromo browser and matrix at FTDNA.
                      marietta: Yes I do, plus I keep a Excel sheet of all my matches Longest Segments.

                      Just because matches have overlapping Longest Segments does not mean that they share a common ancestor.

                      The interesting thing about comparing shared Longest Segments with the Matrix tool is that some of the matches that have long, say 25 cm Longest Segments, do not match with the Matrix tool.

                      I have not figured out why.

                      Have you tried it and found none matches?

                      Comment


                      • #12
                        If you have two matches on the same segment but they don't show as "in common" with each other on the Matrix, you have one of two situations.

                        1) one matches your paternal bit and one matches your maternal bit on that segment

                        OR

                        2) they share that segment but don't have enough total cM shared to show up as matches in common.

                        If they're on Gedmatch you can figure out which is the case.

                        Comment


                        • #13
                          how I deal with identical chromosome segments

                          Here are a few observations I have made, knowing as little as I do but having matches on FTDNA as well as on Ancestry. Maybe I'm missing something but I don't even expect to compare matching chromosome segments in detail unless I can use gedmatch. Obviously I can only use gedmatch, if the matches I'm interested in, have also uploaded there. This may be quite limiting but to me it's just a reality I have to accept.
                          Once you upload to gedmatch, you will also have another huge set of matches to consider and use and you can compare your matches there in a much more efficient way than in Ancestry or FTDNA.
                          Having made all the connections, comparisons, contacts and requests I can put together from Ancestry and FTDNA then I apply the tools on gedmatch. In this way, I have then had excellent success dealing with matching segments, determining family lines and establishing previously unknown ancestors.
                          One thing I learned from my experiences with Ydna at FTDNA, that may apply to finding several people with matching segments and no supporting paper trail, is that there are often "NPE's" in a person's ancestry. "Non Paternal Events", cause a person to believe in a certain paper trail that is not supported by dna evidence. In other words, many non recorded adoptions occurred and many children were fathered by someone other than the person who was recorded on paper as, or otherwise assumed to be, the father.
                          Just a few thoughts from someone who is also trying to understand a complex science.

                          Comment

                          Working...
                          X