Announcement

Collapse
No announcement yet.

Improve X matching

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

  • Improve X matching

    Still disappointed with what FTDNA calls "X-match" (all of mine are less than 5 cM and more than likely noise), and that a match to someone only on the X chromosome will not appear in FamilyFinder. I've made a couple of important breakthroughs with matches in 23andMe where we share very substantial segments on the X with no autosomal segments. One of them shares a 72 cM X segment with me and no autosomal segments. I'd love to know if any others who match me only on the X have tested at FTDNA and if I can confirm the source of the segments through triangulation. On the one hand FTDNA has argued that X matches are too distant to be helpful, yet they flag any autosomal match as an "X-match" when it really isn't. Doesn't make sense to me.
    Last edited by ToddH; 24th January 2019, 01:20 PM.

  • #2
    Would love an example of a 72 cM X match with no matching autosomal segments.

    Comment


    • #3
      Here's what it looks like. The match is a 6th cousin:


      Screen Shot 2019-01-24 at 3.43.16 PM.png

      Comment


      • #4
        Originally posted by ToddH View Post
        Still disappointed with what FTDNA calls "X-match" (all of mine are less than 5 cM and more than likely noise), and that a match to someone only on the X chromosome will not appear in FamilyFinder. ... On the one hand FTDNA has argued that X matches are too distant to be helpful, yet they flag any autosomal match as an "X-match" when it really isn't. Doesn't make sense to me.
        On FTDNA, total cMs shared is the shared cMs on the autosomal chromosomes, it does not include any cMs shared on the X-Chromosome
        - their listing of an X-match is simply letting you know that in addition to the total cMs shared on the autosomal chromosomes, there is also some cMs shared on the X-chromosome.
        Last edited by loobster; 24th January 2019, 06:49 PM.

        Comment


        • #5
          Originally posted by loobster View Post

          On FTDNA, total cMs shared is the shared cMs on the autosomal chromosomes, it does not include any cMs shared on the X-Chromosome
          - their listing of an X-match is simply letting you know that in addition to the total cMs shared on the autosomal chromosomes, there is also some cMs shared on the X-chromosome.
          Yes I know and I'm saying it is very unhelpful. The "X-matches" I have share X segments of under 3 cM, or noise. They shouldn't be counted as X matches at all, it is misleading. A decent X match should start at around 10 cM. And we should be allowed to have matches who share significant X-DNA with no significant segments on the autosomes. It should be easy to do.

          Comment


          • #6
            There is a difference between "doesn't make any sense", on the one hand and "doesn't seem ideal", "is not the optimal way to do things", "isn't what I want", etc. on the other.

            What FTDNA is doing with regard to X-matches makes sense. It just is not what you think they should be doing.

            FTDNA's saying you have an "X-match" when the amount shared on the X is small is misleading only because you CHOOSE to interpret "X-Match" to mean a significant Match on X -- when that is NOT what they are saying.

            Comment


            • #7
              In my opinion, the X matches that FTDNA alerts us to need to be segments that (1) are long enough, (2) have enough SNP's, and (3) are sufficiently unlikely to arise by chance alone (in the sense that huge numbers of people might happen to have that same sequence?), so that there is a reasonable expectation that they might be worth following up. Whether FTDNA is meeting those criteria with it's current practices, I can't tell. There just isn't enough information. But, based on experience with autosomal matches, I have some doubts.

              Comment


              • #8
                I think FTDNA should only report X matches if they are long enough to be meaningful. By doing it the way they are, they confuse people who know very little or nothing about DNA and who jump to false conclusions when they see X matches.

                The only branch of her family tree where Mom can't have an X match is the Colonial American side and that's where all her X matches are, some sharing several small segments. Beyond family members, she has only one significant X match. When I saw that one, I knew she had to be a descendant of one of Mom's recent French or German immigrant ancestors.

                There isn't even a way to sort on X segment length to find significant matches.

                Comment


                • #9
                  Good point on the inability to sort the X matches! In the absence of evidence to the contrary, and in spite of the different mechanism of inheritance, I would like to see X matches handled with the same matching and inclusion criteria, and the same kind of tools, as ordinary autosomal matches. I don't see an evidence-based reason that they should have different criteria for minimum length, minimum number of SNP's, etc., and we should be able to sort them, triangulate them, and otherwise play with them just as we do with autosomal segments. The only difference should be in the genealogical interpretation that we place on them.

                  Comment


                  • #10
                    Originally posted by loobster View Post
                    There is a difference between "doesn't make any sense", on the one hand and "doesn't seem ideal", "is not the optimal way to do things", "isn't what I want", etc. on the other.

                    What FTDNA is doing with regard to X-matches makes sense. It just is not what you think they should be doing.

                    FTDNA's saying you have an "X-match" when the amount shared on the X is small is misleading only because you CHOOSE to interpret "X-Match" to mean a significant Match on X -- when that is NOT what they are saying.
                    You are attacking my wording while ignoring the point I am making. What do you think of my point? Have you been able to make constructive use of what FTDNA is calling an "X-match"? Do you think they could improve the way they present "X-matches"?

                    Again, I know exactly "what they are saying" when they label a match as an X-match. It is not a mystery. It STILL doesn't make sense to me that they would use that label for bits of DNA that are noise. It doesn't make sense that this company would not allow you to see matches that share significant segments with you only on the X, while most matches share significant segments on only one autosomal chromosome (1-22).

                    I was part of a group of people who pressured FTDNA a few years ago to lower their total cM threshold, because it was preventing matching between people of different ethnic backgrounds in many cases. It "didn't make sense" to me for them to set the threshold so high and prevent matching. They asked me for examples of what was happening which I and several others did. They examined the situation and complaints and eventually lowered their threshold. I am hoping they will listen to complaints about how their X matching is not serving customers well. A simple adjustment to their algorithm would make a big difference for the better.
                    Last edited by ToddH; 7th February 2019, 07:50 PM.

                    Comment

                    Working...
                    X