Announcement

Collapse
No announcement yet.

A modest proposal for FTDNA

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

  • A modest proposal for FTDNA

    I find that the nomenclature used among various customers of atDNA (Family Finder) when they contact me, reflects what I consider a misleading label on the DASHBOARD. I would suggest that there is no such thing as a MATCH in Family Finder, no one can actually match anyone else (and let's not get into an angels-on-pinheads argy-bargy about identical twins.)

    I would prefer that the button under Family Finder read as SHARES.

    There is enough lack of precision in interpreting theses results, without suggesting that these results are anything but a report of shared chromosome segments: they can certainly reinforce traditional genealogy, and can break "brick walls" in some cases, but suggesting that the people we find in Family Finder are MATCHES has a connotation that suggests more certainty than there actually is.

    I applaud FTDNA for using centiMorgans as the reporting value instead of the ambiguous language used on that other site, but I find that the term MATCHES has far more validity in YDNA than in autosomal DNA.

  • #2
    I would leave matches, but stressed the probability angle. And its dependency on shared centiMorgans and other factors. Right now that is deeply buried in the Family Tree DNA Learning Center and treated very lightly.

    Why?

    Knowing about the shared DNA, one has then to figure out whether the other person fits (matches) into their family tree or more precisely one of their family trees.

    Merry Christmas - Mr W

    P.S.
    And then some come here to learn about Distribution of genealogical relationships for given amounts of shared DNA initially from http://isogg.org/wiki/Autosomal_DNA_statistics

    Comment


    • #3
      Why leave MATCHES when they are not?

      The subject says it all.

      Comment


      • #4
        Continuing thread

        I respectfully submit that words have meaning. Using a term for convenience when another more appropriate term is more on point, leads to fuzzy thinking. Logic and reason do have a place in these discussions. If someone sincerely thinks that MATCHES is a better term for atDNA, please explain.

        The fact that nearly everyone has to post a link to ISOGG in all of these interchanges tells it all.

        Using the term SHARES makes it more clear that the larger the share, the more likely to be worth investigating. You cannot truly have different MATCHES, you either match or you do not. As stated, in the context of Y DNA, MATCHES is more intuitive, because that is what you are looking for, and you can find them, but in atDNA, it is counterintuitive and therefore does not aid in clarifying the nature of atDNA.
        Last edited by clintonslayton76; 28 December 2015, 10:13 AM. Reason: corrections

        Comment


        • #5
          It is the same with Y-DNA. Two men share STRs (or SNPs) so they match each other. We have no idea whether they are identical twins, or 5th cousins, etc., probabilistic/statistical interpretation can be used.

          There could be a partial match, when two men do not share all tested STRs (or SNPs). A partial match is still a match. However, that more clearly leads to the probabilistic/statistical interpretation of the results.

          With autosomal DNA, there are just more SNPs tested as compared to Y-DNA. In my opinion, that does not significantly change meaning of sharing and matching.

          Mr W

          P.S.
          I am in no way representing FTDNA, nor influencing their decisions, so I will not continue in this thread after this post.

          Comment


          • #6
            Originally posted by clintonslayton76 View Post
            I respectfully submit that words have meaning.
            Using the term SHARES makes it more clear that the larger the share, the more likely to be worth investigating. .
            There is another company that uses the term "sharing genomes." That unfortunately can imply to newbies that they are sharing private genetic information and many opt not to compare. For that reason, I think that "match," although possibly less accurate, is preferable to "share."

            Comment


            • #7
              I realize...

              that with another major service called GEDMatch that I am micturating in the wind, but I still do not like the terminology. I am jumping over this dead horse and moving on.

              Comment


              • #8
                Personally, I think the word "Match" is more accurate than the word "Share" - the DNA in me is only in Me - it is not also in you; I am not sharing my body, nor the DNA in it - which is an actual physical part of me - with anyone else. "Share" might be accurate if I donate an organ to be transplanted into you - otherwise, no. What we have is a partial Matching.

                Comment


                • #9
                  Technically speaking nobody here can prove a segment match. The match is based only on a set of markers that can be on either chromosome in a pair. Many people prefer the term HIR (half identical region). Well you can’t even prove the region is half-identical because important SNPs could also be missing within the Illumina chips. Not all markers are tested, and even the ones that are tested are not always correct. A certain number of mismatches can be tolerated.

                  Even phasing doesn't prove a match. You can separate out the sequence on each chromosome and still not have an identical block because you have not shown every position is identical between those tested. It is true that phasing and increased size both improve the probability of IBD (identical by descent) for our purposes.

                  A match is really only a best guess based on the markers tested. We have not come up with better terminology. Fortunately our success stories show that this "matching" can have utility.
                  Last edited by Kathy Johnston; 2 January 2016, 08:49 AM.

                  Comment


                  • #10
                    Kathy stated "because you have not shown every position is identical between those tested"

                    Triangulation, the process by which we compare a large segment among a large group of people who purportedly match it, to make sure they all match each other at the same spot on every pairing, does ensure that the segment is real and belongs to the whole group.

                    At the present time, that requires asking our matches who are ICW (in common with) each other here to look in their personal chromosome browsers to see if they match our matches at the same spot, or alternatively everybody uploading our data to GEDmatch to verify on there.

                    Comment


                    • #11
                      re: Loobster's objection, and...

                      Kathy Johnson's comment that what is there seems to be working do not overcome my objections, but some cases of institutional lethargy are somewhat justified if any other alternative would cause even more confusion and thus less usage by being obscure. I have attempted to convey to my Project members the theoretical limits of atDNA on the issue of "proving a segment match," but as she states, this terminology works by FTDNA's lights. By my lights, this is "a high probability of a segment match leading to a possible share that might be investigatable through genealogy."

                      (The FF DNA that is IN YOU is never visible to you on FF. Perhaps Kathy can explain how any contributor can <i>ever see what their atDNA account looks like in FF?</i> For instance, I had to test my son in order to be able to log in as him to confirm if my posted Family Tree was really at the privacy level that I wished. There has to be a better way.)

                      Loobster, the atDNA that is IN YOU does not invalidate that the atDNA IN YOU is in fact a SHARE with those who are genetically in your family. It is not only IN YOU but is in fact SHARED back to the threshold of measurement with all of your ancestors and current "blood" relatives. If we did not suspect that, we would not be on this board because there would be no DNA testing for genetic genealogy. I see no connection between the term in this context and sharing organs, etc.

                      As stated, it was not my intention to keep beating a dead horse, but I am glad that the issue has aroused comment.
                      Last edited by clintonslayton76; 3 January 2016, 08:31 AM. Reason: clarification

                      Comment


                      • #12
                        I prefer "match" to "share". And I hate "sharing genomes" because it sounds like you and the other person are going to be able to see one another's entire genome (which they don't even test, of course), rather than just the tiny segment or segments you share. I can see why people with little understanding of the test would refuse to share genomes!!!

                        Comment


                        • #13
                          Originally posted by khazaria View Post
                          Kathy stated "because you have not shown every position is identical between those tested"

                          Triangulation, the process by which we compare a large segment among a large group of people who purportedly match it, to make sure they all match each other at the same spot on every pairing, does ensure that the segment is real and belongs to the whole group.

                          At the present time, that requires asking our matches who are ICW (in common with) each other here to look in their personal chromosome browsers to see if they match our matches at the same spot, or alternatively everybody uploading our data to GEDmatch to verify on there.
                          We assume that a "segment" is really a "match" based on the identification of a matching marker at each position tested but only a fraction of SNPs are actually tested. Take a look at public databases like hapmap.org, ALFRED, 1000 genomes project etc. There you will see how many SNPs (mutations that are variable) are missing from our raw data. There is no guarantee we are covering all the important ones.

                          What I find interesting is that some of the no-calls will be the same within a triangulated group. Those should probably be investigated further because there is likely an adjacent SNP that triggered that no-call. I assume these are recent, regional or private SNPs that are probably significant.

                          Comment


                          • #14
                            Originally posted by clintonslayton76 View Post

                            the atDNA IN YOU is in fact a SHARE with those who are genetically in your family. It is not only IN YOU but is in fact SHARED back to the threshold of measurement with all of your ancestors and current "blood" relatives. If we did not suspect that, we would not be on this board because there would be no DNA testing for genetic genealogy. I see no connection between the term in this context and sharing organs, etc.
                            OK, I'm lost -- What do you mean by "SHARE" and "SHARED"? --
                            (I thought what we were talking about was basically that DNA - or DNA segments or parts of DNA -- in two folks appeared to match and so the DNA in the two folks probably came about thru descending from the same ancestors - but to me, this is MATCHING of the DNA suggesting sharing of ancestors.)
                            Last edited by loobster; 3 January 2016, 04:14 PM.

                            Comment


                            • #15
                              Originally posted by clintonslayton76 View Post
                              [----] The FF DNA that is IN YOU is never visible to you on FF. Perhaps Kathy can explain how any contributor can ever see what their atDNA account looks like in FF? [----]
                              The first sentence is not true.

                              Any contributor can download her or his Family Finder results and see the values (in a spreadsheet).

                              Mr W

                              Comment

                              Working...
                              X