Announcement

Collapse
No announcement yet.

Number of Y-DNA markers, statistics, etc.

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

  • Number of Y-DNA markers, statistics, etc.

    I think, maybe misquided, that I have a pretty good grasp regarding the science behind Y-DNA testing and the prediction algorithms for MRCA's etc. But I remain confused regarding how the notion of common ancestors is applied to different marker tests. For example, I have my results for all 37 markers via FT DNA. I match one person at 25 and a host of folks at one and two mutations from 25; however at 37 markers I only "match" other participants in all the databases at 32 markers or less. Therefore FT DNA reports that I have no matches at 37 markers...but they report the matches at 12 and 25 markers and present percents of probability of having a common ancestor at as high as 99%. How can this be when I match NOBODY within 5 markers at 37? Why don't the results at 37 "trump" all the other reports of matches resulting in me having absolutely no "matches" in the database? I hope I've explained this as well as possible and appreciate any info.

  • #2
    Matching at more markers increases the likelihood that the MRCA lived more recently. If you match someone at 25 markers and the probability of a MRCA within the last 8 generations is, say, 99%, then matching someone at all 37 markers might mean that the probability is now 99.99%. Neither is a guarantee; they are still statistics.

    The way the results are reported, if you have a 2-step match with someone at the 25 marker level and that person did not test 37 markers but only 25, then obviously that person will not appear under the 37 marker results. If the person did do the 37 test, but there are an additional 3 markers in the set beyond 25 that did not match with you--for a total of 5 mismatches out of 37--that person will also not appear in the 37 column. This does not mean that the 2-step match at the 25 level is not significant; it just means that the additional markers from 26 to 37 did not do anything to increase the probability of the MRCA being more recent.

    Comment


    • #3
      Thanks for the response! Just to be sure I have this right. If I have a perfect match with someone at 25 markers then FT DNA predicts a 61.17% probability that we have a common ancestor within the last 100 years. Next we are both tested to 37 markers and have none or only a very few additional matches. The probability then stays at 61.7% irrespective of the possibility of up to 12 mismatches of the 37 total?

      Comment


      • #4
        Originally posted by cycleman
        I think, maybe misquided, that I have a pretty good grasp regarding the science behind Y-DNA testing and the prediction algorithms for MRCA's etc. But I remain confused regarding how the notion of common ancestors is applied to different marker tests. For example, I have my results for all 37 markers via FT DNA. I match one person at 25 and a host of folks at one and two mutations from 25; however at 37 markers I only "match" other participants in all the databases at 32 markers or less. Therefore FT DNA reports that I have no matches at 37 markers...but they report the matches at 12 and 25 markers and present percents of probability of having a common ancestor at as high as 99%. How can this be when I match NOBODY within 5 markers at 37? Why don't the results at 37 "trump" all the other reports of matches resulting in me having absolutely no "matches" in the database? I hope I've explained this as well as possible and appreciate any info.
        most people have no matches at 25 or 37 . the 99% is probably at 600 yrs but you can match at 100 yrs my son is 33 at 100 and 97 at 600 well its 21years.

        Comment


        • #5
          The MRCA calculations are based on statistics as you may already know.

          I think I can answer your question indirectly by example:

          I match 12 out of 12 markers with about 30 people, but all with different surnames from mine (the MRCA assumes a common surname for the percentages). One of these we will call Bill Smith.

          Bill then upgraded his test to 25 markers he then only matched 22 out of 25. While my 4th cousin matched 11 out of 12 markers, when my cousin upgraded he matched 23 out of 25, markers and the markers that did not match were the more variable ones. Equally another person we'll call him Joe Jones matched 23 out of 25.

          So far my 4th cousin has not upgraded to 37 markers, but I have and there are no close matches with me anywhere.

          So in summary 12 marker results are only significant if the surname matches. Furthermore people with 11 out of 12 markers can be more closely related to you than the perfect matches.

          Two results with a match at 23 out of 25 are not necessarily equal in terms of relatedness.

          A close match at the 37 marker level will in all likelihood trump results at 12 and 25 markers. From a pure statistical point of view this should be true. It is important not to forget the surname.

          One last point (it was already made but I will repeat it), the statistics say there is only a 20% chance that my 4th cousin and I share a common ancestor in the last 200 years. When in fact I know that it is much closer to 100%, our common ancestor was born in the late 1700s.

          I hope this helps

          Comment


          • #6
            Originally posted by cycleman
            Thanks for the response! Just to be sure I have this right. If I have a perfect match with someone at 25 markers then FT DNA predicts a 61.17% probability that we have a common ancestor within the last 100 years. Next we are both tested to 37 markers and have none or only a very few additional matches. The probability then stays at 61.7% irrespective of the possibility of up to 12 mismatches of the 37 total?
            Hi Cycleman,

            No, No, No. Once you both have upgraded to 37 markers they are ones that determine TMRCA. The results of the 25 markers become a mute point once you both upgrade.

            The TMRCA tables for 12, 25, and 37 markers each assume that these are all the markers there are to be tested.

            The TMRCA for 12 markers assumes that there are ONLY 12 markers available for testing. The probability calculations are based on this assumption that there are only 12 markers to be tested. If there are only 12 markers and you match 12 for 12, there is a 50% probabilty that you share a common ancestor within 14 generations.

            If you both upgrade to 25 markers and match 23 for 25 there is now a 50% probability you share a common ancestor within 28 generations. The 12 marker TMRCA is no longer valid and means absolutely nothing. It is superceded by the 25 marker TMRCA table.

            If you both upgrade to 37 markers and match 32 for 37 there is now a 50% probability you share a common ancestor within 42 generations. The 12 marker TMRCA and the 25 marker TMRCA are no longer valid and mean absolutely nothing. They are superceded by the 37 marker TMRCA table.

            John

            Comment


            • #7
              Thanks John....but why does FT DNA continue to list my "matches" at 25 complete with a link to calculate the MRCA, etc.? I have no matches at 37 but am listed for 20+ matches at 25 markers. What you say certainly makes sense and is compatible with my first assumption that the 37 test would "trump" the 12 and 25 test; however, FT DNA certainly doesn't make this clear by continuing to list ones matches and presenting statistics indicating that they are "genetic cousins."

              Comment


              • #8
                cycle
                if you have a strong relationship it will be there in all three
                you wont be12-12 18-25 35-37 that 18-25 has to be 23-25

                i know that seems simplistic it is sometimes you dont see the trees thru the brush

                Comment


                • #9
                  Originally posted by jablair
                  Hi Cycleman,

                  If there are only 12 markers and you match 12 for 12, there is a 50% probabilty that you share a common ancestor within 14 generations.

                  John
                  I'm a newbie so please excuse me if this has been asked hundreds of times before. I have just received my twelve markers results which, I'm sorry to say, show little in common with most of my fellow Clan Project members. A serach. however, has revealed 7 others with all 12 markers identical. None has the same surname and geographically they seem very spread out. On the basis of the above responses is this meaningless unless I upgrade to 25 or 37 markers?

                  Comment


                  • #10
                    Jock,

                    A 12 for 12 match with someone of a different surname is basically useless unless there is compelling evidence, based on conventional research, that you are connected to that surname. In fact, a 12 for 12 match with someone with your surname is far from conclusive unless there is a paper trail indicating the 2 of you are related. I would strongly recommend that someone with a 12/12 upgrade to 25 markers, if they have the same surname, but I would not encourage upgrading to 25 markers just to go a "fishing trip" to see if I match with someone of a different surname, unless there is a specific reason to think you are connected to that surname.

                    The history of the Blair surname indicates a number of name changes, i.e. a Blair male taking a different surname or some other male taking the Blair surname. I do watch certain other Surname Projects where I know there might be a connection.

                    John

                    Comment


                    • #11
                      Originally posted by jablair
                      Jock,

                      A 12 for 12 match with someone of a different surname is basically useless unless there is compelling evidence, based on conventional research, that you are connected to that surname. In fact, a 12 for 12 match with someone with your surname is far from conclusive unless there is a paper trail indicating the 2 of you are related. I would strongly recommend that someone with a 12/12 upgrade to 25 markers, if they have the same surname, but I would not encourage upgrading to 25 markers just to go a "fishing trip" to see if I match with someone of a different surname, unless there is a specific reason to think you are connected to that surname.

                      The history of the Blair surname indicates a number of name changes, i.e. a Blair male taking a different surname or some other male taking the Blair surname. I do watch certain other Surname Projects where I know there might be a connection.

                      John

                      there are some places i disagree with john but i agree with most of it
                      i went to 37 because people will test who i have no idea exist. so i am baiting the trap how else will know if i dont do 37

                      also being in the same project effects the computing of the closeness . this is why i have the icon with my son in denning and i dont with a member of the chelsea or my son doesnt in frosinone unless i change the name on the web personal web page [gap] at that time the icon switches to member who match in another project

                      also about matching 12/12 s i have about 50+12/12s half are askenazi THAT MEANT SOMETHING. when i started to see 25 close matches with historys of forced conversions it meant more

                      Comment


                      • #12
                        Wrong calculation TMRCA?

                        I got 10 exact matches on 12 markers test.
                        The calculater show that there is a 50% probility that each
                        of these matches is 15 generations to TMRCA.

                        But, when I upgrade to 25 marker, I got only 18-21 markers match.
                        The calculator show that there is a 50% probility of these matches
                        40 to 50 generations to TMRCA.

                        Don't say that I have to share a Surname with the persons who have
                        the exact matches! The input to the calculator has no parameter
                        for surename. Calculation is based on pure statistical data for the
                        mutate rate in the Y-chromosom.

                        Comment

                        Working...
                        X