Announcement

Collapse
No announcement yet.

46 marker y-DNA match but MRCA way off

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

  • 46 marker y-DNA match but MRCA way off

    Several years ago, I submitted a DNA sample to have a 33-marker y-chrome test completed. A month or so ago, I succeeded in convincing a male with whom I share a surname to submit a sample as well. We always suspected we were closely related, and the test results (posted on Ancestry.com) suggested our MRCA was as recent as 3 generations ago.
    The problem is, we both have indisputable proof that or MRCA must have been at least 6 generations ago, inasmuch as we both have paper trails identifying our ancestors back through five gens.
    In response to this odd result, I opted for an "upgrade" of my older results to a 46-marker y-chrome test. When the new results were posted, this fellow and I emerged with exact matches on all 46 markers, and Ancestry now computes our MRCA to be 1 generation ago.
    That is patently false.
    Can anyone out there help me figure this out? Could it be that, through at least six generations, our y-DNA has either undergone zero mutations, or it has undergone exactly the same mutations? What are the chances of that?

  • #2
    When you open the FTDNATiP, what are the percetages listed for the MRCA between the two of you when you enter in that you do not share an ancestor in the last 6 generations?
    Last edited by kokeb; 17 May 2011, 12:21 PM.

    Comment


    • #3
      Originally posted by kokeb View Post
      When you open the FTDNATiP, what are the percetages listed for the MRCA between the two of you when you enter in that you do not share an ancestor in the last 6 generations?
      I'm new to this site and this forum. What is a FTDNA TiP? and how do I open it?

      Comment


      • #4
        Originally posted by MastoDon View Post
        I'm new to this site and this forum. What is a FTDNA TiP? and how do I open it?
        To the far right of the matches name there should be a blue and orange icon. Click on it to open the FTDNATiP menu. This tool allows you to calculate the MRCA percentages.

        Comment


        • #5
          Hi MastoDon,

          33 and 46 is an actual Ancestry test, not a Family Tree DNA test, right?

          When I had 67 FTDNA markers tested and entered them manually at Ancestry, it gave me hundreds of matches within the last 250-500 years. Most had tested the Ancestry 33 panel.

          When I got my additional nine markers from a free SMGF kit (these are also tested at Ancestry) and added them, I had zero matches.

          So I suspect the Ancestry MRCA calculator is just waaaaay crummy.

          Comment


          • #6
            Originally posted by MastoDon View Post
            Several years ago, I submitted a DNA sample to have a 33-marker y-chrome test completed. A month or so ago, I succeeded in convincing a male with whom I share a surname to submit a sample as well. We always suspected we were closely related, and the test results (posted on Ancestry.com) suggested our MRCA was as recent as 3 generations ago.
            The problem is, we both have indisputable proof that or MRCA must have been at least 6 generations ago, inasmuch as we both have paper trails identifying our ancestors back through five gens.
            In response to this odd result, I opted for an "upgrade" of my older results to a 46-marker y-chrome test. When the new results were posted, this fellow and I emerged with exact matches on all 46 markers, and Ancestry now computes our MRCA to be 1 generation ago.
            That is patently false.
            Can anyone out there help me figure this out? Could it be that, through at least six generations, our y-DNA has either undergone zero mutations, or it has undergone exactly the same mutations? What are the chances of that?
            I'm not familiar with how Ancestry.com shows the results for matches, so I can only speculate. I uploaded my own results there, but have zero matches. The wording they use to explain the concept of MRCA might be confusing or misleading. Can you copy and paste the exact wording from your match?

            Calculations for MRCA are based on probabilities, so there's no way to definitively say it is within so many generations. It can only tell you the chance that's the case. The FAQ here has a table which explains how closely related you and your match could be, based on the genetic distance (GD) and number of markers tested. For 37 markers (the closest to 33), a GD of 0 means a 50% chance you share an ancestor within 5 generations, and 95% chance within 8 generations. I'm not sure how much the probability changes with 46 markers, which is in between 37 and 67, and which markers are tested could make a difference, since they have different observed mutation rates. But at 67 markers, the probabilities go up to 50% chance within 3 generations, and 90% chance within 5 generations.

            Originally posted by MastoDon View Post
            I'm new to this site and this forum. What is a FTDNA TiP? and how do I open it?
            That's unique to FTDNA, so it doesn't apply to your results, since you tested with Ancestry.com.

            Comment


            • #7
              Originally posted by nathanm View Post
              That's unique to FTDNA, so it doesn't apply to your results, since you tested with Ancestry.com.


              Yes. Sorry, I didn't see that you were talking about Ancestry.com. The info I was asking for and providing is for FTDNA only.


              As far as Ancestry.com's calculations... I posted my results and my father's results to their database and they have my father matching me at 4 generations instead of 1. I prefer FTDNA's MRCA calculations...though I know that won't help you much if both of your results are not in this database.
              Last edited by kokeb; 17 May 2011, 01:37 PM.

              Comment


              • #8
                Originally posted by kokeb View Post
                Yes. Sorry, I didn't see that you were talking about Ancestry.com. The info I was asking for and providing is for FTDNA only.


                As far as Ancestry.com's calculations... I posted my results and my father's results to their database and they have my father matching me at 4 generations instead of 1. I prefer FTDNA's MRCA calculations...though I know that won't help you much if both of your results are not in this database.
                Yes, I am underwhelmed by the Ancestry commitment to usefulness. I was able to hand-copy my 46-marker results from Ancestry to the SMGF in order to do a search for matches. Does FTDNA allow the same flexibility (that is, for a person to import DNA results from a "competitors" source)?

                Comment


                • #9
                  Originally posted by MastoDon View Post
                  When the new results were posted, this fellow and I emerged with exact matches on all 46 markers, and Ancestry now computes our MRCA to be 1 generation ago.
                  That is patently false.
                  Does Ancestry give a range of generations? One generation would be the "Maximum Likelihood Estimate" or "Most Likely Estimate" -- more brothers will match exactly than first cousins, and more first cousins than second cousins, and so on down the line.

                  But the absolute probabilities are rather low, so another common measure is how many generations do you need to have a 50% or 95% chance of identifying the common ancestor.

                  Comment


                  • #10
                    Originally posted by Ann Turner View Post
                    Does Ancestry give a range of generations? One generation would be the "Maximum Likelihood Estimate" or "Most Likely Estimate" -- more brothers will match exactly than first cousins, and more first cousins than second cousins, and so on down the line.

                    But the absolute probabilities are rather low, so another common measure is how many generations do you need to have a 50% or 95% chance of identifying the common ancestor.

                    The Ancestry.com results just give you a single column called "MRCA" in your results list. In the column is a single number that is supposed to represent the MRCA generation value; no percentage is given (at least I don't see any on the results page or any link to send me to such information.)
                    Last edited by kokeb; 17 May 2011, 01:57 PM.

                    Comment


                    • #11
                      Originally posted by MastoDon View Post
                      Yes, I am underwhelmed by the Ancestry commitment to usefulness. I was able to hand-copy my 46-marker results from Ancestry to the SMGF in order to do a search for matches. Does FTDNA allow the same flexibility (that is, for a person to import DNA results from a "competitors" source)?
                      Not at FTDNA directly, but they run a free database called Ysearch. Unfortunately, it doesn't match against their whole database, just people who have transferred their results. But it's better than nothing, and matches are more likely to respond, since they're pro-active.

                      Comment


                      • #12
                        This is curious, but I don't think a problem necessarily.

                        Mutations are random. They're based on typical observed patterns in a very large population--specific individuals within that population may vary considerably. There seems little doubt that you two are related--it's only a question of precisely how closely related you are.

                        And no one who works with DNA tests will honestly and in good conscience claim to you that it represents a very precise clock. There's just too much variability among individuals within the population.

                        Now the fact that you have a 2nd set of tests that project a dramatically closer relationship is very interesting, and does suggest that one or the other or maybe even both of the tests contain a reporting error. But again, given the range of normal variation of individuals within the population, I don't think either result is totally out of the question--and both clearly show that you two are related fairly closely.

                        Anecdotally, I've heard from one group of fellahs, "A", "B" and "C" that show a genetic distance ("GD") pattern that is as weird as anything you've presented here. "A" and "B" share an ancestor in colonial America and their GD is consistent, something like GD of 5 at 67. BUT "A" and "C" are 100% exact matches -- GD of 0 at 67--and "C" is a Brit who cannot possibly have a common ancestor with "A" after 1600. All the same rare Yorkshire surname.

                        Just pure-D freaky. There are mutliple tests done here, and the odds of the precise range of mistakes being made on the tests of both "A" and "C" seems even less likely than the odds that these GD are accurate.

                        Not to take away from your concern about the discrepancies between the tests, but overall, I wouldn't be too concerned that neither test says a MRCA precisely 3 generations ago.


                        Originally posted by MastoDon View Post
                        Several years ago, I submitted a DNA sample to have a 33-marker y-chrome test completed. A month or so ago, I succeeded in convincing a male with whom I share a surname to submit a sample as well. We always suspected we were closely related, and the test results (posted on Ancestry.com) suggested our MRCA was as recent as 3 generations ago.
                        The problem is, we both have indisputable proof that or MRCA must have been at least 6 generations ago, inasmuch as we both have paper trails identifying our ancestors back through five gens.
                        In response to this odd result, I opted for an "upgrade" of my older results to a 46-marker y-chrome test. When the new results were posted, this fellow and I emerged with exact matches on all 46 markers, and Ancestry now computes our MRCA to be 1 generation ago.
                        That is patently false.
                        Can anyone out there help me figure this out? Could it be that, through at least six generations, our y-DNA has either undergone zero mutations, or it has undergone exactly the same mutations? What are the chances of that?

                        Comment


                        • #13
                          Mcra

                          This is one I'd like to figure out! I've also got a 46/46 match on Y-dna with a cousin where our common ancestor is known to be 12 generations back. I just started the ftdna 67-tests and I'm beginning to get a clearer picture of the statistical probabilities. I am becoming convinced that some of my "close" matches are farther back than 400 years.

                          Comment


                          • #14
                            Originally posted by MastoDon View Post
                            Can anyone out there help me figure this out? Could it be that, through at least six generations, our y-DNA has either undergone zero mutations, or it has undergone exactly the same mutations? What are the chances of that?
                            Yes it is possible. Mutation rates are based on averages. If you flip a coin 10 times, on average you'll have 5 heads and 5 tails, but it is possible to get either 10 heads or 10 tails. If you did this only 2 times and the first time you got all heads and the second time you got all tails your average would be 5 heads and 5 tails.

                            So your question could be, is it possible that we aren't average? And the answer is yes.

                            Comment

                            Working...
                            X