Announcement

Collapse
No announcement yet.

Double First Cousins Shared DNA?

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

  • Double First Cousins Shared DNA?

    I know this question is answered widely on line as being 25% but that seems wrong when I do the rough calculations in my head.

    Half siblings typically share between 20-30% of their DNA. Half siblings share two of their four grand parents.

    Double first cousins share all four of their grand parents. how can they also be said to share roughly 20-30% of their DNA given that? It seems like something in the range of 32.5-42.5% would make more sense.

    Does anyone have real world numbers for double first cousins? Or an explanation for why they would one average share the same amount of DNA as half siblings?

  • #2
    Think of Double Cousins (all 1st cousins) in they DNA they share of their Great Grandparents.

    They(Double Cousins) may share 4 Grandparents, but their parents received different mixtures of their 4 Grandparents(Cousins Great Grandparents).
    Last edited by prairielad; 31 January 2014, 11:56 AM.

    Comment


    • #3
      Understood but the same is true of half siblings, only half siblings share only two grandparents.

      What I'd really love to see are some real world numbers.

      Comment


      • #4
        Two of my double 1st cousins have tested. With one I share a total of 1511.22 cM (longest segment 105.55 cM). With her brother I share 1287.81 cM (longest segment 70.99 cM).

        Comment


        • #5
          Thank you.

          Interesting, if I'm doing the math right one would be a 42% match and the other 36%?

          If I'm right on the math then that would be right in line with what I'd expect, maybe a bit on the high side.

          I'm going to try to work through the logic here...

          Full siblings on average share 50% of their DNA. In the average case double first cousins could therefor share between 0-50% of the DNA they get from each parent. On the extreme ends this means you could have a double first cousin with no genetic relationship on the one end or essentially an identical twin on the other.

          So, in the average case each parent of the double first cousins shares 50% of their DNA with their sibling, or roughly 1800cM. The remaining DNA for each sibling makes up roughly 1800cM as well. So the total pool is roughly 7200cM with each half of the pool being discreet. What this means is that there is a one in two chance of any particular segment being inherited by each of the double first cousins. Given those odds the expectation should be a number much higher than 25% shared.

          I think that kind of makes sense but even with a couple of data points I can't believe so many people could have been getting this wrong.

          Any other thoughts or data points?

          Comment


          • #6
            Hmm, it strikes me that part of the problem here is that half identical and full identical matches tend to be talked about interchangeably and they aren't the same. In reality don't we have closer to 7200cM of total autosomal DNA that can be shared? It seems like it would be less confusing if we always talked about the total number. Half siblings will typically have around 1800cM, all half shared (assuming no other close relationship) while full siblings having around 3600 with some of that being fully shared and the rest being half shared.

            Somebody please help me here. I feel like I'm getting further and further out on the limb and I'm not sure how far down the ground is. :-)

            Comment


            • #7
              Originally posted by 1_mke View Post
              Hmm, it strikes me that part of the problem here is that half identical and full identical matches tend to be talked about interchangeably and they aren't the same. In reality don't we have closer to 7200cM of total autosomal DNA that can be shared? It seems like it would be less confusing if we always talked about the total number. Half siblings will typically have around 1800cM, all half shared (assuming no other close relationship) while full siblings having around 3600 with some of that being fully shared and the rest being half shared.

              Somebody please help me here. I feel like I'm getting further and further out on the limb and I'm not sure how far down the ground is. :-)
              If you are talking about real world results from the FTDNA Family Finder test then no, those numbers will not work. The total of both lineages would be 6770cM summed.

              Since FTDNA does not count full identical matches twice but rather counts them like half identical matches, double cousins will end up sharing less segments cM then real they would if the full and half segments were accounted for separately. This is true for full siblings in the FTDNA FF test as well. Full siblings sharing 55% may end up appearing to share only 37% in the FF test.

              Comment


              • #8
                Originally posted by 1_mke View Post
                Hmm, it strikes me that part of the problem here is that half identical and full identical matches tend to be talked about interchangeably and they aren't the same. In reality don't we have closer to 7200cM of total autosomal DNA that can be shared? It seems like it would be less confusing if we always talked about the total number. Half siblings will typically have around 1800cM, all half shared (assuming no other close relationship) while full siblings having around 3600 with some of that being fully shared and the rest being half shared.

                Somebody please help me here. I feel like I'm getting further and further out on the limb and I'm not sure how far down the ground is. :-)
                You can't use total numbers for cousins and siblings because both are not compared to just one grandparent (i.e. 25%) but rather are composite results of sharing with multiple grandparents in sometimes the same locations (full instead of half identical regions).

                If you wanted to use a total number, it would be the original sibling share between the ancestors of the branches. Those two siblings form the basis of the SNP pool that can be compared in their descendants.

                Comment


                • #9
                  Matt,
                  If I'm understanding you correctly it seems like the root problem is the half versus full sharing and the fact that most places seem to treat both the same when they are in fact not.

                  Because of this double first cousins would show as sharing more DNA than half siblings because double first cousins would have sections of DNA where they matched on both halves while half siblings would not?

                  Comment


                  • #10
                    Originally posted by 1_mke View Post
                    Matt,
                    If I'm understanding you correctly it seems like the root problem is the half versus full sharing and the fact that most places seem to treat both the same when they are in fact not.

                    Because of this double first cousins would show as sharing more DNA than half siblings because double first cousins would have sections of DNA where they matched on both halves while half siblings would not?
                    This chart helps me get my head around it... Doesn't include doubles, though.

                    Bob
                    Attached Files

                    Comment


                    • #11
                      Double first cousins should be twice the cousin number, less 1/64th or 1.56% [the overlap]. Around 23-24%. I think I got that right...

                      Bob

                      Comment


                      • #12
                        That is a good chart. Thank you.

                        Is seems like my basic premise is correct. Double first cousins are not the same as Uncle/Nephew and half siblings from a DNA perspective even though they all get lumped in the "25% alike" bucket.

                        Comment


                        • #13
                          Originally posted by 1_mke View Post
                          Matt,
                          If I'm understanding you correctly it seems like the root problem is the half versus full sharing and the fact that most places seem to treat both the same when they are in fact not.

                          Because of this double first cousins would show as sharing more DNA than half siblings because double first cousins would have sections of DNA where they matched on both halves while half siblings would not?
                          No its not a problem unless one is trying to calculate % shared instead of cM and then only siblings really share significant sizes and amounts of full identical regions.

                          When I compare my granddaughter to her full sibling brother, they share tons of FIR segments but when I compare her to my other grandson, her cousin, they only share one FIR segment and it is a small one. I'm sure that would hold true if I kept checking the other grandkids too.

                          It is really only full siblings where FIR calculations are needed and FTDNA does not report % shared anyway so it really doesn't matter for them at this point, not until they want to report % shared that is. Taking all the FTDNA tests in the system it is easy to find the nominal ranges where full siblings share and then say with confidence that when the shared amount is xxcM, those two people are predicted as full siblings.

                          I guess I don't know what you are trying to determine. I thought you were just not able to make % shared amounts work for you and if taking into account full sibs, the reason is the lack of using FIR segments in the FTDNA comparison. Cousins, double or single, will not share enough FIR segments to really affect a % shared calculation. Of course double cousins may share some FIR segments but most of the time they will simply share more HIR segments because there are two more grandparents to compare against in the mix.

                          Matt.

                          Comment


                          • #14
                            Originally posted by 1_mke View Post
                            I know this question is answered widely on line as being 25% but that seems wrong when I do the rough calculations in my head.

                            Half siblings typically share between 20-30% of their DNA. Half siblings share two of their four grand parents.

                            Double first cousins share all four of their grand parents. how can they also be said to share roughly 20-30% of their DNA given that? It seems like something in the range of 32.5-42.5% would make more sense.

                            Does anyone have real world numbers for double first cousins? Or an explanation for why they would one average share the same amount of DNA as half siblings?
                            I've went back to see what you were originally asking.

                            There is a difference between shared and inherited, especially cousins. Cousins share DNA when they both have the same segment from the same grandparent in the same location. Everything is a matter of recombination and how all of the unique ancestral segments map out on the grandchildren's chromosomes.

                            % shared for siblings, cousins and so on is just an estimate based on the expected amount inherited. People do not always inherit exactly 25% from each grandparent and they do not always inherit the same segments from the grandparent that their siblings do, which is why siblings do not share 100% of each side with each other. There is a ton of variation in how people compare to their grandparents and cousins and the variation is based solely on the fact that recombination causes every person's inheritance to be unique.

                            I think you see that as you are saying that half siblings share something like 30% of their DNA. My half sister and I share 1540cM, or calculated that is 22% shared.

                            Probably the difference in what you expect and what you see is simply that the extra recombination event for cousins strongly affects their ability to share full identical segments in the first place as they are not comparing between four grandparents like full siblings would, or two grandparents like half siblings would, but comparing between four grandparents with half of the information removed and since half of the information is removed differently per side, they really have a lot less places they can share DNA to start with.


                            Matt.

                            Comment


                            • #15
                              Matt,
                              The genesis for this thread was somebody I know asking me how much DNA double first cousins shared. Initially I just quoted the number I'd seen mentioned which was 25% but the more I thought about it the more I started to think that there was something wrong with that number.

                              Double first cousins are arguably closer than half siblings since they likely share some amount of DNA that is identical on both halves. As you pointed out, he only other relationship where this is true is full siblings.

                              Comment

                              Working...
                              X