Announcement

Collapse
No announcement yet.

Inaccurate X-Chromosome Results?

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

  • Inaccurate X-Chromosome Results?

    I compared my Mom's X-chromosome versus two of my brothers' X results
    and got the following results:

    1) Brother One - shared 185.60 cM on the X-chromosome with his mother.
    2) Brother Two - shared 194.58 cM on the X-chromosome with his mother.

    Is there really a 10 cM difference between the old chip and the new one for the X-chromosome?

    Or, is there a ton of miscalls that are occurring here?

    Thanks!
    Attached Files

  • #2
    It has to do with newer chips recording opposite side of DNA strand. Normally it does not effect the matching algorithm, but on when there is a bunch of values close together on a chromosome the algorithm rejects it.

    One chip records some values in forward orientation, another in reverse orientation. A for T and C for G.

    I have mentioned this a few times, but have never gotten a clear answer if it is an issue they are going to resolve.

    I have noticed these missed sections on a few chromosomes, one I posted about which has 43 SNPs in a row on chromosome 13
    http://forums.familytreedna.com/showthread.php?t=35387

    In regards to X the section that misses has the following Build 36 positions with the forward/reverse orientation values. If you compare your raw data kits, you will see this. One will be G, the other C, and one will be T, the other A

    122664306
    125051065
    126309782
    126687359
    129374997
    132407803
    133547050
    133552958
    134304672
    136473956
    140983624
    148575315
    148575735
    148577057
    148577205
    148582655
    148588026
    148588379
    148593002
    148595188
    148597358
    148604219
    153386506

    Edit
    Attachment is my 3 sister to my Father. My Father tested on chip used between May 2012 and April 2013. Sisters test on chips used after.
    Attached Files
    Last edited by prairielad; 10 July 2014, 09:47 AM.

    Comment


    • #3
      Originally posted by prairielad View Post
      It has to do with newer chips recording opposite side of DNA strand. Normally it does not effect the matching algorithm, but on when there is a bunch of values close together on a chromosome the algorithm rejects it.

      One chip records some values in forward orientation, another in reverse orientation. A for T and C for G.

      I have mentioned this a few times, but have never gotten a clear answer if it is an issue they are going to resolve.

      I have noticed these missed sections on a few chromosomes, one I posted about which has 43 SNPs in a row on chromosome 13
      http://forums.familytreedna.com/showthread.php?t=35387

      In regards to X the section that misses has the following Build 36 positions with the forward/reverse orientation values. If you compare your raw data kits, you will see this. One will be G, the other C, and one will be T, the other A

      122664306
      125051065
      126309782
      126687359
      129374997
      132407803
      133547050
      133552958
      134304672
      136473956
      140983624
      148575315
      148575735
      148577057
      148577205
      148582655
      148588026
      148588379
      148593002
      148595188
      148597358
      148604219
      153386506

      Edit
      Attachment is my 3 sister to my Father. My Father tested on chip used between May 2012 and April 2013. Sisters test on chips used after.
      Looks like FTDNA might want to work on this issue. Our educational efforts have convinced people that a father must give all his daughters exact copies of his X chromosome (except at the very ends of the chromosome in the pseudo-autisomal regions). Many people are depending on this matching algorithm to show evidence for paternity. Mothers are expected to match their sons along the entire length of the X as well. Hopefully those who don't have exact copies will be patient as FTDNA fixes the matching algorithm.

      Comment


      • #4
        I was just questioning this same thing on an autosomal discussion group on facebook. I first tested in 2006 with the National Genographic Project, then ordered a FF test in Dec. 2012 that was processed in the early part of 2013. My mom was tested in mid 2013 and two more sisters between then and June 2014. While both sisters match our mother entirely on the X, I have several small gaps. Is this due to different chips? Comparing with our mother, there are additional small gaps on some chromosomes, 3 of which affect each sister alone, and 3 of which affect me with one of my sisters in the exact same location, two with one sister, one with the other sister. Can anyone explain?
        Last edited by ShellyH; 18 July 2014, 10:19 PM.

        Comment


        • #5
          Originally posted by ShellyH View Post
          I was just questioning this same thing on an autosomal discussion group on facebook. I first tested in 2006 with the National Genographic Project, then ordered a FF test in Dec. 2012 that was processed in the early part of 2013. My mom was tested in mid 2013 and two more sisters between then and June 2014. While both sisters match our mother entirely on the X, I have several small gaps. Is this due to different chips? Comparing with our mother, there are additional small gaps on some chromosomes, 3 of which affect each sister alone, and 3 of which affect me with one of my sisters in the exact same location, two with one sister, one with the other sister. Can anyone explain?
          First if you are talking about a comparison of parent-to-child only there will be gaps for a variety of reasons. One reason is that there are some SNP poor regions where there is not enough information but that usually causes a gray zone.

          Another reason would be that actual gaps are occurring because the chips are different as Prairielad has explained. See the image below where the double helix is visualized. A and T always pair. C and G always pair. A SNP change (polymorphism) will be different so don't confuse a mutation with what I am saying. If a section is read on the wrong strand in the same place, the sequence will be different. Each chromosome has a double helix. You want to always read the green strand or always read the red strand but don't use different strands in different software versions and expect the comparisons to be equivalent. These won't be the same.
          http://en.wikipedia.org/wiki/Single-...le:Dna-SNP.svg

          I have seen some corrections made by FTDNA between parent and child lately but I don't know if this is being done for everyone. Best to ask the Help Desk. I am still seeing tiny gaps where there should be solid matching. Mutations can cause gaps but this should be rare. As I understand it, a one point mutation usually doesn't show up as a mismatch if all the surrounding SNPs are the same. Some companies may smooth out the results more than others.

          Now if you are comparing sibs with each other, then don't expect exact full matching with some exceptions.

          Full sisters are expected to match the father entirely on the X and they will match each other (with the exception of a few insignificant SNPs) on the entire X coming from their father. Right now I would not worry about a few small gaps.

          Full matching doesn't usually occur when sisters look at the X segments they got from their mother. In a sister-to-sister comparison, there should also be some half-matching regions where they match the father but not the mother. You won't see the full matching segment cutoff points at FTDNA but if you upload to GEDmatch or David Pike's tool you will see fully identical regions. There you will be able to see the crossover points (start and end points) for the segments coming from the mother on the X. Half matching versus full matching is hidden in the FTDNA browser.

          Brothers are not expected to match along any entire chromosome with a sibling but it can happen. They usually match their parents entirely on the autosomes with the exception of the gray zones. The X chromosome is where we probably see the most holes or gaps in them because males only have one X and so a mismatch with their mom is more obvious and creates confusion. Again, if you are seeing a few tiny gaps, I would not worry about it but I think FTDNA will have to correct some of these holes.

          Comment


          • #6
            Thanks for all your responses.

            I think I read somewhere that the raw data is now Build 37 but the chromosome browser display is still based on Build 36.

            Still, a 10 cM gap is pretty big.

            Comment


            • #7
              Thank you for the explanation, Kathy. Much of it I understood, but I'll read through it again to be sure I am understanding it fully.

              In this case to which I referred, my mother was the main kit and I was running the comparison of her three daughters to her, all at once in the chromosome browser to that I could see it all in one place. I wasn't comparing the sisters to one another in this instance except for how we compare to our mother.

              My X looks worse until the cM is lowered to 1+, but even then there is a gap. Comparing myself to my mother, my X is in 6 segments: 77.45, a decent sized gap, then 56.1, 3.85, 32.99, and 13.08. My sisters (T & L below) are solid and include data in the segment where I (S below) have the gap.

              Chromosome 2 - S & T have two (same) segments, L is solid.

              Chromosome 5 - T & L have two (different) segments, S is solid.

              Chromosome 7 - S has two segments, T & L are solid.

              Chromosome 8 - S & T have two (same) segments, L is solid.

              Chromosome 11 - S & L have two (same) segments, T is solid.

              So, each of us has gaps to our mom, but I (S) just got more of them.

              Mom was in batch 529
              S was in batch 497
              T was in batch 538
              L was in batch 568

              Do you think this suggests a need to check with the Help Desk, or is this actually pretty common?
              Last edited by ShellyH; 19 July 2014, 04:19 PM.

              Comment


              • #8
                Originally posted by Kathy Johnston View Post

                ..............



                I have seen some corrections made by FTDNA between parent and child lately but I don't know if this is being done for everyone. Best to ask the Help Desk. I am still seeing tiny gaps where there should be solid matching. Mutations can cause gaps but this should be rare. As I understand it, a one point mutation usually doesn't show up as a mismatch if all the surrounding SNPs are the same. Some companies may smooth out the results more than others.

                .........
                I concur that most gaps between parent/child have seemed to be corrected on most of the chromosomes which was due to the way different chip versions recorded the reverse/forward orientation values.

                If I compare my kits to father now, it only shows the larger gaps on the X chromosome.
                EDIT And the end of chromosome 13

                Originally posted by ShellyH View Post
                Thank you for the explanation, Kathy. Much of it I understood, but I'll read through it again to be sure I am understanding it fully.

                In this case to which I referred, my mother was the main kit and I was running the comparison of her three daughters to her, all at once in the chromosome browser to that I could see it all in one place. I wasn't comparing the sisters to one another in this instance except for how we compare to our mother.

                My X looks worse until the cM is lowered to 1+, but even then there is a gap. Comparing myself to my mother, my X is in 6 segments: 77.45, a decent sized gap, then 56.1, 3.85, 32.99, and 13.08. My sisters (T & L below) are solid and include data in the segment where I (S below) have the gap.

                Chromosome 2 - S & T have two (same) segments, L is solid.

                Chromosome 5 - T & L have two (different) segments, S is solid.

                Chromosome 7 - S has two segments, T & L are solid.

                Chromosome 8 - S & T have two (same) segments, L is solid.

                Chromosome 11 - S & L have two (same) segments, T is solid.

                So, each of us has gaps to our mom, but I (S) just got more of them.

                Mom was in batch 529
                S was in batch 497
                T was in batch 538
                L was in batch 568

                Do you think this suggests a need to check with the Help Desk, or is this actually pretty common?

                The other thing to note is the number of no calls listed in a segment of non matching, this to will cause a gap if there is a higher number in this section (total number of no-calls between both kits being compared in respect to gap in parent/child comparison)
                Last edited by prairielad; 19 July 2014, 06:01 PM.

                Comment


                • #9
                  Originally posted by ShellyH View Post
                  Thank you for the explanation, Kathy. Much of it I understood, but I'll read through it again to be sure I am understanding it fully.

                  In this case to which I referred, my mother was the main kit and I was running the comparison of her three daughters to her, all at once in the chromosome browser to that I could see it all in one place. I wasn't comparing the sisters to one another in this instance except for how we compare to our mother.

                  My X looks worse until the cM is lowered to 1+, but even then there is a gap. Comparing myself to my mother, my X is in 6 segments: 77.45, a decent sized gap, then 56.1, 3.85, 32.99, and 13.08. My sisters (T & L below) are solid and include data in the segment where I (S below) have the gap.

                  Chromosome 2 - S & T have two (same) segments, L is solid.

                  Chromosome 5 - T & L have two (different) segments, S is solid.

                  Chromosome 7 - S has two segments, T & L are solid.

                  Chromosome 8 - S & T have two (same) segments, L is solid.

                  Chromosome 11 - S & L have two (same) segments, T is solid.

                  So, each of us has gaps to our mom, but I (S) just got more of them.

                  Mom was in batch 529
                  S was in batch 497
                  T was in batch 538
                  L was in batch 568

                  Do you think this suggests a need to check with the Help Desk, or is this actually pretty common?
                  I would be interested in hearing what the Help Desk has to say about this since it is a common problem but it should be fixable.

                  Comment


                  • #10
                    Thanks again, everyone. I'll report back when I contact them. Hopefully whatever I find will be helpful for others as well.

                    Comment


                    • #11
                      Originally posted by prairielad View Post
                      It has to do with newer chips recording opposite side of DNA strand. Normally it does not effect the matching algorithm, but on when there is a bunch of values close together on a chromosome the algorithm rejects it.

                      One chip records some values in forward orientation, another in reverse orientation. A for T and C for G.

                      I have mentioned this a few times, but have never gotten a clear answer if it is an issue they are going to resolve.

                      I have noticed these missed sections on a few chromosomes, one I posted about which has 43 SNPs in a row on chromosome 13
                      http://forums.familytreedna.com/showthread.php?t=35387

                      In regards to X the section that misses has the following Build 36 positions with the forward/reverse orientation values. If you compare your raw data kits, you will see this. One will be G, the other C, and one will be T, the other A

                      122664306
                      125051065
                      126309782
                      126687359
                      129374997
                      132407803
                      133547050
                      133552958
                      134304672
                      136473956
                      140983624
                      148575315
                      148575735
                      148577057
                      148577205
                      148582655
                      148588026
                      148588379
                      148593002
                      148595188
                      148597358
                      148604219
                      153386506

                      Edit
                      Attachment is my 3 sister to my Father. My Father tested on chip used between May 2012 and April 2013. Sisters test on chips used after.
                      I was so thankful FTDNA was able to find a vial of my deceased father's DNA and do Family Finder recently. He died in 2005 and we originally were told there was not enough DNA. I can use my brother's vial for Y DNA SNPs and STRs but I sure wanted to save something of an autosomal test from my Dad. I feel very blessed.

                      I see where my father and I don't match in the region above on the X because of the problem you mention, Prairielad. One still has to change all Cs to Gs and vice versa and change all Ts to As and vice versa for those particular markers in order to get a full match along the majority of the X chromosome. The software is still reading opposite strands in the double helix in some areas.

                      Comment

                      Working...
                      X