Announcement

Collapse
No announcement yet.

Do I need 7cm total or a 7cm segment?

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

  • Do I need 7cm total or a 7cm segment?

    I understand that in order for a match on the X to be meaningful, it should be more than 7cm...but I'm unclear on whether that's the total number or a specific segment.

    For example, I have one alleged second-to-fourth cousin with whom I'm especially eager to figure out the match. Our GedMatch X-DNA Comparison shows these segments:

    3.2cm, 286 SNPs
    4.2cm, 307 SNPs
    3.3cm, 235 SNPs
    3.7cm, 299 SNPs
    4.6cm, 338 SNPs
    3.3cm, 204 SNPs

    So we have 22.3cm total...but our largest segment is 4.6cm.

    We are both female. Do we really match on the X, or is this just noise?

  • #2
    Originally posted by RandomGenie View Post
    I understand that in order for a match on the X to be meaningful, it should be more than 7cm...but I'm unclear on whether that's the total number or a specific segment.

    For example, I have one alleged second-to-fourth cousin with whom I'm especially eager to figure out the match. Our GedMatch X-DNA Comparison shows these segments:

    3.2cm, 286 SNPs
    4.2cm, 307 SNPs
    3.3cm, 235 SNPs
    3.7cm, 299 SNPs
    4.6cm, 338 SNPs
    3.3cm, 204 SNPs

    So we have 22.3cm total...but our largest segment is 4.6cm.

    We are both female. Do we really match on the X, or is this just noise?
    It depends on what you call as meaningful as the term differs from people to people. If you ask me are they related, then yes like beyond 14 to 20 generations. If you ask me are they related in genealogical times like 5 to 8 generations, then no.

    I haven't seen any noise or accidental matching segments with anyone at 3cM for autosomal/xdna with either Neanderthals or Denisovans in GedMatch. Noise don't occur at 3cM.
    Last edited by felix; 5th December 2013, 04:14 PM.

    Comment


    • #3
      For me, at least in terms of this project, "meaningful" means I can eliminate some lines as possible connections if the cousin matches me on the X, since that can only happen on certain lines (i.e. those without a father-son connection).

      It sounds like you're saying that's NOT the case here, and that this cousin could match me on any line, not just the ones that would result in a match on the X.

      Comment


      • #4
        Originally posted by RandomGenie View Post
        For me, at least in terms of this project, "meaningful" means I can eliminate some lines as possible connections if the cousin matches me on the X, since that can only happen on certain lines (i.e. those without a father-son connection).

        It sounds like you're saying that's NOT the case here, and that this cousin could match me on any line, not just the ones that would result in a match on the X.
        Yes, because of cousin marriages which might have happened beyond what you are aware of.

        Yes, it is possible that a male can share his father's X-DNA, provided if the father's mother and the male's mother (or the father's wife) are related without a father-son in the path, in near or distant future.

        Comment


        • #5
          Originally posted by felix View Post
          Yes, because of cousin marriages which might have happened beyond what you are aware of.

          Yes, it is possible that a male can share his father's X-DNA, provided if the father's mother and the male's mother (or the father's wife) are related without a father-son in the path, in near or distant future.
          Hi,

          What Felix says here is true. It is most significant if you are from an endogenous population like Ashkenazic Jews, but it comes up in other populations with significant repeated intermarriage between near or distant cousins.

          Comment


          • #6
            Originally posted by RandomGenie View Post
            For me, at least in terms of this project, "meaningful" means I can eliminate some lines as possible connections if the cousin matches me on the X, since that can only happen on certain lines (i.e. those without a father-son connection).

            It sounds like you're saying that's NOT the case here, and that this cousin could match me on any line, not just the ones that would result in a match on the X.
            Hi,

            To use the X like this, I suggest only using longer matching segments that are highly likely to be recently identical by decent.

            Comment


            • #7
              Originally posted by Rebekah Canada View Post
              Hi,

              What Felix says here is true. It is most significant if you are from an endogenous population like Ashkenazic Jews, but it comes up in other populations with significant repeated intermarriage between near or distant cousins.
              Felix and Rekekah:

              These are confusing statements to people new to DNA testing. It is always good in a case like this to ensure a distinction is made that inheritance, and sharing are not always the same in a FF test result. A man can not "inherit" X DNA from his father. Yes a man can "share" X DNA from his father's side but it is not actually inherited, it is shared as a result of the test matching his mother's X to people of close relation in an endogamous situation. The reason a man is a man is due to the fact that a man inherits a Y chromosome, not X, from his father.

              Sharing vs. Inheritance. It is important to distinguish these are different in a statement such as both of the ones just made for people new to genetic genealogy. X DNA matching will include DNA from the opposite lineage due to a female having two X chromosome alleles, but that doesn't mean X is inherited, just matched, in certain situations.

              Matt.
              Last edited by mkdexter; 6th December 2013, 07:50 AM.

              Comment


              • #8
                Originally posted by RandomGenie View Post
                I understand that in order for a match on the X to be meaningful, it should be more than 7cm...but I'm unclear on whether that's the total number or a specific segment.

                For example, I have one alleged second-to-fourth cousin with whom I'm especially eager to figure out the match. Our GedMatch X-DNA Comparison shows these segments:

                3.2cm, 286 SNPs
                4.2cm, 307 SNPs
                3.3cm, 235 SNPs
                3.7cm, 299 SNPs
                4.6cm, 338 SNPs
                3.3cm, 204 SNPs

                So we have 22.3cm total...but our largest segment is 4.6cm.

                We are both female. Do we really match on the X, or is this just noise?
                To provide you a different perspective, I believe the answer is that it is noise. Some people may want to discuss what noise is but that would be a different question and different topic. In your case, to your question - You will not receive a match like that from the FF test once X matching is instituted as it will require a larger matching segment first.

                So, would other tests like 23andMe, or the future FF X match say this is a X chromosome match? The answer is a resounding, "No". Regardless of the fact that it may be a very rare outlier, most test parameters will not consider this a match.

                The fact that you are both female indicates the matching algorithm in gedmatch has to distinguish between two alleles, of which it does not know which came from the paternal side and which came from the maternal side. To work around this, a test like 23andMe requires two female kits match at a 1200 SNPs or more. Gedmatch does not have that requirement and therefore unless it is manually set to 7cM in the one-one comparison, it will pick up small segment matches like this, even if they are noise.

                Matt.
                Last edited by mkdexter; 6th December 2013, 08:01 AM.

                Comment


                • #9
                  This discussion would not be complete without mentioning that matching segments on the X chromosome are significantly different than matching segments on the other 22 chromosomes. That difference becomes even more significant as the 'match' becomes more distant. Put simply, the X chromosome has very unusual inheritance characteristics. As an extreme example, a 10 cM match on an autosomal chromosome might be interpreted as a 5th or 6th cousin. A 10 cM match on the X chromosome could also be a 5th or 6th cousin - or it could easily be twice as distant - or anything in between. Throw in the greater chance of random segments combining across that many generations, and the X chromosome becomes very unreliable for determining the degree of a relationship. ...and as Matt has pointed out, trying to make any use of very small matching segments is likely to be misleading.

                  On the other hand, when you have a reliable match on the X, it can be invaluable for eliminating approximately half of the possible paths of common ancestry.

                  Comment


                  • #10
                    Originally posted by Rebekah Canada View Post
                    Hi,

                    To use the X like this, I suggest only using longer matching segments that are highly likely to be recently identical by decent.
                    One way to determine if an X is recent is if you also have autosomal matches at the same time with the same person. Not all these autosomal segments have to come from the ancestor carrying the X. The non-X segment could even be IBD (identical by descent) from a spouse (mate) of an "X" ancestor.

                    We have found that for adoptees and for those looking for a biological father, an X pattern of inheritance can really narrow the playing field when you have both an X + autosomal match. You can also occasionally hone in on the geographical origin of that segment within a genealogical time period based on a good pedigree.

                    Comment


                    • #11
                      Originally posted by JOlson View Post
                      <snip>
                      On the other hand, when you have a reliable match on the X, it can be invaluable for eliminating approximately half of the possible paths of common ancestry.
                      This is a very important aspect of the X chromosome, so thanks for bringing it up. I want to discuss the math but I am going to do that elsewhere on an advanced X chromosome discussion page so I don't turn off the newbies right off the bat.

                      Comment

                      Working...
                      X