Announcement

Collapse
No announcement yet.

21/29 & Surname match

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

  • 21/29 & Surname match

    I finally found someone whith my Surname, and who matches a reasonable number of markers. He matches 21 of 29 markers. His results are from a different lab. The surname and the STR matches combined suggest the probablity that we are related. The seven marker mismatch seems rather high. I have a couple of questions regardng this. First, is a mismatch by two indicative to two distinct mutations? That is if my DYS385b is 14, and the other person's is 12, dose that mean there were two different mutations probably representing two different generations?

    Secondly, am I correct in assuming that mismatches on 7 markers probably indicates far more than 7 generations? The surname is on the order of 1300 years old, so there is plenty of time for more than 7 generations.

  • #2
    Originally posted by Hetware
    First, is a mismatch by two indicative to two distinct mutations? That is if my DYS385b is 14, and the other person's is 12, dose that mean there were two different mutations probably representing two different generations?


    Secondly, am I correct in assuming that mismatches on 7 markers probably indicates far more than 7 generations? The surname is on the order of 1300 years old, so there is plenty of time for more than 7 generations.

    I think the jury is still out as to whether the ste-wise (I step permutation) or the infinite allele (passibly many steps per mutation) model is appropriate.


    You are proper (IMHO) in your phrasing of *probably* more than 7 generations for 7 mutations. Based on the general assupmption of of a mutation rate of .004 ... about 4 mutations per 1000 'transmission events'.
    For example, if you are looking at a 25 marker comparison then 4 generations are required for 100 transmission events .. for a common ancestor 4 generations ago .. there would be 200 transmission events and based on our origianal assumption .. we might expect 1 mutation.
    You can see it would take a while to add up to 7 mutations but it is statistically *possible* .. just not very likely.

    We can muddy the waters a bit, however. It does seem that certain populations are more prone to producing mutations than others. I've been led to believe I'm on one of those populations (haplogroup N3a ) and I'm about 5 mutations from a family that I'm afairly certain is related..we are looking for the link .. we are 31/37 on our Y-DNA .. they (4 cousins) are very close to one another .. only 1 marker differnce among them and it is the CDYa marker which is considered to be fast moving or even "flakey' but seems to be useful for helping to separate close family members.

    I hope this helped you .. I'm still learning myself after 4 years of being tested and studying and reading.
    Good luck to you,
    ChrisS

    Comment


    • #3
      21/29 is a mismatch. The chance that you are related within the period that surnames have been around (even if it is as high as 1300 years) is extremely slim.

      Comment


      • #4
        Originally posted by bob_allison
        21/29 is a mismatch. The chance that you are related within the period that surnames have been around (even if it is as high as 1300 years) is extremely slim.
        If I take the estimate of .004 errors per transmission event, assume 20 years per generation and 13 centuries, I get 5 * 13 * 29 * .004, I get 7.54 mutations. That would be more than are present in the data. If I take 25 years per generation, I get 6.03 which is one less than the data shows. There is also a person with a variant spelling who seems to have a pattern that correlates with mine and the other person with my name.

        There is also the possibility that the surname goes back even further if we consider that it is likely a tribal name dating back to the time of Caesar. This is both fascinating and frustrating. I am trying to piece together some ancient legends and speculative connections that go back to the time of Caesar. So far I have found nothing to disprove my hypothesis, nor have I found much that solidly confirms any part - other than the broad picture that I am probably of European male ancestry going back to before the current era.

        Comment


        • #5
          Originally posted by Hetware
          If I take the estimate of .004 errors per transmission event, assume 20 years per generation and 13 centuries, I get 5 * 13 * 29 * .004, I get 7.54 mutations. That would be more than are present in the data. If I take 25 years per generation, I get 6.03 which is one less than the data shows. There is also a person with a variant spelling who seems to have a pattern that correlates with mine and the other person with my name.

          There is also the possibility that the surname goes back even further if we consider that it is likely a tribal name dating back to the time of Caesar. This is both fascinating and frustrating. I am trying to piece together some ancient legends and speculative connections that go back to the time of Caesar. So far I have found nothing to disprove my hypothesis, nor have I found much that solidly confirms any part - other than the broad picture that I am probably of European male ancestry going back to before the current era.
          Hetware,

          If you're able to prove or at least get reasonable indications that you have a commmon ancestor 1300 years back that would be exceptional.

          Even though by FTDNA's own criteria anything above a genetic distance of 4 is too far off to be considered related, we know that there are always exceptions.
          http://www.familytreedna.com/GDRules_25.html
          http://www.familytreedna.com/GDRules_37.html

          For example, in my surname project the two participants most separated have a 4 step distance (21/25) and we're dealing with a genealogical timeframe of about 400 years. So a 7 step distance in a 1300 year period doesn't seem that farfetched, provided that there are other reasons to believe in a shared ancestry.

          And check the quote below that I snatched from the DNA Genealogy List:
          In Gusmao's recent paper, "Mutation Rates a Y Chromosome Specific
          Microsatellites," there is a table combining data from all father-son studies. Based on the 137 mutations found in 64,273 "allele transfers" (number of markers tested x number of father/son pairs), the distribution was as follows:

          85 one-step increases
          46 one-step decreases
          3 two-step increases
          1 two-step decrease
          1 three-step increase
          1 four-step decrease
          As you can see most mutations seem to be of a single step but there are a few exceptions.

          Victor

          Comment


          • #6
            I would have to agree with Bob on that one.

            You have not factored in the probability of a break in the genetic line, which over the course of 1300 years is very likely. Admittedly there are no good estimates at how often this happens, I have read 1% per generation as one estimate. Take this over 65 generations and it equates to a risk that can't be ignored.

            One result is far too little to go on. I hope you will find closer matches in the future and you may find that these take you in a different direction in terms of marker values.

            Good Luck!

            Comment

            Working...
            X