Announcement

Collapse
No announcement yet.

Genetic Distance - what is "close"?

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

  • Genetic Distance - what is "close"?

    If anyone can help me with layman's terms on genetic distance I would be most appreciative. What maximum distance is close?

    For example, if I run a genetic search on ySearch, use 37 markers, max distance of 4, I get 22 hits, 6 of which are Doty, 3 are a form of Beaty, 2 McConnels and 2 Vances. Clearly I am related to Doty somehow, right? But what I know of the paper trail, there are no Dotys.

    ySearch ID is 52PF4.

    Anybody? Bueller?

  • #2
    "close" is a relative term.

    Human and chimpanzee are "close" relatives after all.

    "Humanoids" exist since say 5 Million years.
    Lets say, "Humans" exist since 1 Million years.
    And "modern humans", well lets round it up to 100.000 years.

    All people in the same haplogroup have at least a common anchestor, say 20.000 years ago (some haplogroups/subclades even less).

    All lighteyed (Blue, grey, green) humans have a common anchestor about 6.000 years ago. (but in any line. This must not be the Y-DNA or mtdna line.) If its true, that the mutation, that causes light eyes, had really only happened in one single human, in all of history.

    Mathematically, there is a 98% chance that you have a common anchestor in the y-DNA line, with someone, wich has a genetic distance of 4 on 37 markers, in the last 2.700 years.

    BUT

    This only means, that the lineages of 2 brothers, who lived 1.400 years ago, must be at least 33/37 by now. But you could also match 33/37 with someone by pure random chance.

    Its like with sound or image data.
    there is "wanted signal" and "backgrund noise".

    So more far away the common anchestor is, so weaker becomes the "Wanted signal" and the contrast to the "background noise" vannishes. Finaly, you cannot tell apart the both.

    I think, nobody really knows about these things for sure yet.
    Its like the weather forecast. All this knowledge, all this technology and its still close to beeing useless. Ha ha ha....

    But you may read this:
    http://www.ysearch.org/help.asp?uid=

    Edit: Oh yes, almost forgot: In most European countries, surnames exist since 500 years only.
    Someone must be 36/37 (genetical distance 1) to have a 90% chance of common anchestor in the past 500 years!
    Last edited by Daniel72; 8 February 2010, 10:29 PM.

    Comment


    • #3
      Thanks for the overview. And I like the signal to noise ratio analogy. That is helpful.

      Comment


      • #4
        That sounds like a lot of close matches - my closest match is one person at 30/37, and then 5 more at 28/37 or 29/37.

        You also have some matches at 34/37, one of whom has tested as R1b1b2a1b5, and it looks like you fit in the Leinster cluster - you might want to join this project:

        http://www.familytreedna.com/public/...icwebsite.aspx

        Comment


        • #5
          Originally posted by GST View Post
          That sounds like a lot of close matches - my closest match is one person at 30/37, and then 5 more at 28/37 or 29/37.

          You also have some matches at 34/37, one of whom has tested as R1b1b2a1b5, and it looks like you fit in the Leinster cluster - you might want to join this project:

          http://www.familytreedna.com/public/...icwebsite.aspx
          Interesting stuff! That's all the close matches you have? I have 1 at 25/25, 40 at 24/25, 230 at 23/25, 1 at 35/37 and 17 at 33/37 or 34/37.

          I need to spend more time figuring out what this means and what to do with it.

          Comment


          • #6
            OK I matched up at Ancestry with three others different surname. One was 42 o/o 46 the second one was 43 o/o 46 and the third one was 44 o/o 46. Ancestry say there are within 14, 10, and 6 generations. Now I reading 33 o/o 37 is 2,700 years and 36 o/o 37 is 500 years. LOL!!! BLAH!!!!

            Comment


            • #7
              Seems to me that it's not unusual, with some sweat and elbow grease, to get your paper trail back a couple hundred years. So if you start getting genetic distances of 1 or 2 then you're honing in on things.

              Comment


              • #8
                Originally posted by EdwardRHill View Post
                OK I matched up at Ancestry with three others different surname. One was 42 o/o 46 the second one was 43 o/o 46 and the third one was 44 o/o 46. Ancestry say there are within 14, 10, and 6 generations. Now I reading 33 o/o 37 is 2,700 years and 36 o/o 37 is 500 years. LOL!!! BLAH!!!!
                Ok, lets take an example...

                your 44/46

                They tell you its 6 Generations?

                Well, if one just uses an average mutation rate per marker one gets in that case:

                1% chance, that its your own father.
                10% chance for a common anchestor in the last 4 generations.
                30% chance for last 6 generations
                50% chance for the last 9 generations
                75% for the last 13 generations
                90% chance for the last 18 Generations
                98% chance for the last 25 generations

                Its about chances
                Also, as I mentioned above, these chances are not even really usefull.
                Because, its not really a 1% chance that its your father. But its rather a 1% chance to match only 44/46 with your father.

                Same goes for the other things.
                There is a 98% chance that you match 44/46 with someone who is related to you in the last 25 Generations. It does not really mean, that you are related to 98% of the 44/46 matches out there. And thats what many forget. There will be 44/46 matches that are simply "noise".


                Markers mutate randomly. Thats why one cannot tell for sure.
                BUt its known roughly how often this happens.

                Also, one does not know the lengh of a generation.
                I used 30 years in the example few posts earlier. Plus the 98% chance. to make sure.

                Also I calculated it wrong and correctedb only 1 of the numbers (used them twice). It was 1300 years, not 2700.

                And here is a tool to calculate it:
                http://dna-project.clan-donald-usa.org/tmrca.htm

                Recall that there are 2 trigger events in every generation if you use this.

                Thats because:
                Say you compare yourself to your brother. Then the Y-DNA sequence of your father had 2 chances to mutate. With you and with your brother.
                Last edited by Daniel72; 9 February 2010, 06:42 PM.

                Comment


                • #9
                  Originally posted by GST View Post
                  That sounds like a lot of close matches - my closest match is one person at 30/37, and then 5 more at 28/37 or 29/37.

                  You also have some matches at 34/37, one of whom has tested as R1b1b2a1b5, and it looks like you fit in the Leinster cluster - you might want to join this project:

                  http://www.familytreedna.com/public/...icwebsite.aspx
                  I just got my DYS464X results back and I do indeed fit into the 2c2g Leinster cluster - good call on your part!

                  Comment

                  Working...
                  X