Announcement

Collapse
No announcement yet.

American Indian admixture in White Americans

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

  • Originally posted by thetick View Post
    By confound I think nathan means the PF tool will match your genetic segments to the closest population. You get 50 % of your autosomal dna from each parent and there's no guarantee you get equal weights of their ethnicities in your genes and consequently PF.

    Take this example. This is purely for illustration, but could technically happen. In reality all degrees are possible, but you are much more likely to get ethnicities spread evenly.

    Suppose a 50% German 50% / Chinese mother and a 50% African / 50% Pakistani father have two children.

    One child could theoretically get 50% German and 50% African while second child could get 50% Chinese / 50% Pakistani. The children's PF would literally be on the other side of the world.
    Thanks Tick, I understand that better now.

    Comment


    • Originally posted by thetick View Post
      By confound I think nathan means the PF tool will match your genetic segments to the closest population. You get 50 % of your autosomal dna from each parent and there's no guarantee you get equal weights of their ethnicities in your genes and consequently PF.

      Take this example. This is purely for illustration, but could technically happen. In reality all degrees are possible, but you are much more likely to get ethnicities spread evenly.

      Suppose a 50% German 50% / Chinese mother and a 50% African / 50% Pakistani father have two children.

      One child could theoretically get 50% German and 50% African while second child could get 50% Chinese / 50% Pakistani. The children's PF would literally be on the other side of the world.
      Yes, that's essentially what I meant, from a macro-perspective. I'll try to explain it another way at more of a micro-level. Biogeographical ancestry (BGA) analysis uses SNPs known as ancestry informative markers (AIMs). They're chosen specifically because they're found in different percentages between diverse populations.

      For example, imagine a hypothetical SNP known to have the allele CC in 95% of Africans, AA in 95% of Europeans, and either AA, CC, or AC in more or less equal quantities in Asians and Native Americans. You can see actual statistics by SNP and population at the ALlele FREquency Database (ALFRED). Now suppose a person has one African parent (CC) and one European (AA). Obviously, their allele at this SNP has to be AC, which by itself would lead us to believe (incorrectly) the person was Asian (or Native American).

      Then multiply this comparison several thousand times, done by a computer that can only take the data at face value. FTDNA uses around a quarter million SNPs in their PF algorithm. There are several articles explaining new methodologies to whittle the number of AIMs used down to a hundred or so, because BGA can be extremely computation intensive. But those studies use a limited set of samples from known populations, so they're more proofs of concept than anything.
      Last edited by nathanm; 12th April 2012, 08:25 PM.

      Comment


      • Originally posted by nathanm View Post
        Yes, that's essentially what I meant, from a macro-perspective. I'll try to explain it another way at more of a micro-level. Biogeographical ancestry (BGA) analysis uses SNPs known as ancestry informative markers (AIMs). They're chosen specifically because they're found in different percentages between diverse populations.

        For example, imagine a hypothetical SNP known to have the allele CC in 95% of Africans, AA in 95% of Europeans, and either AA, CC, or AC in more or less equal quantities in Asians and Native Americans. You can see actual statistics by SNP and population at the ALlele FREquency Database (ALFRED). Now suppose a person has one African parent (CC) and one European (AA). Obviously, their allele at this SNP has to be AC, which by itself would lead us to believe (incorrectly) the person was Asian (or Native American).

        Then multiply this comparison several thousand times, done by a computer that can only take the data at face value. FTDNA uses around a quarter million SNPs in their PF algorithm. There are several articles explaining new methodologies to whittle the number of AIMs used down to a hundred or so, because BGA can be extremely computation intensive. But those studies use a limited set of samples from known populations, so they're more proofs of concept than anything.
        nathanm thank you so much for the micro-level explanation! You made it very easy to understand, awesome!!

        Comment


        • Well, now that I finally ordered the FF test, maybe I'll see whether or not I have any significant non-Caucasian elements in my autosomal DNA.

          Comment


          • Originally posted by PDHOTLEN View Post
            Well, now that I finally ordered the FF test, maybe I'll see whether or not I have any significant non-Caucasian elements in my autosomal DNA.
            If by some chance I were to find a bit of Native American DNA in my FF results, where would it have come from? Outside of an unknown "non-paternal event" along the way, it could only come from pre-Revolutionary War North Carolina. And not to forget that HVR1 U5 match on Mitosearch that supposedly traces back to a Tuscorora maternal ancestress of theirs. Hmm... Lost Colony, perhaps?

            Comment


            • Originally posted by PDHOTLEN View Post
              If by some chance I were to find a bit of Native American DNA in my FF results, where would it have come from? Outside of an unknown "non-paternal event" along the way, it could only come from pre-Revolutionary War North Carolina. And not to forget that HVR1 U5 match on Mitosearch that supposedly traces back to a Tuscorora maternal ancestress of theirs. Hmm... Lost Colony, perhaps?
              Speaking of the Lost Colony of Roanoke (NC), I was just looking at the Lost Colony Project (mitochondrial). And there is a sample (#195018) who is very close to mine. Although that person/line is apparently not in the U5 Project, I think it should be. It looks to maybe belong in the group (right next to mine) called U5b2b1* Group A2. The HVR1 is 16270, 16292 & 16362. The HVR2 is mine minus my 228A. Hmm...

              Comment


              • Originally posted by rainbow View Post
                I had the autosomal/admixture test thru www.ancestrybydna.com. There are no Indian names in my family tree, but it is incomplete. My result is 83% European, 17% Native American. My ancestors immigrated to America in the 1600s, 1700s,1800,& 1900s, and stayed on the east coast. I must be descended from some of the original northeast tribes.



                Of my 176 markers, 7 are unknown.
                I had this test done in 2006 - do you happen to know what markers are tested and their ds numbers by any chance? I would love the list. When they sent them to me they were in a form like #982 which does not correspond to anything I know of in order to research them.

                Comment

                Working...
                X