Announcement

Collapse
No announcement yet.

Advice needed on DNATribes

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

  • #16
    Probably gonna sounds really thick now but why can you put your results into omnipop and the mounties database etc, but then need DNA fingerprint to analyse your DNA for the ENFSI? That database of me, obviously would be good to look at. Isn't it just the same stuff?

    Comment


    • #17
      Burto,

      Every database is somewhat different - ENFSI's db has Euro populations not contained in DNAT's db. One would only use DNA-F if one wanted more markers to use in STR database searches.

      Josh W,

      "One SNP is not as informative as one STR, since either SNP allele will be shared with a large percentage of the population. It would take more than 100 SNP's to equal 13 STR's in power ..." from Trace Your Roots with DNA, Smolenyak and Turner

      ABDNA 2.5 uses 175 SNP's

      There was a recent paper that proposed one could nail the predominant ancestry of a testee with as few as 10 SNP's.

      The problems many have with ABDNA are their methodologies and their measure of minor ancestral components. ABDNA reached conclusions from very small data sets from derived populations - American college students, I heard. And their accounting for minor admixture, especially when the results fall within their own confidence range, is annoying and useless.

      I took the ABDNA and would probably try again if the test was significantly improved. If DNATribes had been available at the time of my purchase decision I believe I would have elected for that test instead. I have done the DNAT Basic and then Upgrade. The Upgrade included an Update of my entire profile via their improved algorithm and expanded database of populations - a good value!

      Generally, I prefer the openess of the STR standard as opposed to the closed/proprietary character of DNAPrint's methodology. I like being able to match to actual populations rather than generic categories. And I prefer being able to work on my own with my markers over passive acceptance of DNAP's pronouncements.

      Tom

      Comment


      • #18
        Tomcat, you raise many important points. One of my pet complaints on this forum has been about proprietary databases. Why can't all the commercial and academic anonymous databases be combined into a single megadatabase for those not seeking immediate family history.
        Since ABDNA does not publish statistics about the discriminating power of its algorithm versus other strategies it is hard to determine which is the most valid approach. The confidence interval issue while perhaps confusing to laypeople is not a limitation. In general, statistical scores including those used byDnaTribes are subject to the same consideration. ABDNA claims that its 3.0 version will provide a discriminating strategy to identify specific national groups, but seeing is believing. On the other hand, some of the responses to the DnaTribes' forum raise serious questions about the value of that particular approach (I don't mean all STR approaches).
        Josh
        Last edited by josh w.; 21 July 2006, 12:28 PM.

        Comment


        • #19
          But once i got my results from DNATribes, would I be able to check them myself against Omnipop, ENFSI etc.?

          Comment


          • #20
            Originally posted by burto
            But once i got my results from DNATribes, would I be able to check them myself against Omnipop, ENFSI etc.?
            YES! On ENFSI you could get results for 8 of your 13 markers and all 13 markers would be accepted by Omnipop and RCMP.

            Tom

            Comment


            • #21
              Originally posted by tomcat
              YES! On ENFSI you could get results for 8 of your 13 markers and all 13 markers would be accepted by Omnipop and RCMP.

              Tom
              Is there a new version of Ancestry by DNA2.5? (not referring to the Euro 3.0) Ethno-Ancestry was supposed to come out with their own version last fall. Also I did not receive the alleles from the 2.5 test. Has anyone been able to get them on request so they can compare to DNA Tribes alleles and other databases.
              Tried Omnipop but for some reason file will not open.

              Comment


              • #22
                As far as I know, they are still using the same test. I checked their site last night and did not see any changes or news. Perhaps they are working on a 3.0 version, maybe not. I also wonder when Ethnoancestry will finally come out with their own version. I guess it is proving to be more difficult than they had anticipated or perhaps they just want to make sure it will be the most accurate test on the market. I have no faith in Ancestrybydna when it comes to estimating minority racial admixture. I've read of a French woman who was told she was "25% Amerindian" when no one among her known ancestors was from the Americas. On the other hand, some people who are full-blooded Americans and who had reason to believe they had some native genes, were told they had none. Plus many other cases which are really puzzling.

                Comment


                • #23
                  Tomcat, perhaps my prayers have been answered. Lgmayka noted on another thread that Ftdna just bought out DnaFingerprint.

                  Comment


                  • #24
                    [QUOTE=josh w.]Tomcat, you raise many important points. One of my pet complaints on this forum has been about proprietary databases. Why can't all the commercial and academic anonymous databases be combined into a single megadatabase for those not seeking immediate family history.[QUOTE=josh w.]

                    As per your more recent post, mergers (FTDNA and DNA-F) is one way company db's get merged. When DNAP bought Trace Genetics they promoted TG's big Native American db as a chief benefit. Other than that I can think of a number of reasons why academic db's would not of-their-own-course consolidate - differing data sets, differing standards, the lack of a repository/host etc.

                    Sorensen (SMGF) aims to have, eventually, a monster db of related x-, y-, mt-, and autosomal-DNA results but only their y-DNA db is currently available. I do think project administrators could, today, welcome other DNA test results from project participants - combining mt- with x-DNA results seems natural but y-DNA projects could also benefit from x data, and both types of projects ought to welcome autosomal STR results.

                    And it would be a GOOD THING if there was, at least, a website with descriptions and links to all public DNA db's where one could search, post results etc.

                    Originally posted by josh w.
                    Since ABDNA does not publish statistics about the discriminating power of its algorithm versus other strategies it is hard to determine which is the most valid approach. The confidence interval issue while perhaps confusing to laypeople is not a limitation. In general, statistical scores including those used byDnaTribes are subject to the same consideration. ABDNA claims that its 3.0 version will provide a discriminating strategy to identify specific national groups, but seeing is believing. On the other hand, some of the responses to the DnaTribes' forum raise serious questions about the value of that particular approach (I don't mean all STR approaches). Josh
                    DNAP has been catching heat since they launched their test and the complaints of today are identical to the complaints of the past - generally, the incomprehensibility of results - results that don't match the paper trail, that don't pass "the mirror test" and minor admixture numbers that can't stand the test of DNAP's own confidence ranges! They also get a lot of informed criticism from the statistically-savvy for their methodology. Only DNAP can say if the long-term volume of complaint approaches statistical significance!

                    But I personally feel that some of that frustration comes from the way DNAP presents themselves. Have you plowed through the dense texts on ABDNA's site? They are not very informative and seem to exist solely as potential legal defense. And when their (apparent) defensiveness combines with their "closedness" as to their methods and their (apparent) distance form the concerns of their customer the gestalt has an authoritarian odor. And all that is capped with the arrival of of a (cheesey) gold-foil-stamped certificate of ancestry. (Here I would insert and eye-rolling Emoticon were one available).

                    Tom

                    Comment

                    Working...
                    X