Announcement

Collapse
No announcement yet.

Autosomal Confusion: Omnipop vs. DNA Tribes vs. ENSFI

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

  • Autosomal Confusion: Omnipop vs. DNA Tribes vs. ENSFI

    I embarked on autosomal testing out of simple curiosity. I knew that this field -- at least as it pertains to commercial genetic genealogy -- is in a very early (embryonic?) stage. Therefore, I decided in advance to lower my expectations and take the results with a massive grain of salt.

    However, even an entire Morton's salt container would not have prepared me for what has followed.

    I have used all of the above three tools (one is a company) after getting my data set back from Tribes. But while they agree on some points (I am part Portuguese, apparently. who knew?), they FLATLY contradict one another in more important ways.

    Let me preface this further by saying this is not meant to be a rant: DNA Tribes analyzed my sample with LIGHTNING speed and has been very reponsive to my enquiries. Moreover, they gave me the data I needed to use the other two tools listed above (the Omnipop Excel calculator and the ENSFI database). So far, so good.

    However -- and this is a very big however -- one or more of the above is just plain, old-fashioned wrong. And not by a matter of degree.
    What they don't agree on is where my closest matches are. This of course goes to the root of the whole enterprise, and the utility of paying for autosomal testing.

    I am now left not knowing what to think, or whom to trust. Does anyone who has tried all of the above have similar experiences? I have read the Tribes Kerchner blog and seen that others have been confused. I don't have that problem. I am just getting utterly contradictory info.

  • #2
    Originally posted by NormanGalway
    I embarked on autosomal testing out of simple curiosity. I knew that this field -- at least as it pertains to commercial genetic genealogy -- is in a very early (embryonic?) stage. Therefore, I decided in advance to lower my expectations and take the results with a massive grain of salt.

    However, even an entire Morton's salt container would not have prepared me for what has followed.

    I have used all of the above three tools (one is a company) after getting my data set back from Tribes. But while they agree on some points (I am part Portuguese, apparently. who knew?), they FLATLY contradict one another in more important ways.

    Let me preface this further by saying this is not meant to be a rant: DNA Tribes analyzed my sample with LIGHTNING speed and has been very reponsive to my enquiries. Moreover, they gave me the data I needed to use the other two tools listed above (the Omnipop Excel calculator and the ENSFI database). So far, so good.

    However -- and this is a very big however -- one or more of the above is just plain, old-fashioned wrong. And not by a matter of degree.
    What they don't agree on is where my closest matches are. This of course goes to the root of the whole enterprise, and the utility of paying for autosomal testing.

    I am now left not knowing what to think, or whom to trust. Does anyone who has tried all of the above have similar experiences? I have read the Tribes Kerchner blog and seen that others have been confused. I don't have that problem. I am just getting utterly contradictory info.
    Omnipop is an American forensics tool designed for American forensic analysis. It seems to contain American populations that most likely are of mixed ethnicity due to being in the "melting pot" as the US is sometimes called. ENFSI is a European service and contains populations from Europe. So, my guess is, if you want to know your dominant American ethnicity, Omnipop is your tool, if you want to know your dominant European ethnicity, ENFSI is your tool. It's not that one tool is better than the other, they are just designed to serve different geographical areas.

    Comment


    • #3
      Eki, you are right about ENSFI being all about Europe.

      And it gave me different results than Tribes. Apparently, I am mostly of Nordic origin and Tribes does not really have much coverage there yet.

      However, Tribes is supposed to be able to ascertain the presence of non-European admixture -- at the macro level at least. The Continent Match section of their report is the most robust statistically (they claim).

      This section - in my case at least - gave me a 100% reading with 0% of anything else.

      The RCMP database, being Canadian, also contains American populations in depth. This database basically "agreed" with Tribes.

      Omnipop does not do this - it just spits out matches by population and gives the relevant sample size of each. So you are left to try to interpret or guess the results.

      For example, any one allele may belong to many different populations, from all over the world, even in fairly close frequencies. So it is ambiguous.
      Last edited by NormanGalway; 27 August 2006, 12:26 PM.

      Comment


      • #4
        Anyway, according to ENSFI, I am most closely correlated with the Finns. So perhaps we have something in common.

        Looking at the Omnipop results, I see that they have poor coverage of many European populations, and my "matches" can only be found in populations 10,000X larger (not sure I am expressing this correctly)-- ie, are of much lower resolution than the ones on ENSFI.

        Comment


        • #5
          Originally posted by NormanGalway
          I am now left not knowing what to think, or whom to trust. Does anyone who has tried all of the above have similar experiences? I have read the Tribes Kerchner blog and seen that others have been confused. I don't have that problem. I am just getting utterly contradictory info.
          NormanGalway,

          I looked at the Kerchner log and I noticed that practically everyone has similar matches. We can't all be from the same parts of the world, so our matches are obviously not our actual places of origin. I figure they have a limited supply of DNA samples and that's what we're matched against. We're probably placed in whichever group we come closest to. So I guess it's not possible to find our true places of origin at this stage. Maybe in time.

          Comment


          • #6
            DNATribes results are also puzzling to me and I'm a FAN!

            For instance, I got several Native and Global Matches to Turkish but no score was better than 6.9 (and that was one of my BETTER scores). However when I compare my alleles to the Turkish sample.pdf I find that I share fourteen with that sample including two homozygotic markers and one heterozygotic marker and that we share alleles on eleven of thirteen markers.

            Curiously I got a score of 4.4 for Northern Portugal but my comparison to the Portugal sample.pdf turns-up sixteen matches including two homozygotic and two heterozygotic markers and shared alleles on twelve of thirteen markers. To my simple mind that looks better than the Turkish match.

            On the other hand, I do think the sample.pdf's are mostly "place-holders" as there is no single profile that could contain all the diversity in Turkey or Portugal.

            I'm sticking with ENFSI until the Tribes db SWELLS. Although ENFSI rates me as Swiss, Spanish, Dutch, Danish, French or Swedish, in that order. And they're sticking by their story - I changed all those alleles I share with the Ojibwe sample.pdf to homozygotic entries of the remaining allele and got Swiss, Spanish, Italian, French and Portuguese, in that order.

            Interestingly my Tribes report did show a cluster of Italian matches, a line actually, that "pointed" in both it's geography and strength gradient to the Italian Swiss border. And the ENFSI Scandinavian matches may fit my mysterious maternal grandfather surnamed Nelson and the French match might confirm a hitherto unsubstantiated bit of family lore that my maternal grandmother was French-Indian. I have no idea how the Spanish and Dutch figure into the story as my father is Eastern European.

            Tom
            Last edited by tomcat; 28 August 2006, 01:11 AM.

            Comment


            • #7
              interesting...

              I ordered the extra two markers from DNA Fingerprint to round out the results and lo and behold Omnipop now agrees with ENSFI. However, they still "disagree" with Tribes because Tribes lacks Scandinavian data.

              Lesson Learned: take everything with a grain of salt until you get the complete panel.

              You get the Tribes results and add D2 and D19 from DNA-F in order to do this.

              Comment


              • #8
                Hi,
                NormanGalway, I have sent you a PM as I need you're opinion on my Omnipop scores....also could anyone give me instructions on how to use ENFSI....I can't find the calculator and it says things like "not able to enable" along the top of the page and weird stuff in general

                Comment


                • #9
                  ENFSI seems to be in the middle of a site renovation but you can still do work there - choose Calculate. You will have to scroll down through site renovation notes to get to Calculator and Results.

                  Comment


                  • #10
                    Has DnaTribes been informed of the problem of not considering all of the standard markers of forensic databases. Why should there have to be individual adjustments if they could do it themselves.

                    Comment


                    • #11
                      I told them those extra two markers changed my results significantly, so they must be aware of it now...

                      They changed the order of my matches, among other things (like changing my Omnipop results).

                      Comment


                      • #12
                        Originally posted by NormanGalway
                        I told them those extra two markers changed my results significantly, so they must be aware of it now...

                        They changed the order of my matches, among other things (like changing my Omnipop results).
                        I can't figure out where to find the Omnipop database. The only thing I found was an Excel document. Could you please let me know where it's at? Thanks.

                        Comment


                        • #13
                          DNA markers.

                          I would like to order the other 2 markers D2 and D19 but whats the purpose. Surely its not just so you can get a better match in databases is it. I know DNA Fingerprint does this but do I need to order it from Germany since that it where my actual samples are from DNA Tribes. I do know that FTDNA bought Fingerprint out. How much does it cost for the extra 2 and what advantage is there of getting the extra 2. What is the purpose? Also what about the other panel 2 that Fingerprint offers? Plus what about ordering x testing. Have a nice day! Maria

                          Comment


                          • #14
                            Originally posted by haplogroupc
                            I can't figure out where to find the Omnipop database. The only thing I found was an Excel document. Could you please let me know where it's at? Thanks.
                            Hi,
                            That's what you use...delete the data in the marker table and add your own then press the find frequency button or whatever it's called just underneath.

                            Comment


                            • #15
                              It's about $17. the purpose is to get more accurate information. what other purpose could there be?

                              If you read my comments above, you will see that they made quite a difference in my case.

                              Comment

                              Working...
                              X