Announcement

Collapse
No announcement yet.

Polish DNA Tribes Test

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

  • Polish DNA Tribes Test

    Has anyone who knows they are Polish taken the DNA Tribes test? Did you get matches with Poland? My relative took the test and got several matches with Poland and we're not sure how accurate this is. It would be interesting to know the results of people who know they are Polish.

  • #2
    I have no known Polish heritage: Tribes showed a tiny reading for me ("Poor" MLI.)

    I got a better, more accurate read from the ENSFI using Tribes data:



    Don't know if you've tried that (Thanks Tom!), but it opens up a whole new realm of data with far larger samples...

    does your relative show any tiny amounts of Asian? (not in the Continent Match, but for Native or Global Match)?

    The reason I ask is that I have heard anecdotally that some Poles show a little trace of (probably) Mongol or other Asian DNA.
    Last edited by NormanGalway; 26 August 2006, 09:33 PM.

    Comment


    • #3
      Originally posted by haplogroupc
      Has anyone who knows they are Polish taken the DNA Tribes test? Did you get matches with Poland? My relative took the test and got several matches with Poland and we're not sure how accurate this is. It would be interesting to know the results of people who know they are Polish.
      I have never taken an autosomal DNA test, but when I read of the results of those who have, I can often make a correlation between their autosomal results and their yDNA/mtDNA results (or, for a woman, her mtDNA results and the yDNA results of her father).

      In this case, Poland (along with Ukraine and Belarus) is the R1a nation (over 50%). A man with yDNA haplogroup R1a, or a woman whose father is R1a, is likely to get an autosomal match with Poland. Ironically, such individuals may also get a match with "south Asia," meaning India, because R1a is quite common there too (at least in northern India).

      Comment


      • #4
        Originally posted by NormanGalway
        I have no known Polish heritage: Tribes showed a tiny reading for me ("Poor" MLI.)

        I got a better, more accurate read from the ENSFI using Tribes data:



        Don't know if you've tried that (Thanks Tom!), but it opens up a whole new realm of data with far larger samples...

        does your relative show any tiny amounts of Asian? (not in the Continent Match, but for Native or Global Match)?

        The reason I ask is that I have heard anecdotally that some Poles show a little trace of (probably) Mongol or other Asian DNA.
        NormanGalway,

        My relative didn't get any Asian matches but she got weak matches with nearby countries like Germany, Bosnia-Herzogovina and Romania. Her top three matches on the Native Population list were Polish and they range from "good" to "weak" on the scale. She also got a weak match with Poland on that same list. Her Global Population matches were also Polish at the top of the list. We're not sure what to make of the test. Is she Polish? She doesn’t know anything about her European ancestry.

        I've tried to use the ENSFI database but I'm not sure how to use it. Do I type in all markers or just one at a time? Which numbers equal a closer match than the others? The bigger numbers or the smaller numbers?

        Comment


        • #5
          Originally posted by lgmayka
          I have never taken an autosomal DNA test, but when I read of the results of those who have, I can often make a correlation between their autosomal results and their yDNA/mtDNA results (or, for a woman, her mtDNA results and the yDNA results of her father).

          In this case, Poland (along with Ukraine and Belarus) is the R1a nation (over 50%). A man with yDNA haplogroup R1a, or a woman whose father is R1a, is likely to get an autosomal match with Poland. Ironically, such individuals may also get a match with "south Asia," meaning India, because R1a is quite common there too (at least in northern India).
          Igmayka,

          Are you planning to take the test? It would be interesting to see what your results would be. My relative's situation is a little tricky because part of her ancestors are Native American and part are from Europe. So her mtDNA is Native American but she has no idea what the Y-DNA is of her male relatives. She and I are closely related and I got a weak match with Belarus myself and also matches with India.

          Comment


          • #6
            haplogroup c,

            I am hardly the expert, and have my own issues with autosomal testing that I've just started a thread about. My issues with statistics are of an altogether different nature.

            You should start by entering all of your data. Two columns will be blank because Tribes does not test for those. Then, if you want to do further investigation, you can pull them apart one by one if you want to see how often each individual allele appears in different populations. I learned that trick from Tomcat.

            My understanding (thanks Tom!) is that with Ensfi, you can can look at the population numbers to get an idea of how many samples they are out of. In the left hand column, the higher the number the closer the match. In the Baldwin size correction column (I think it's the third across), the larger the number of zeroes following the E (exponent) the lower the frequency of the match. So for example, I may have a "5" next to Polish. Then I look across and find that it is out of 100 samples (this is an utterly phony example because these are out of millions).

            However, I may have a "5" next to another country or region, yet it be out of 10,000 samples. Clearly, I am closer to my polish matches. Throw a stone at a 100 poles and I am more likely to find a match - I'd have to throw the stone at 10,000 of the other area to find a similar number of matches.

            does that make sense? I hope not, because I've managed to confuse myself...

            Comment


            • #7
              My father's line is from the mountainous Poland-Slovakia border. Surnames and history books indicate I am of Polish descent.

              The DNA Tribes test gave me a shockingly strong match with a group in India, followed by six or seven other Indian populations, then:

              Lithuanian
              Belorussian
              German
              Turkish Aegean
              Vienna but not Austria in general
              Naples
              Lodz Poland
              Japanese (?)
              Western Poland
              Norwegian
              Italian

              I had no match with the other Polish populations.

              Results of a previous DNA Tribes test gave me a strong match to Polish Tatars. Don't know where that went on round 2.

              So these results tell me my ancestors on the male side were a mixture of Slavicized Germans from Silesia, and people from where Lithuania and Belorussia meet.

              The Turkish Aegean match is part of my Y-DNA story. My Y-DNA is J1 with no matches to any Semitic samples and with near matches in the Caucasus and Anatolia. My direct male ancestor must have been someone from just north of the Caucasus have tagged along with "Tatars" to the Polish-Lithuanian kingdom.

              Jim

              Comment


              • #8
                Originally posted by lgmayka
                I have never taken an autosomal DNA test, but when I read of the results of those who have, I can often make a correlation between their autosomal results and their yDNA/mtDNA results (or, for a woman, her mtDNA results and the yDNA results of her father)...
                This is sound advice and similar to DNATribes' advice on interpreting results for "admixture" posted on their blog in April of this year and available in their archive. But as more than one parent contributed to your autosome, your results might "cluster" in more than one region.

                The guidance I got on using ENFSI is the following -

                ENFSI only accepts eight of the thirteen markers you got from Tribes but even such a partial profile can be useful. You can get the other two markers needed for the ENFSI standard (D2 and D19 for a total of ten) from DNA-Fingerprint (an FTDNA company).

                ENSFI uses shorthand for their markers and results. Their D8 corresponds to our D8S1179. Their numerical results are in scientific notation - a whole number followed by an E# e.g. 2.356E10. The E stands for exponent, basically a lot of zeros. Without the use of an exponent abbreviation the numerical results would take-up much space and be less conveniently read.

                I usually sort results by their E#'s first, selecting those results with the LOWEST E#(s). And then sort the lowest E# results by the size of the prefacing number. So, 2.356E10 is better than 2.356E12 and 2.578E10 is better than 2.356E10.

                Tom
                Last edited by tomcat; 27 August 2006, 10:32 AM.

                Comment


                • #9
                  This a copy-and-paste from an exchange regarding ENFSI etc on the DNA and Anthropology mail list from earlier this year (I have edited-out names of participants). >> Indicate prior posts to which the writer is responding. The writer is Thomas Krahn of DNA-Fingerprint (now an FTDNA company)..

                  "Just let me add some comments to the discussion:

                  CODIS markers were chosen to have no relation to the health status of
                  a person. The time when CODIS was started, nobody has thought about
                  ethnical relations at all and population studies had just started. The
                  CODIS database was- and still is frequently attacked by data protection
                  activists. So the forensic DNA typing community tries its best to
                  negotiate any connection with phenotypes.

                  > there are a few scientific and forensic articles that do the same and
                  > try to use the CODIS markers as a biogeographical test to infer
                  > phenotypes.

                  Now, that the data are existing, there is no reason not to mine them.

                  > I think we would be better served to look at more ancestry-informative
                  > markers and lower are expectations about confirming recent minority
                  > ancestry.

                  The value of "ancestry-informative" markers is strongly overestimated.
                  There are just a few examples that are really typical for some continents:
                  A very short Penta E allele (2.2 or 3.2) is significant for African
                  ancestry. A very long FGA allele (>30) is also significant for African ancestry. An incomplete THO1 (9.3 or 8.3) is typical for Europeans.

                  However, those signs aren't obligatory. Just take a couple of autosomal
                  markers in combination and you will get the maximum possible information
                  from your autosomal DNA. It does NOT matter if the markers are selected
                  or if they aren't. If there is a clear tendency in one direction you
                  will soon find it out. But if the limit is reached, you can test 100,
                  1000, 10000 or more autosomal markers and you will not find out anything
                  new.

                  > I do think we are just at the beginning of this research and look
                  > forward to an integration of Y NDA, mtDNA, atDNA, X DNA and HLA
                  > testing.

                  The connection of all types of DNA information in the same database is
                  very interesting because there may be possible some "transfer
                  conclusions" one day (if enough data are sampled).
                  Imagine you are a female and find a micro-haplotype block on your XSTR
                  markers. Now you can search all males in the XSTR database with the same haploblock and try to look for clusters in their Y-haplogroups. Because we know quite a lot about Y-haplogroups already, a female can narrow down the most probable region of her ancestral origin this way.

                  > > >>The European STR-base.org website S. shared in a prior
                  > > >>posting on this thread yesterday is thus very interesting. You can
                  > run a
                  > > >>subset of your 13/26 atSTR marker through the European STR-Base at the
                  > > >>as sort of a free "add-on"
                  > to any
                  > > >>CODIS test you have already taken. If you have your CODIS atSTR test
                  > > >>results handy, try the European atSTR database search at to get some
                  > > >>extra utility out of your CODIS test results:
                  > > >>http://www.str-base.org/


                  STR-base actually doesn't use exactly the CODIS markers, but an
                  intersection of the autosomal STR markers used by European police
                  laboratories.
                  D2 and D19 are part of the Identifiler Kit from Applied Biosystems and
                  are frequently used.

                  > > > I went back there and entered
                  > > > my marker values, and perhaps now someone could explain to me a little
                  > > > what the numbers and headings mean. Here are the headings:
                  > > >
                  > > > 1. Actual Matching Probability - I take it this is the raw figure.

                  This figure tells you how likely it is that you will find another person
                  in the selected country(ies) that has exactly the same alleles.
                  If you would get a matching probability of 0.01 it would mean that in
                  100 persons from this country you will find one that matches your
                  profile completely. Normaly theese numbers are very much smaller,
                  because it is very unlikely that you will find any matching person in
                  the whole world at all if you consider a complete profile.

                  > > > 2. Balding Size Bias Correction - Is this an amount to be added as a
                  > > > correction, or is it the corrected figure?

                  See:

                  The theta correction factor can be entered in the field for "FST" in the
                  allele input form.
                  This correction factor has something to do with the difference of a
                  population sample in comparison to the whole population.
                  Some alleles in the population can be enhanced due to selection
                  processes and founder effects. At court the forensic expert wants to be
                  on the safe side if it comes to a prosecution of an offender. So they
                  correct the matching probability to a higher value if they want to
                  exclude a random match.
                  The number in the table is already the corrected result.

                  > > > 3. Balding & Nichols FST - What is an FST??

                  "coancestry coefficient", theta or Fst, which is the correlation
                  between two alleles sampled from distinct individuals
                  within a subpopulation. You enter this correction factor manually into
                  the input box.
                  Usually 0.01 is a good choice. The ENFSI database authors say that they
                  have measured a real value of ~0.001 by comparing different
                  subpopulations, but to be on the safe side they use a default of 0.01.
                  See:


                  > > > 4. NRC II Upper Bound Confidence Intervals - ??

                  Just to take account of possible sampling and testing errors.

                  > > > So which is the most definitive for me to look to as a result?

                  If you only compare populations against each other you needn't consider
                  the correction factors at all, because the theta / Fst is the same for
                  all populations.
                  However, you get an idea of validity of your biogeographical conclusions
                  if you consider the confidence intervalls in relation to the absolute
                  matching probability values.

                  > > > Then, there are the numbers. Could someone possible explain to me
                  > which
                  > > > of these is a higher probability -- that is, could you tell me
                  > whether I
                  > > > have these (below) in the correct order, from higher to lower
                  > > > probability, or are they backwards, or scrambled? It was explained to
                  > > > me very quickly once, but I obviously didn't understand it well enough
                  > > > to retain the knowledge.
                  > > >
                  > > > 1.0008E-15 North Ireland
                  > > > 9.0891E-15 Belgium
                  > > > 3.1851E-16 Austria
                  > > > 8.0932E-16 Estonia

                  From this short list it seems to be most likely to find a profile
                  similar to yours in Belgium."
                  Last edited by tomcat; 27 August 2006, 11:17 AM.

                  Comment


                  • #10
                    Originally posted by haplogroupc
                    My relative didn't get any Asian matches but she got weak matches with nearby countries like Germany, Bosnia-Herzogovina and Romania. Her top three matches on the Native Population list were Polish and they range from "good" to "weak" on the scale. She also got a weak match with Poland on that same list. Her Global Population matches were also Polish at the top of the list. We're not sure what to make of the test. Is she Polish? She doesn’t know anything about her European ancestry.
                    Since my yDNA haplogroup is I1b, I would also expect to get at least weak matches with Bosnia and Romania, hotbeds of I1b. (Modern Poland is about 10% I1b, but this percentage may be higher in southern Poland.) But I have to think my strongest matches would be Polish, since essentially all of my ancestors have been Polish for many hundreds of years.

                    However, my mother's brother (and hence also my mother's father) is of yDNA haplogroup G, an apparent remnant of Alanic warriors who settled in Poland (to escape the Huns!) about 1500 years ago. So I suppose I could get some weak match with the Caucasus.

                    But no, I do not intend to spend money on this particular test, sorry.

                    Comment


                    • #11
                      Originally posted by Jim Honeychuck
                      My father's line is from the mountainous Poland-Slovakia border. Surnames and history books indicate I am of Polish descent.

                      The DNA Tribes test gave me a shockingly strong match with a group in India, followed by six or seven other Indian populations, then:

                      Lithuanian
                      Belorussian
                      German
                      Turkish Aegean
                      Vienna but not Austria in general
                      Naples
                      Lodz Poland
                      Japanese (?)
                      Western Poland
                      Norwegian
                      Italian

                      I had no match with the other Polish populations.

                      Results of a previous DNA Tribes test gave me a strong match to Polish Tatars. Don't know where that went on round 2.

                      So these results tell me my ancestors on the male side were a mixture of Slavicized Germans from Silesia, and people from where Lithuania and Belorussia meet.

                      The Turkish Aegean match is part of my Y-DNA story. My Y-DNA is J1 with no matches to any Semitic samples and with near matches in the Caucasus and Anatolia. My direct male ancestor must have been someone from just north of the Caucasus have tagged along with "Tatars" to the Polish-Lithuanian kingdom.

                      Jim
                      Jim Honeychuck,

                      Do you have any idea why you got so many matches with India? I noticed that the Romany (Gypsy) sample had alot of matches with India since that's where they originate from. I also got alot of matches with India on my test. I wonder if this could mean that we have Romany ancestry as well.

                      Comment


                      • #12
                        Originally posted by tomcat
                        This is sound advice and similar to DNATribes' advice on interpreting results for "admixture" posted on their blog in April of this year and available in their archive. But as more than one parent contributed to your autosome, your results might "cluster" in more than one region.

                        The guidance I got on using ENFSI is the following -

                        ENFSI only accepts eight of the thirteen markers you got from Tribes but even such a partial profile can be useful. You can get the other two markers needed for the ENFSI standard (D2 and D19 for a total of ten) from DNA-Fingerprint (an FTDNA company).

                        ENSFI uses shorthand for their markers and results. Their D8 corresponds to our D8S1179. Their numerical results are in scientific notation - a whole number followed by an E# e.g. 2.356E10. The E stands for exponent, basically a lot of zeros. Without the use of an exponent abbreviation the numerical results would take-up much space and be less conveniently read.

                        I usually sort results by their E#'s first, selecting those results with the LOWEST E#(s). And then sort the lowest E# results by the size of the prefacing number. So, 2.356E10 is better than 2.356E12 and 2.578E10 is better than 2.356E10.

                        Tom
                        tomcat,

                        Thank you for all the information. I read the DNA Tribes April blog and from what I understand, the results are more geographic than anything. If we have ancestors from a certain part of the world, we will get matches with that general area. It's not about revealing the specific country our ancestors came from.

                        By the way, what is the XSTR test? Is that a different test and can a female take it?

                        Comment


                        • #13
                          Originally posted by lgmayka
                          Since my yDNA haplogroup is I1b, I would also expect to get at least weak matches with Bosnia and Romania, hotbeds of I1b. (Modern Poland is about 10% I1b, but this percentage may be higher in southern Poland.) But I have to think my strongest matches would be Polish, since essentially all of my ancestors have been Polish for many hundreds of years.

                          However, my mother's brother (and hence also my mother's father) is of yDNA haplogroup G, an apparent remnant of Alanic warriors who settled in Poland (to escape the Huns!) about 1500 years ago. So I suppose I could get some weak match with the Caucasus.

                          But no, I do not intend to spend money on this particular test, sorry.
                          lgmayka,

                          Thanks for the info. I can see how Y-DNA is related to autosomal DNA. Maybe that's why I got those weak matches with those other countries. I never looked at it that way before.

                          By the way, you're fortunate to know your ancestry. Do the names Fabet, Severa and Fhelma sound Polish to you? I mean have you ever heard them used as Polish names? I'm trying to find where they originate from.

                          Comment


                          • #14
                            Originally posted by haplogroupc
                            ... By the way, what is the XSTR test? Is that a different test and can a female take it?
                            DNA-Fingerprint offers two panels of XSTR's and they host a db of results. There is at present only a small db of XSTR results linked to geography/ethnicity but I expect that will grow. DNA-Fingerprint is now owned by FTDNA (see press-release FTDNA homepage). I expect an announcement in the nearer future from FTDNA, but you can email Thomas Krahn at DNA-Fingerprint, now, and discuss the matter. The XSTR results db is supposed to move to FTDNA and I suppose it will have a matching capability.

                            You can take an X-chromosome test. Your two X's are composites of your maternal and paternal grandparents respectively. So X-testing will be a great way to jump a generation to look for information about ancestries. The only issue for women is a means to discern from which line each X descends. Testing a brother is the usual means.

                            But in your case, with samples from both your parents (?) you could go even deeper by using their DNA directly.

                            Tom

                            Comment


                            • #15
                              X DNA confusion ...

                              Tomcat - my limited understanding has been that I would learn nothing more from having my mother's DNA tested than mine (e.g., my having done a MT-DNA plus test). If I understand you, my mother's DNA will (or could, ultimately) reveal additional information because of the two X's.

                              ???

                              Comment

                              Working...
                              X