Announcement

Collapse
No announcement yet.

Multiple Results-No idea how to interpret now.

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

  • Multiple Results-No idea how to interpret now.

    I initially tested with Ancestry.com and received results that weren't overly surprising. But now I've uploaded my raw DNA to MyHeritage and FTDNA, as well as run them through pretty much every test on GEDMatch and now I'm not sure what to consider an outlier or overlap. Could someone that's well-versed in this help me figure out what my ancestry might actually look like?

    I do understand this is no where near an exact science.
    My known ancestry it Italian on my father's side but have only been able to trace back to his 4 grandparents being in Italy. I have all of his g-grandparents names and they are Italian, but nothing definitive that they were from Italy.

    On my mother's side my tree goes back to roughly 1300's on her father's side to French Canada and then back to England and France. Her mother's side traces back to England and Ireland mid 1800s and drops off there on one side, and England and France back to early 1800s before the line disappears on the other.

    My various test results are below. Thank you so much for any help! I'm still getting my feet wet on all of this.

    Ancestry:
    Europe South: 47%
    Europe West: 29%
    Caucasus: 12%
    Ireland/Scotland/Wales: 5%
    Great Britain: 2%
    Scandinavia: 2%
    Iberian: 2%
    Middle Eastern: 1%

    MyHeritage:
    Greek: 37.6%
    Iberian: 27.6%
    North and West Europe: 23.8%
    West Asian: 11%

    FTDNA:
    British Isles: 27%
    Iberia: 17%
    Southeastern Europe: 17%
    Eastern Europe: 9%
    Asia Minor: 28%
    Ashkenazi: <1%
    North and Central American: <1%
    South American: <1%

    GED Match K13:
    North Atlantic: 28.4%
    Baltic: 13.17%
    W Mediterranean: 21.20%
    West Asian: 13.06%
    Eastern Mediterranean: 18.71%
    Red Sea: 2.94%
    Amerindian: 2.01%

    K13 Oracle4:
    Using 1 population approximation:
    1 North_Italian @ 8.718374
    2 Tuscan @ 8.750091
    3 Greek_Thessaly @ 12.118048
    4 Italian_Abruzzo @ 12.131498
    5 Romanian @ 12.995146
    6 Bulgarian @ 13.430166
    7 West_Sicilian @ 14.165439
    8 Serbian @ 16.096497
    9 Central_Greek @ 16.285183
    10 Portuguese @ 16.301798

    Using 2 populations approximation:
    1 50% Central_Greek +50% French @ 4.048017


    Using 3 populations approximation:
    1 50% North_Italian +25% Turkish +25% West_German @ 3.217606


    Using 4 populations approximation:
    ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++
    1 French_Basque + Georgian_Jewish + Hungarian + Tuscan @ 2.202214
    2 Armenian + Serbian + Southwest_French + Spanish_Murcia @ 2.234617
    3 Armenian + Serbian + Southwest_French + Spanish_Valencia @ 2.265323
    4 Armenian + Serbian + Southwest_French + Spanish_Castilla_Y_Leon @ 2.308479
    5 Armenian + Serbian + Spanish_Cantabria + Spanish_Cataluna @ 2.330874
    6 Armenian + Serbian + Southwest_French + Spanish_Cataluna @ 2.338314
    7 Armenian + Portuguese + Serbian + Southwest_French @ 2.352947
    8 Armenian + Serbian + Spanish_Cantabria + Spanish_Valencia @ 2.366077
    9 Armenian + Romanian + Southwest_French + Spanish_Cataluna @ 2.411644
    10 Armenian + Serbian + Spanish_Cantabria + Spanish_Castilla_Y_Leon @ 2.413403
    11 Armenian + Romanian + Southwest_French + Spanish_Castilla_Y_Leon @ 2.414363
    12 Armenian + Serbian + Southwest_French + Spanish_Extremadura @ 2.432635
    13 Georgian_Jewish + Romanian + Southwest_French + Spanish_Cataluna @ 2.462730



    GED Match K15:
    North Sea: 18.37%
    Atlantic: 18.65%
    Baltic: 9.4%
    East Europe: 3.39%
    West Mediterranean: 16.46%
    West Asian: 11.81%
    East Mediterranean: 17.35%
    Red Sea: 3.02%
    Amerindian: 1.72%

    K15 Oracle4:
    Using 1 population approximation:
    1 Tuscan @ 7.560107
    2 North_Italian @ 9.185580
    3 Greek_Thessaly @ 10.388011
    4 Italian_Abruzzo @ 11.630812
    5 Romanian @ 11.743538
    6 Bulgarian @ 11.829917
    7 Greek @ 12.778710
    8 Serbian @ 13.868921
    9 West_Sicilian @ 13.947838
    10 Central_Greek @ 14.776825

    Using 2 populations approximation:
    1 50% Central_Greek +50% French @ 3.411801


    Using 3 populations approximation:
    1 50% Central_Greek +25% North_Dutch +25% North_Italian @ 2.885310


    Using 4 populations approximation:
    ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ +
    1 Central_Greek + Greek_Thessaly + Southwest_English + Tuscan @ 2.802962
    2 Greek_Thessaly + Irish + South_Italian + Tuscan @ 2.879542
    3 Central_Greek + Central_Greek + North_Dutch + North_Italian @ 2.885310
    4 Central_Greek + North_Dutch + North_Italian + South_Italian @ 2.908030
    5 Greek_Thessaly + South_Italian + Southwest_English + Tuscan @ 2.926807
    6 Central_Greek + Greek_Thessaly + Southeast_English + Tuscan @ 2.933780
    7 Central_Greek + Greek_Thessaly + Irish + Tuscan @ 2.970186
    8 Greek_Thessaly + Orcadian + South_Italian + Tuscan @ 3.001070
    9 Greek_Thessaly + South_Italian + Southeast_English + Tuscan @ 3.003525
    10 Central_Greek + Italian_Abruzzo + North_Italian + West_German @ 3.015851
    11 Greek_Thessaly + North_Dutch + South_Italian + Tuscan @ 3.016721
    12 Central_Greek + Central_Greek + North_German + North_Italian @ 3.018565
    13 Central_Greek + North_Italian + South_Italian + Swedish @ 3.021908
    14 Central_Greek + Central_Greek + North_Italian + Southwest_English @ 3.022295

  • #2
    Have you tried Gedmatch's MDLP Project K16 Modern? I personally have a mix of two major European ethnic classifications, total of 4 minor sub-classifications. K16 Modern seems to do much better than any other model I have seen in returning geographically consistent results. I've often wondered whether this success is specific to these two ethnic groups. I have no 'French' ancestry per se, but my mother's Baden-Wuerttemberg ancestry is sometimes so returned, presumably due to the proximity with Alsace-Loraine.

    Comment


    • #3
      Originally posted by Frederator View Post
      Have you tried Gedmatch's MDLP Project K16 Modern? I personally have a mix of two major European ethnic classifications, total of 4 minor sub-classifications. K16 Modern seems to do much better than any other model I have seen in returning geographically consistent results. I've often wondered whether this success is specific to these two ethnic groups. I have no 'French' ancestry per se, but my mother's Baden-Wuerttemberg ancestry is sometimes so returned, presumably due to the proximity with Alsace-Loraine.

      My full K16 results are below. The Bulgarian and Romanian showed up in a lot of my results. I'm not sure if their proximity to the Caucasus is showing up as that? I imagine anything West Asian is from the Italian ancestry, but since I can't trace it back more than 3 generations I'm not sure.

      Some tests show English and Irish but no French. Some show only Italian and Eastern Europe.

      Admix Results (sorted):

      # Population Percent
      1 Neolithic 29.53
      2 Caucasian 28.72
      3 Steppe 17.85
      4 NorthEastEuropean 16.28
      5 NearEast 4.19
      6 Indian 1.67
      7 NorthAfrican 1.38



      Using 1 population approximation:
      1 Italian_NorthIitaly @ 5.550799
      2 Italian_Friul @ 5.668580
      3 Romanian_Gorj @ 5.741944
      4 Macedonian_Macedonia @ 6.000121
      5 Montenegrian_Montenegro @ 6.517783
      6 Italian_Tuscany @ 6.581297
      7 Kosovar_Kosovo @ 6.634248
      8 Swiss_Switzerland @ 6.687255
      9 Corsican_Corsica @ 6.696495
      10 Romanian_Apuseni @ 6.949284
      11 Bulgarian_Bulgaria @ 6.996163
      12 Italian_Bergamo @ 7.082696
      13 Romanian_Romania @ 7.262941
      14 Provencal_Provence @ 7.541073
      15 Serbian_Serbia @ 7.688632
      16 Gagauz_Gagauzia @ 7.997632
      17 Moldavian_Molodva @ 8.448658
      18 Bulgarian_Bulgaria @ 8.874867
      19 Greek_Greece @ 9.172812
      20 Albanian_Albania @ 9.562369

      Using 2 populations approximation:
      1 50% French_France +50% Greek_Athens @ 2.557600


      Using 3 populations approximation:
      1 50% Greek_Macedonia +25% Macedonian_Macedonia +25% Spanish_Pais_Vasco @ 1.912798


      Using 4 populations approximation:
      ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++
      1 Armenian_Vardnis + German_Germany + Macedonian_Macedonia + Spanish_Pais_Vasco @ 1.289357
      2 Armenian_Vardnis + Bulgarian_Bulgaria + German_Germany + Spanish_Pais_Vasco @ 1.356438
      3 Armenian_Vardnis + French_SouthFrance + German_Germany + Macedonian_Macedonia @ 1.400778
      4 Armenian_Vardnis + French_SouthFrance + German_Germany + Montenegrian_Montenegro @ 1.471132
      5 Albanian_Albania + Armenian_Vardnis + Basque_Spain + Pole_EastPoland @ 1.477526
      6 Armenian_Vardnis + Bulgarian_Bulgaria + French_SouthFrance + German_Germany @ 1.491447
      7 Armenian_Vardnis + Basque_Spain + Greek_Thessaloniki + Pole_EastPoland @ 1.518440
      8 Armenian_Vardnis + Basque_Spain + Kosovar_Kosovo + Pole_EastPoland @ 1.521441
      9 Armenian_Vardnis + Basque_Spain + Greek_Greece + Pole_EastPoland @ 1.525240
      10 Armenian_Armenia + Basque_Spain + Kosovar_Kosovo + Sorb_Lusatia @ 1.531886
      11 Armenian_Vardnis + Basque_Spain + Croat_Croatia + Macedonian_Macedonia @ 1.534951
      12 Armenian_Armenia + German_Germany + Montenegrian_Montenegro + Spanish_Pais_Vasco @ 1.539626
      13 Albanian_Albania + Armenian_Vardnis + Basque_Spain + Sorb_Lusatia @ 1.550202
      14 Armenian_Vardnis + German_Germany + Montenegrian_Montenegro + Spanish_Pais_Vasco @ 1.556342
      15 Armenian_Vardnis + Basque_Spain + Montenegrian_Montenegro + Serbian_Bosnia-Herzegovina @ 1.557404
      16 Armenian_Vardnis + Basque_Spain + Bulgarian_Bulgaria + Croat_Croatia @ 1.557853
      17 Armenian_Vardnis + Basque_Spain + Greek_Greece + Sorb_Lusatia @ 1.558477
      18 Armenian_Vardnis + Basque_Spain + Macedonian_Macedonia + Serbian_Bosnia-Herzegovina @ 1.564194
      19 Armenian_Vanatur_Hrazdan + Bulgarian_Bulgaria + German_Germany + Spanish_Pais_Vasco @ 1.564244
      20 Armenian_Vardnis + Basque_Spain + Greek_Thessaloniki + Sorb_Lusatia @ 1.578326

      Comment


      • #4
        Interesting.

        My methodology for providing personal clarification around the dizzying array of sometimes trivial population references is to group them into major geographic (not necessarily ethnic) categories, and then look for consistency among the admixture scenarios. Here's what my method shows for your K16 results:

        Group # 1: 'Armenian'--20 instances.

        Group # 2: 'Albanian'--8 Greek, 8 Serbian, 2 Albanian, 2 Croat. Although clearly in the minority, I designated this group as 'Albanian' just because you told me your father's heritage is Italian, and I know that there has or had historically been a considerable minority of ethnic Italians on the Adriatic coast of Albania.

        Group # 3: 'German' -4 Bulgarian, 8 Pole, 8 German. This might be the least geographically coherent group in my mind, partially because Germans are comprised of so many very different subcategories. I am tempted to wonder whether some of the explicitly 'German' readings here relate to immigrant ancestors from Alsace-Loraine to Quebec, but on the other hand their presence in this group could refer to Ostdeutsch people, barely distinguishable with Poles or Silesian Czechs. None the less, by default with the consistency of the other groups, this is the major category that most closely fits for these K16 reference populations.

        Group # 4: 'Pyrenees'--17 Spanish, 3 French.

        Note that the total # of instances for each group totals to 20, the number of different admixture scenarios reported by Gedmatch's K16 utility.

        It would be great if you had at least one of your parents test. Then we could look at their admixtures separately to get some more clarity. I find that even the best of calculators struggle to accurately portray even just 3 major ethnic classifications.

        My sense is that Groups # 1 and # 2 represent your father's Italian ancestry, Group # 4 represents most of your mother's French ancestry, and Group # 3 represents a mix of her French and British Isles ancestry. If you have less than a full 25% of British Isles ancestry, the parameters for K16 force it to average it out with non-British ancestry, consistently muddying the waters.

        In my experience, even for Irish-born persons, K16 will sometimes report some fraction of Irish ancestry results as Ostdeutsch or ancestry from north-eastern Europe. I imagine this is even more likely to happen in where a donor has less than a full 25% of their actual ancestry from Ireland.
        Last edited by Frederator; 7th November 2017, 01:09 PM.

        Comment


        • #5
          Wow. That's impressive.
          I hadn't thought of working out that way. How does this reflect the results I've received from Ancestry, MyHeritage and FTDNA though? I've read the Middle Eastern/Caucasian/West Asian can be attributed to the Italian ancestry, but it seems an awfully high percentage to be ancient DNA.

          The Eastern Europe results across the board is also throwing me for a bit of a loop, as well as the Iberian peninsula. Are these regularly attributed to Italian or French ancestry? I have to say the Group#3 seems strange to me as well.

          I am not as familiar with the regional names as I am the ethnicity names so that's why I'm referring to those.

          I do plan on getting a test for my father, since his ancestry is the bigger of the two mysteries.

          Are there any areas the K16 tends to exclude in it's results that may not be coming up for me?

          Additionally, is there a general length of time going back where it makes sense to stop looking for ethnicity? I'm not sure at what migration points it's just all too muddled. I know it's more personal than anything.

          Comment


          • #6
            Originally posted by euromutt View Post
            Wow. That's impressive.
            I hadn't thought of working out that way. How does this reflect the results I've received from Ancestry, MyHeritage and FTDNA though? I've read the Middle Eastern/Caucasian/West Asian can be attributed to the Italian ancestry, but it seems an awfully high percentage to be ancient DNA.
            I don't feel familiar enough with their reference populations to do a similar summarization. Maybe this is a personal limitation of mine rather than the methodology of summarization itself, because I focus so obsessively on such a small group of northern European major ethnic groups.


            Originally posted by euromutt View Post
            The Eastern Europe results across the board is also throwing me for a bit of a loop, as well as the Iberian peninsula. Are these regularly attributed to Italian or French ancestry? I have to say the Group#3 seems strange to me as well.
            A correspondence between actual Italian ancestry and Eastern European reference populations makes total sense to me, just because of what little I know of Italian history. As I mentioned in my previous response, I know there were Italian colonies in Albania. But, maybe more apropos of your own situation, Sicily and southern Italy were heavily colonized by the Greeks. Maybe the most appropriate tag for Group # 2 would be 'Greek'.

            I think the Iberian stuff likely refers to your Quebecois ancestry. I myself don't have any Quebecois ancestry, but I do have a couple of matches with significant Louisiana Cajun ancestry, a closely related group. While the founding population of the Quebecois came from all over France, maybe it's possible that a significant proportion of your specific ancestors came from the southern parts of France? The ones near the border with Spain? The tag "Velasco" in a lot of your Spanish reference populations clearly alludes to the Basque country, right on the border between France and Spain.



            Originally posted by euromutt View Post
            Are there any areas the K16 tends to exclude in it's results that may not be coming up for me?
            I think you're on to something here. I think the builders of K16 may have had more knowledge or generally done a better job of building representative reference populations for northern Europe rather than southern Europe, where I guess most of your ancestry comes from. Maybe using my summarization methodology would produce more useful results if I became more familiar with the reference population of the other models.

            Originally posted by euromutt View Post
            Additionally, is there a general length of time going back where it makes sense to stop looking for ethnicity? I'm not sure at what migration points it's just all too muddled. I know it's more personal than anything.
            My sense is that any ancestor whose ethnic group represents less than 25% of your total ethnic makeup on paper will not be reported distinctly on any calculator. That is to say, any ancestor more remote than a grandparent.

            I'm not sure if that's because these 4-factor analyses at Gedmatch (the only ones I even partially trust) inherently average contributions which are individually less than 25%, or because there is some more fundamental limit on the accuracy of the calculators in general, no matter how many factors they allow.

            But also remember that your ethnic composition per your paper trail is based only on population averages for grandparental contribution. There is significant variation in actual practice.

            For example, my father's background is, on paper, 25% Irish and 75% German, mostly (50%) from the east end of Germany. The MDLP Project K16 reading for my own phased paternal results just about perfectly matches this paper trail.

            But my brother's phased paternal results show 100% German, 75% from the eastern end of the country. I know for a fact based on a detailed census of the segments where I can confidently attribute a specific MRCA, my brother inherited a lot fewer of my father's 'Irish' segments than I did. No doubt his paternal phased results for eastern Germany per K16 are exaggerated since his Irish segments must be less than 25% of the total.

            In actual practice for individual donors, the distribution curve for contribution % of paternal grandfathers is bi-modal, either 40% or 60% (although averaging 50% among all individuals in the population). I guess I got closer to 60% of our father's store of our Irish great grandfather's DNA and my brother closer to 40%.

            So detecting the ethnic contribution even for an ancestor as recent as a grandparent can be tricky.
            Last edited by Frederator; 7th November 2017, 02:40 PM.

            Comment

            Working...
            X