Page 1 of 3 123 LastLast
Results 1 to 10 of 29

Thread: Help interpreting GEDmatch results

  1. #1
    Registered Users
    Posts
    1,265
    Sex
    Location
    Viseu
    Ethnicity
    Romanised Paesuri
    Nationality
    Portuguese
    Y-DNA
    E-BY36858
    mtDNA
    H20

    Asturias Portugal 1143 Portugal 1485 Portugal Order of Christ

    Help interpreting GEDmatch results

    Hello everyone, I'm starting this thread to basically ask for help interpreting my results, particularly on eurogenes GEDmatch K13, K15 and K36 calculators.
    What I want to know from people who are more experienced than me is if this deviation is normal, or if there's reason to believe this is real and a product of uncommon ancestry - such as undersampling from my place of origin which could be hiding a different structure. It's a bit of a long one with many numbers, so please bear with me.
    While very clearly Iberian, the results seem to deviate from the typical Portuguese results in having a markedly western, almost Basque-like, trend.




    For starters all my known ancestors up to late 1800s are of Portuguese origin, with typical Portuguese family names, so I'm not expecting to have any significant amount of foreign ancestry in the recent past. 75% of them are from the interior north of the country, the remaining 25% from a few kms west of Porto. With that in mind, and knowing that historically the interior north is poor and not a place of destination for migrants or traders, when I took the 23andme test I wasn't expecting a high amount of foreign hits on the ancestry composition, however I wasn't expecting such a low amount of it, it seems non-natives avoided it like the plague.


    Iberian 84.5% (Portugal)
    Italian 2.0%
    Broadly Southern European 7.7%
    Broadly Northwestern European 3.1%
    Broadly European 2.1%

    So that's 94.2% southern Euro, which kind of took me by surprise as most Portuguese score significantly less. I checked some information to confirm my suspicions, firstly the eupedia maps, a secondly a website with a small collection of samples http://www.iberiandna.com/ . Eupedia maps might not exactly be scientific literature, but for the purpose of simple comparison they work okay.
    Iberian component map from 23andme lists Portugal as roughly 60-70% - so markedly lower than mine - the remaining common components (British, etc) follow the opposite trend, with mine being considerably lower than average. iberiandna.com data agrees with this, and eventhough the number of samples on certain regions is indeed very low, my results seem to be more in line with Navarra, Pais Vasco and Cantabria - all with over 90% South and minimal North Euro. Obviously this not not strictly true, so I decided to convert the 23andme V5 files to V3 and uploaded them to GEDmatch to further look into it. alexfritz asked me to check the number of SNPs involved in the calculations, and they all seem solid enough for reliable results.


    I started with K13, the results weren't uncommon, but the oracle already gives hints of a less-than-optimal fit with Portuguese, as it never shows up on mixed mode.
    The 4 pop approximation, however, sometimes seeks a very western source.

    Admix Results (sorted):

    # Population Percent
    1 North_Atlantic 39.42
    2 West_Med 28.51
    3 East_Med 12.04
    4 Baltic 9.58
    5 Red_Sea 4.47
    6 Northeast_African 2.60
    7 Sub-Saharan 1.13
    8 West_Asian 1.05

    Using 1 population approximation:
    1 Spanish_Castilla_La_Mancha @ 3.309796
    2 Spanish_Castilla_Y_Leon @ 3.629709
    3 Spanish_Extremadura @ 3.937467
    4 Spanish_Galicia @ 4.226626
    5 Spanish_Murcia @ 4.278359
    6 Portuguese @ 4.760566
    7 Spanish_Cantabria @ 4.761722
    8 Spanish_Andalucia @ 4.834264
    9 Spanish_Valencia @ 5.333640
    10 Spanish_Aragon @ 5.788502
    11 Spanish_Cataluna @ 6.111313
    12 Southwest_French @ 7.455640
    13 North_Italian @ 13.241318
    14 French @ 13.557904
    15 French_Basque @ 18.597870
    16 Tuscan @ 19.949224
    17 South_Dutch @ 20.338776
    18 West_German @ 20.832329
    19 Southeast_English @ 24.712744
    20 Southwest_English @ 24.838282

    Using 2 populations approximation:
    1 50% Spanish_Castilla_La_Mancha +50% Spanish_Galicia @ 2.842885


    Using 3 populations approximation:
    1 50% French_Basque +25% South_Dutch +25% Tunisian @ 2.554944


    Using 4 populations approximation:
    ++++++++++++++++++++++++++++++++++++++++++++++++++ +++++++++++++++++++++++++++++
    1 French_Basque + French_Basque + South_Dutch + Tunisian @ 2.554944
    2 French_Basque + French_Basque + Tunisian + West_German @ 2.665070
    3 French + French_Basque + French_Basque + Tunisian @ 2.665232
    4 Spanish_Andalucia + Spanish_Aragon + Spanish_Galicia + Spanish_Galicia @ 2.775472
    5 Spanish_Aragon + Spanish_Castilla_La_Mancha + Spanish_Extremadura + Spanish_Galicia @ 2.787880
    6 Algerian + French_Basque + French_Basque + West_German @ 2.803577
    7 Spanish_Aragon + Spanish_Castilla_Y_Leon + Spanish_Extremadura + Spanish_Galicia @ 2.808777
    8 Spanish_Aragon + Spanish_Castilla_La_Mancha + Spanish_Galicia + Spanish_Galicia @ 2.810401
    9 Spanish_Castilla_La_Mancha + Spanish_Castilla_La_Mancha + Spanish_Extremadura + Spanish_Galicia @ 2.812029
    10 Spanish_Castilla_La_Mancha + Spanish_Castilla_La_Mancha + Spanish_Castilla_La_Mancha + Spanish_Galicia @ 2.827235
    11 Algerian + French + French_Basque + French_Basque @ 2.831906

    ph2ter plotted me on his neat K13 map, my position shows the western trend, but not exaggeratedly so. Please note that the sampled individuals marked in white are also Portuguese.





    I get a lot of Basque on the 4-pop, but maybe this isn't too uncommon, so I jumped into the K15 to see what it says:

    # Population Percent
    1 Atlantic 34.66
    2 West_Med 22.82
    3 North_Sea 15.44
    4 East_Med 9.05
    5 Baltic 6.16
    6 Red_Sea 4.62
    7 Northeast_African 2.90
    8 Eastern_Euro 2.17
    9 Sub-Saharan 1.05

    Using 1 population approximation:
    1 Spanish_Aragon @ 4.667961
    2 Spanish_Castilla_La_Mancha @ 5.150952
    3 Spanish_Valencia @ 5.897346
    4 Spanish_Andalucia @ 6.064717
    5 Spanish_Cantabria @ 7.080356
    6 Southwest_French @ 7.101790
    7 Spanish_Castilla_Y_Leon @ 7.859376
    8 Spanish_Murcia @ 8.023123
    9 Spanish_Extremadura @ 8.708567
    10 Spanish_Cataluna @ 9.743428
    11 Portuguese @ 10.739885
    12 Spanish_Galicia @ 12.926400
    13 French_Basque @ 13.731232
    14 North_Italian @ 15.190049
    15 French @ 18.670847
    16 Tuscan @ 21.475231
    17 South_Dutch @ 22.907410
    18 Austrian @ 26.084274
    19 Southwest_English @ 26.341728
    20 Serbian @ 26.759363

    Using 2 populations approximation:
    1 50% Spanish_Aragon +50% Spanish_Aragon @ 4.667961


    Using 3 populations approximation:
    1 50% French_Basque +25% Southwest_French +25% Tunisian @ 3.362905


    Using 4 populations approximation:
    ++++++++++++++++++++++++++++++++++++++++++++++++++ +++++++++++++++++++++++++++++++++
    1 Algerian + French_Basque + French_Basque + Southwest_French @ 3.127435
    2 French_Basque + French_Basque + Southwest_French + Tunisian @ 3.362905
    3 Algerian + French_Basque + French_Basque + Spanish_Aragon @ 3.564496
    4 French_Basque + French_Basque + Moroccan + Southwest_French @ 3.593437
    5 French_Basque + French_Basque + Spanish_Cantabria + Tunisian @ 3.631227
    6 Algerian + French_Basque + French_Basque + Spanish_Cantabria @ 3.637549
    7 French_Basque + French_Basque + Mozabite_Berber + Spanish_Cataluna @ 3.704723
    8 French_Basque + French_Basque + Moroccan + Spanish_Cataluna @ 3.714605
    9 French_Basque + French_Basque + Mozabite_Berber + Southwest_French @ 3.745634
    10 French_Basque + French_Basque + Spanish_Cataluna + Tunisian @ 3.760424
    11 Algerian + French_Basque + French_Basque + Spanish_Cataluna @ 3.765542
    12 Algerian + French_Basque + French_Basque + Spanish_Castilla_Y_Leon @ 3.806783
    13 French_Basque + French_Basque + Spanish_Castilla_Y_Leon + Tunisian @ 3.832284
    14 Algerian + French_Basque + French_Basque + Spanish_Valencia @ 3.854127
    15 French_Basque + French_Basque + Mozabite_Berber + Spanish_Castilla_Y_Leon @ 3.868355
    16 French_Basque + French_Basque + Spanish_Aragon + Tunisian @ 3.907134
    17 French + French_Basque + French_Basque + Mozabite_Berber @ 3.909878
    18 French_Basque + French_Basque + Moroccan + Spanish_Castilla_Y_Leon @ 3.915656
    19 French_Basque + French_Basque + Moroccan + Spanish_Aragon @ 3.920760
    20 Algerian + French_Basque + French_Basque + Spanish_Castilla_La_Mancha @ 3.944761



    So every single 4-pop mix model gives me at least a whopping 50% Basque. The position on the PCA makes perfect sense, given the results, so I decided to compare them with the average results from Portugal and various regions on Spain to get a grasp of what is going on. I tried to divide the components into somewhat related groups so they'd be easier to compare.


    I realise some values might differ due to statistical variation (ie lower East Med+Red Sea+West Asian with NE Africa included, I guess), and that is fine and might be roughly what happened in the K13, however I'm pretty sure that gigantic 34.66 Atlantic and minuscule 15.44 North Sea is not, especially since it's also following the trend of having "toned down" Basque-like results.


    Lastly I decided to try the K36, and compared them to the average results form Portugal, and again there's one thing that really stands out.
    I'll just post some of the relevant components instead of all the clutter.
    Basque 5.57
    East_Balkan 2.68
    East_Central_Euro 2.08
    East_Med 1.61
    Eastern_Euro 3.28
    French 4.20
    Iberian 32.03
    Italian 13.70
    Near_Eastern 4.87
    North_African 6.20
    North_Atlantic 8.16
    North_Sea 6.56
    West_Med 6.65

    Here's the K36 similarity map, again showing me rather disconnected form pretty much everyone bar other Iberians.








    So, to get back on the original point, do you think this could be simple random DNA shenanigans, something with the algorithm that might be unfit with me, or is it possible my results are just naturally uncommon for my ancestry? Having such high Iberian/Basque/Atlantic-related and low North Sea makes me believe it could be a product of very low Germanic ancestry dating back to the migration period (ph2ter agreed with it), so more "archaic"? Or could the region have been resettled with a greater amount of individuals with Basque ancestry during the Reconquista period?

    Please share your thoughts, I'm a bit confused with this and wanted to get a better understanding of what is going on.
    Last edited by Ruderico; 10-01-2018 at 02:17 PM.
    YDNA - E-Y31991>PF4428>BY36857. Domingos Rodrigues, b. circa 1680 Hidden Content , Viana do Castelo, Portugal
    mtDNA - H20. Maria Josefa de Almeida, b. circa 1750 Hidden Content , Porto, Portugal

    Global25 PCA West Eurasia dataset Hidden Content
    Hidden Content


    [1] "distance%=1.7726"

    Ruderico

    Celtiberian,77.6
    Roman_Imperial_proxy,13.6
    Guanche,8.8

  2. The Following 8 Users Say Thank You to Ruderico For This Useful Post:

     Amarelo (05-14-2019),  Camulogène Rix (10-01-2018),  kingjohn (10-09-2018),  Robert1 (10-09-2018),  rxavierflima (10-04-2018),  Shadogowah (10-01-2018),  Vivaldi4Stagioni (10-01-2018),  Vrump (10-02-2018)

  3. #2
    Registered Users
    Posts
    366
    Sex
    Location
    Netherlands
    Ethnicity
    Celtic Iberian Jew?
    Nationality
    Spanish
    Y-DNA
    R-Y22889
    mtDNA
    H3-T152C!

    Spain Andalucia Spanish Empire (1506-1701)
    My completely uneducated opinion is that the key parameter is indeed the difference between the two main contributors Atlantic and North Sea.

    Atlantic is always higher than North Sea both for Spaniards and Portuguese but they level up as you move to the NW of Iberia.

    In Galicia, Atlantic component drops and almost levels up with North Sea that is higher there (23/22) while in Basque country the difference between both is much bigger mostly because Atlantic grows a lot while North sea drops a bit (45/15).

    I think you are right and your profile fits better a Cantabrian/North Castilian one than the Portuguese one in that regard.

    We actually have a similar admixture, if you want to run comparisons, I can pass you my kit # by PM.

    In any case we need more detail about the origin of the samples. Rough groupings by region like Castilla and Leon or Andalusia in Spain actually hide these nuances.

    We have discussed before about the West/East gradient.
    Last edited by Shadogowah; 10-01-2018 at 02:47 PM.
    Do you have a question or topic about genetics in Spain and/or prefer to use Spanish language? Visit the new subforum!

    Hidden Content

  4. The Following User Says Thank You to Shadogowah For This Useful Post:

     Ruderico (10-01-2018)

  5. #3
    Registered Users
    Posts
    1,265
    Sex
    Location
    Viseu
    Ethnicity
    Romanised Paesuri
    Nationality
    Portuguese
    Y-DNA
    E-BY36858
    mtDNA
    H20

    Asturias Portugal 1143 Portugal 1485 Portugal Order of Christ
    Quote Originally Posted by Shadogowah View Post
    I think you are right and your profile fits better a Cantabrian/North Castilian one than the Portuguese one in that regard.
    I wonder if this has anything to do with mountain regions being more isolated and removed from Germanic ancestry. Besides NW Portugal and Galicia, I mean, which was the core of Suebi settlement in the peninsula.

    The region of Viseu has a funny geography which walls it a bit from the outside, maybe it had something to do with it. During the Roman wars it was only during the time of Julius Caesar (by the man himself) that they started to invade this particular area, even modern Galicia was attacked first
    Last edited by Ruderico; 10-01-2018 at 03:31 PM.
    YDNA - E-Y31991>PF4428>BY36857. Domingos Rodrigues, b. circa 1680 Hidden Content , Viana do Castelo, Portugal
    mtDNA - H20. Maria Josefa de Almeida, b. circa 1750 Hidden Content , Porto, Portugal

    Global25 PCA West Eurasia dataset Hidden Content
    Hidden Content


    [1] "distance%=1.7726"

    Ruderico

    Celtiberian,77.6
    Roman_Imperial_proxy,13.6
    Guanche,8.8

  6. #4
    Registered Users
    Posts
    164
    Sex
    Location
    BR
    Nationality
    Brazilian
    Y-DNA
    R1b-Z49
    mtDNA
    T2b

    Brazil
    Just out of curiosity, on the Similarity Map, did you try removing your known ancestry (Portuguese) to see how the map looks like?

  7. #5
    Registered Users
    Posts
    1,265
    Sex
    Location
    Viseu
    Ethnicity
    Romanised Paesuri
    Nationality
    Portuguese
    Y-DNA
    E-BY36858
    mtDNA
    H20

    Asturias Portugal 1143 Portugal 1485 Portugal Order of Christ
    Here's the Global 25 PCA, again trending to the western edge of the Iberian cluster

    YDNA - E-Y31991>PF4428>BY36857. Domingos Rodrigues, b. circa 1680 Hidden Content , Viana do Castelo, Portugal
    mtDNA - H20. Maria Josefa de Almeida, b. circa 1750 Hidden Content , Porto, Portugal

    Global25 PCA West Eurasia dataset Hidden Content
    Hidden Content


    [1] "distance%=1.7726"

    Ruderico

    Celtiberian,77.6
    Roman_Imperial_proxy,13.6
    Guanche,8.8

  8. #6
    Interesting not very different from my results

  9. #7
    Registered Users
    Posts
    1,265
    Sex
    Location
    Viseu
    Ethnicity
    Romanised Paesuri
    Nationality
    Portuguese
    Y-DNA
    E-BY36858
    mtDNA
    H20

    Asturias Portugal 1143 Portugal 1485 Portugal Order of Christ
    Quote Originally Posted by emc View Post
    Just out of curiosity, on the Similarity Map, did you try removing your known ancestry (Portuguese) to see how the map looks like?
    You can do that? All I did was insert the K36 components and that was it
    YDNA - E-Y31991>PF4428>BY36857. Domingos Rodrigues, b. circa 1680 Hidden Content , Viana do Castelo, Portugal
    mtDNA - H20. Maria Josefa de Almeida, b. circa 1750 Hidden Content , Porto, Portugal

    Global25 PCA West Eurasia dataset Hidden Content
    Hidden Content


    [1] "distance%=1.7726"

    Ruderico

    Celtiberian,77.6
    Roman_Imperial_proxy,13.6
    Guanche,8.8

  10. #8
    Registered Users
    Posts
    3,082
    Sex
    Location
    Taiwan
    Ethnicity
    Métis
    Nationality
    Canadian
    Y-DNA
    R-Z198 (DF27)
    mtDNA
    T2B-T152C

    Canada England Scotland Germany Poland France
    Here's a Global 25 PCA which has been hastily cut down to try to get a picture of where you're at in relation to Portugal and Spain:

    Attachment 26412

    Orange is Portugal, Red is Spain, and I coloured the Galicians Purple/Pink (I was curious where they would land)

    Are your closest samples Portuguese? Nope, but you're also clearly within the Portuguese cluster. Interestingly it occupies pretty much the same space as the Spanish cluster. There was a study done on the genetics of Portugal and Spain but IIRC to differentiate the clusters it required the use of finestructure? I'm more confident in speaking about Britain and Ireland, and there we see an overlap between England, Ireland, Wales, and Scotland, but with finestructure we can begin to observe more noticeable differences. I think this might be the situation here on this graph, Spain and Portugal are so close they are more or less indistinguishable without the use of finestructure or something similar. Thus, I don't think your results are abnormal Ruderico.
    Paper trail ancestry to the best of my knowledge:
    English (possibly containing some Welsh ancestry) 31.25%, Eastern European and Eastern German (Galicia, Poland) 25%, Scottish 17.96%, Scotch-Irish 12.5%, French 8.2%, Native American 1.95%, and Colonial American, 3.125%, which cannot be determined with complete certainty: there is Dutch (at least 1.36%) and some English. The rest could include Spanish, Norwegian, German, and French, but these percentages would be minuscule.

  11. The Following 2 Users Say Thank You to sktibo For This Useful Post:

     Robert1 (10-09-2018),  Ruderico (10-03-2018)

  12. #9
    Registered Users
    Posts
    164
    Sex
    Location
    BR
    Nationality
    Brazilian
    Y-DNA
    R1b-Z49
    mtDNA
    T2b

    Brazil
    Quote Originally Posted by Ruderico View Post
    You can do that? All I did was insert the K36 components and that was it
    At the right side there is a box called "Foncion B". At the bottom of it you'll find a list of populations, and you can choose one to remove. I know you're not mixed, but if the map says you're "90% Portuguese", for example, you can remove Portuguese and then it will show what your remaining 10% might be. In my case it didn't help much because North Italians and Portuguese are similar.

  13. The Following User Says Thank You to emc For This Useful Post:

     Ruderico (10-03-2018)

  14. #10
    Registered Users
    Posts
    1,265
    Sex
    Location
    Viseu
    Ethnicity
    Romanised Paesuri
    Nationality
    Portuguese
    Y-DNA
    E-BY36858
    mtDNA
    H20

    Asturias Portugal 1143 Portugal 1485 Portugal Order of Christ
    Quote Originally Posted by sktibo View Post
    Here's a Global 25 PCA which has been hastily cut down to try to get a picture of where you're at in relation to Portugal and Spain:

    Attachment 26412

    Orange is Portugal, Red is Spain, and I coloured the Galicians Purple/Pink (I was curious where they would land)

    Are your closest samples Portuguese? Nope, but you're also clearly within the Portuguese cluster. Interestingly it occupies pretty much the same space as the Spanish cluster. There was a study done on the genetics of Portugal and Spain but IIRC to differentiate the clusters it required the use of finestructure? I'm more confident in speaking about Britain and Ireland, and there we see an overlap between England, Ireland, Wales, and Scotland, but with finestructure we can begin to observe more noticeable differences. I think this might be the situation here on this graph, Spain and Portugal are so close they are more or less indistinguishable without the use of finestructure or something similar. Thus, I don't think your results are abnormal Ruderico.
    Thank you very much!




    Quote Originally Posted by emc View Post
    At the right side there is a box called "Foncion B". At the bottom of it you'll find a list of populations, and you can choose one to remove. I know you're not mixed, but if the map says you're "90% Portuguese", for example, you can remove Portuguese and then it will show what your remaining 10% might be. In my case it didn't help much because North Italians and Portuguese are similar.
    Here it is, no idea what it's supposed to mean though
    YDNA - E-Y31991>PF4428>BY36857. Domingos Rodrigues, b. circa 1680 Hidden Content , Viana do Castelo, Portugal
    mtDNA - H20. Maria Josefa de Almeida, b. circa 1750 Hidden Content , Porto, Portugal

    Global25 PCA West Eurasia dataset Hidden Content
    Hidden Content


    [1] "distance%=1.7726"

    Ruderico

    Celtiberian,77.6
    Roman_Imperial_proxy,13.6
    Guanche,8.8

  15. The Following User Says Thank You to Ruderico For This Useful Post:

     sktibo (10-03-2018)

Page 1 of 3 123 LastLast

Similar Threads

  1. Help interpreting GEDmatch results?
    By Shanatova in forum General
    Replies: 1
    Last Post: 09-10-2018, 12:14 AM
  2. Help Interpreting Gedmatch Results
    By Candycola in forum AncestryDNA
    Replies: 5
    Last Post: 04-09-2018, 03:38 PM
  3. Help with Interpreting GedMatch Results
    By MeloG91 in forum Genealogy
    Replies: 1
    Last Post: 10-10-2017, 01:41 PM
  4. Interpreting my Gedmatch results
    By Inal in forum Genealogy
    Replies: 16
    Last Post: 10-05-2017, 11:06 AM
  5. Help with interpreting Gedmatch results
    By davidram in forum Autosomal (auDNA)
    Replies: 3
    Last Post: 09-04-2014, 07:15 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •