Page 2 of 4 FirstFirst 1234 LastLast
Results 11 to 20 of 38

Thread: Question about Iberia CHL

  1. #11
    Registered Users
    Posts
    293
    Sex
    Y-DNA
    J2a-M319

    Quote Originally Posted by Kale View Post
    F3 outgroup nmonte model

    Iberia_CA
    "Barcin_N" 40.05
    "Greece_N" 17.8
    "Peloponnese_N" 16.95
    "Bichon" 10.2
    "LaBrana1" 9.9
    "KO1" 4.2
    "Mentese_N" 0.8
    "Mota" 0.1
    "distance%=0.1511 / distance=0.001511"
    Columns: Boncuklu_N, Iran_N, Taforalt, Kotias, MA1, Anzick_Kennewick, ElMiron, Iron_Gates_HG, Motala_HG, Ukraine_Mesolithic, Ust_Ishim

    Nothing out of the ordinary. No problems at all fitting to any of the columns.

    Okay, so they basically are a mixture of anatolian farmers and whg.

  2. #12
    Registered Users
    Posts
    2,621
    Sex
    Location
    Groningen
    Ethnicity
    Northwest European
    Nationality
    NL
    Y-DNA
    E-V22

    Netherlands
    Quote Originally Posted by Sorcelow View Post
    Are they simply a fusion between Anatolian farmers and WHG, or do they represent a completely standalone population? They seem incredibly western, more so than WHG.
    Keeps an interesting question, despite the intermingling with the Anatolian farmer (see your conclusion)the K18 makes indeed difference in the results apparently, what makes that that whole Atlantic Europe uo to Norway has Iberia CHL and from NE Dutch (and I guess so NE France to the east) incl. Sweden has Barcin_N.

    Is it indeed LBK/GAC and TRB (with higher Anatolian)? Anyone???

  3. The Following 3 Users Say Thank You to Finn For This Useful Post:

     JMcB (04-07-2019),  Nibelung (04-06-2019),  Sorcelow (04-06-2019)

  4. #13
    Moderator
    Posts
    5,410
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA
    U152>L2>Z367
    mtDNA
    H5a1

    Normandie Netherlands Friesland Finland Orkney
    Quote Originally Posted by Finn View Post
    Keeps an interesting question, despite the intermingling with the Anatolian farmer (see your conclusion)the K18 makes indeed difference in the results apparently, what makes that that whole Atlantic Europe uo to Norway has Iberia CHL and from NE Dutch (and I guess so NE France to the east) incl. Sweden has Barcin_N.

    Is it indeed LBK/GAC and TRB (with higher Anatolian)? Anyone???
    If what you call "K18" is the G25+nMonte-calculator recently elaborated by Sorcelow, the fact that precisely Iberia_CHL is for a great part descending from an Anatolian neolithic population at least strongly correlated to Barcin is a real problem. One of the major strong points of qpAdm (only one of them) is this one, that some people certainly would wrongly consider as a weak point: if you want to model A as a combination of B, C and D, C and D being related ("non orthogonal"), you will get great standard-errors on C and D coefficients. As the standard-errors are not computed in a standard nMonte result, the door is open to serious hidden problems. Back to the question, I can hardly imagine what Iberia_CHL could be, else that the result of the genetic drift acting on a fusion of Anatolia_N (with perhaps a minor contribution of Levant_N) and a WHG autochtonous population. Furthermore, if I'm well understanding that you believe in a genetic imprint of the iberian_CHL populations into all the Atlantic populations (do you?), that's nothing but a fable.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  5. The Following 3 Users Say Thank You to anglesqueville For This Useful Post:

     Finn (04-06-2019),  JMcB (04-07-2019),  Ruderico (04-06-2019)

  6. #14
    Registered Users
    Posts
    2,621
    Sex
    Location
    Groningen
    Ethnicity
    Northwest European
    Nationality
    NL
    Y-DNA
    E-V22

    Netherlands
    Quote Originally Posted by anglesqueville View Post
    If what you call "K18" is the G25+nMonte-calculator recently elaborated by Sorcelow, the fact that precisely Iberia_CHL is for a great part descending from an Anatolian neolithic population at least strongly correlated to Barcin is a real problem. One of the major strong points of qpAdm (only one of them) is this one, that some people certainly would wrongly consider as a weak point: if you want to model A as a combination of B, C and D, C and D being related ("non orthogonal"), you will get great standard-errors on C and D coefficients. As the standard-errors are not computed in a standard nMonte result, the door is open to serious hidden problems. Back to the question, I can hardly imagine what Iberia_CHL could be, else that the result of the genetic drift acting on a fusion of Anatolia_N (with perhaps a minor contribution of Levant_N) and a WHG autochtonous population. Furthermore, if I'm well understanding that you believe in a genetic imprint of the iberian_CHL populations into all the Atlantic populations (do you?), that's nothing but a fable.
    So Iberian CHL looks like Barcin-N combined with WHG....Agree!

    But looking at the outcomes of the model, we see indeed a difference between more West European and NE European results. The difference between me and my countryman Radboud is quite big in this model.

    "sample": "Custom:AGUser_Radboud",
    "fit": 3.2417,
    "Steppe_k18": 50.83,
    "European_Farmer_k18": 40,
    "Anatolian_Farmer_k18": 5.83,
    "Baltic_HG_k18": 3.33

    "sample": "Custom:AGUser_finn",
    "fit": 3.9437,
    "Steppe_k18": 46.67,
    "Anatolian_Farmer_k18": 20.83,
    "European_Farmer_k18": 18.33,
    "Baltic_HG_k18": 14.17

    Norway and Sweden also look kind of different which is more often the case (Sweden has a more NE tendency).
    But not only Anatolian farmer is at stake it seems also a difference between Baltic HG and WHG. It's not weird to think that my HG is from Swifterbant/Ertebølle, unlike the Irish for example.



    So indeed that leads to a thought that there is a difference between the influence from Spain to West Europe, the Atlantic front, and more inland Anatolian through the Balkans up to Central, North Europe. LBK/GAC/TRB. Or?
    Last edited by Finn; 04-06-2019 at 06:23 PM.

  7. The Following 2 Users Say Thank You to Finn For This Useful Post:

     Drewcastle (04-07-2019),  Nibelung (04-06-2019)

  8. #15
    Registered Users
    Posts
    276
    Sex
    Ethnicity
    Assyrian
    Y-DNA
    G1a
    mtDNA
    I5b1

    Sweden
    Quote Originally Posted by Finn View Post
    So Iberian CHL looks like Barcin-N combined with WHG....Agree!

    But looking at the outcomes of the model, we see indeed a difference between more West European and NE European results. The difference between me and my countryman Radboud is quite big in this model.

    "sample": "Custom:AGUser_Radboud",
    "fit": 3.2417,
    "Steppe_k18": 50.83,
    "European_Farmer_k18": 40,
    "Anatolian_Farmer_k18": 5.83,
    "Baltic_HG_k18": 3.33

    "sample": "Custom:AGUser_finn",
    "fit": 3.9437,
    "Steppe_k18": 46.67,
    "Anatolian_Farmer_k18": 20.83,
    "European_Farmer_k18": 18.33,
    "Baltic_HG_k18": 14.17

    Norway and Sweden also look kind of different which is more often the case (Sweden has a more NE tendency).
    But not only Anatolian farmer is at stake it seems also a difference between Baltic HG and WHG. It's not weird to think that my HG is from Swifterbant/Ertebølle, unlike the Irish for example.



    So indeed that leads to a thought that there is a difference between the influence from Spain to West Europe, the Atlantic front, and more inland Anatolian through the Balkans up to Central, North Europe. LBK/GAC/TRB. Or?
    I think in Sorcelow's run if someone or if an average has just slightly more WHG-related ancestry or escessive EHG/SHG than what the Iberian farmers and the Steppe averge do then the whole run gets kinda freaky, as then nMonte rather prefers just giving Barcin/Anatolian and Baltic_HG.
    K18 doesn't include GAC samples. Once they are included the difference between Atlantic and inland Western Europe is not big at all.

    "distance%=3.382"

    Norwegian

    Yamnaya_Kalmykia,48.6
    Poland_GAC,46.8
    Narva_Lithuania,4.6
    Barcin_N,0
    Iberia_Central_CA,0

    "distance%=3.3473"

    Swedish

    Yamnaya_Kalmykia,47.6
    Poland_GAC,44.8
    Narva_Lithuania,7.6
    Barcin_N,0
    Iberia_Central_CA,0

    Only slightly more HG-related admix in the Swedish average as expected as it inlcudes Swedes from all over Sweden.

  9. The Following 2 Users Say Thank You to Helves For This Useful Post:

     anglesqueville (04-07-2019),  JMcB (04-07-2019)

  10. #16
    Registered Users
    Posts
    2,621
    Sex
    Location
    Groningen
    Ethnicity
    Northwest European
    Nationality
    NL
    Y-DNA
    E-V22

    Netherlands
    Quote Originally Posted by Helves View Post
    I think in Sorcelow's run if someone or if an average has just slightly more WHG-related ancestry or escessive EHG/SHG than what the Iberian farmers and the Steppe averge do then the whole run gets kinda freaky, as then nMonte rather prefers just giving Barcin/Anatolian and Baltic_HG.
    K18 doesn't include GAC samples. Once they are included the difference between Atlantic and inland Western Europe is not big at all.

    "distance%=3.382"

    Norwegian

    Yamnaya_Kalmykia,48.6
    Poland_GAC,46.8
    Narva_Lithuania,4.6
    Barcin_N,0
    Iberia_Central_CA,0

    "distance%=3.3473"

    Swedish

    Yamnaya_Kalmykia,47.6
    Poland_GAC,44.8
    Narva_Lithuania,7.6
    Barcin_N,0
    Iberia_Central_CA,0

    Only slightly more HG-related admix in the Swedish average as expected as it inlcudes Swedes from all over Sweden.
    Yes I see, you elaborate it further, that I was looking for, thanks!

    Still it looks like that there is a shift between West Europe and more N(E) Europe, see this results, I diver from Jessie and also from my countryman Radboud (he is from the west , I'm from the northeast). I have Swedish kind of results.....but that's especially due to the HG indeed!

    Poi's tool, pen=0

    The Ertebølle/TRB effect?


    "sample": "Custom:AGUser_Jessie",
    "fit": 3.1391,
    "Yamnaya_Kalmykia": 51.67,
    "Globular_Amphora_Poland": 35,
    "Iberia_Central_CA": 8.33,
    "Narva_Lithuania": 5,
    "Barcin_N": 0,

    "sample": "Custom:AGUser_Radboud",
    "fit": 2.9272,
    "Yamnaya_Kalmykia": 50.83,
    "Globular_Amphora_Poland": 28.33,
    "Iberia_Central_CA": 19.17,
    "Narva_Lithuania": 1.67,
    "Barcin_N": 0,

    "sample": "Custom:AGUser_finn",
    "fit": 3.4897,
    "Yamnaya_Kalmykia": 49.17,
    "Globular_Amphora_Poland": 42.5,
    "Narva_Lithuania": 7.5,
    "Barcin_N": 0.83,
    "Iberia_Central_CA": 0,

    "sample": "Dutch:Average",
    "fit": 2.7847,
    "Yamnaya_Kalmykia": 49.17,
    "Globular_Amphora_Poland": 47.5,
    "Narva_Lithuania": 2.5,
    "Iberia_Central_CA": 0.83,
    "Barcin_N": 0,

    "sample": "Irish:Average",
    "fit": 2.8862,
    "Yamnaya_Kalmykia": 50.83,
    "Globular_Amphora_Poland": 39.17,
    "Iberia_Central_CA": 7.5,
    "Narva_Lithuania": 2.5,
    "Barcin_N": 0,

    "sample": "Norwegian:Average",
    "fit": 3.1287,
    "Yamnaya_Kalmykia": 50.83,
    "Globular_Amphora_Poland": 44.17,
    "Narva_Lithuania": 5,
    "Barcin_N": 0,
    "Iberia_Central_CA": 0,

    "sample": "Swedish:Average",
    "fit": 3.203,
    "Yamnaya_Kalmykia": 49.17,
    "Globular_Amphora_Poland": 35.83,
    "Narva_Lithuania": 9.17,
    "Barcin_N": 5.83,
    "Iberia_Central_CA": 0,

  11. #17
    Moderator
    Posts
    5,410
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA
    U152>L2>Z367
    mtDNA
    H5a1

    Normandie Netherlands Friesland Finland Orkney
    Yes, I strongly agree with Helves. I made myself a quick experiment on the basis of Sorcelow's spreadsheet. I didn't post it because I didn't want to spoil Sorcelow's thread, but I can do it if he wants, and if someone finds it useful. Shortly said, I removed all the african, oceanian and african references (unuseful in the context of this experiment), removed the "Euro-Farmer" in order to solve the problem of overlapping between EEF and Barcin, unscaled everything, and ran with pen=0. My hope was to be able to see a clearer breakdown Euro-HG/Baltic-HG. What eventually happens is only that Baltic_HG is ridiculously dominating on Euro_HG. In particular, just to look on Norw vs Swed:

    [1] "distance%=2.386"

    Norwegian

    Steppe,57
    Anatolian_Farmer,34.6
    Baltic_HG,7.4 ........... no Euro_HG at all.
    Siberian,1


    [1] "distance%=1.9507"

    Swedish

    Steppe,53.8
    Anatolian_Farmer,33.2
    Baltic_HG,12
    Siberian,1


    Another problem, as always, is the use of averages as reference dots. For example:

    [1] "distance%=2.4554"

    Irish

    Steppe,58
    Anatolian_Farmer,35.4
    Euro_HG,6.6

    I guess you would say "yeah! Euro_HG!". Now take a pure irish actual individual:


    [1] "distance%=2.3069"

    Jessie

    Steppe,55.8
    Anatolian_Farmer,33.6
    Baltic_HG,7.2
    Euro_HG,3.4

    Well, in fact, there is here nothing really surprising. The only directing parameter on the HG continuum (between extreme western WHG, say Loschbour or La Brana, and extreme eastern EHG, say the Karelian one) is the affinity to ANE. In other words, between Euro-HG and Baltic-HG, there is also a terrible overlapping, and therefore a terrible hidden fluctuation in the way nMonte chooses the references.
    I know that my opinion on these issues is not very popular, but I repeat that if your goal is to reveal very fine phenomena, these friendly amateur calculators are not reliable tools.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  12. The Following 3 Users Say Thank You to anglesqueville For This Useful Post:

     Drewcastle (04-07-2019),  Helves (04-07-2019),  JMcB (04-07-2019)

  13. #18
    Registered Users
    Posts
    2,621
    Sex
    Location
    Groningen
    Ethnicity
    Northwest European
    Nationality
    NL
    Y-DNA
    E-V22

    Netherlands
    Quote Originally Posted by anglesqueville View Post
    Yes, I strongly agree with Helves. I made myself a quick experiment on the basis of Sorcelow's spreadsheet. I didn't post it because I didn't want to spoil Sorcelow's thread, but I can do it if he wants, and if someone finds it useful. Shortly said, I removed all the african, oceanian and african references (unuseful in the context of this experiment), removed the "Euro-Farmer" in order to solve the problem of overlapping between EEF and Barcin, unscaled everything, and ran with pen=0. My hope was to be able to see a clearer breakdown Euro-HG/Baltic-HG. What eventually happens is only that Baltic_HG is ridiculously dominating on Euro_HG. In particular, just to look on Norw vs Swed:

    [1] "distance%=2.386"

    Norwegian

    Steppe,57
    Anatolian_Farmer,34.6
    Baltic_HG,7.4 ........... no Euro_HG at all.
    Siberian,1


    [1] "distance%=1.9507"

    Swedish

    Steppe,53.8
    Anatolian_Farmer,33.2
    Baltic_HG,12
    Siberian,1


    Another problem, as always, is the use of averages as reference dots. For example:

    [1] "distance%=2.4554"

    Irish

    Steppe,58
    Anatolian_Farmer,35.4
    Euro_HG,6.6

    I guess you would say "yeah! Euro_HG!". Now take a pure irish actual individual:


    [1] "distance%=2.3069"

    Jessie

    Steppe,55.8
    Anatolian_Farmer,33.6
    Baltic_HG,7.2
    Euro_HG,3.4

    Well, in fact, there is here nothing really surprising. The only directing parameter on the HG continuum (between extreme western WHG, say Loschbour or La Brana, and extreme eastern EHG, say the Karelian one) is the affinity to ANE. In other words, between Euro-HG and Baltic-HG, there is also a terrible overlapping, and therefore a terrible hidden fluctuation in the way nMonte chooses the references.
    I know that my opinion on these issues is not very popular, but I repeat that if your goal is to reveal very fine phenomena, these friendly amateur calculators are not reliable tools.
    L'art pour l'art Angles?

    Even with salt I (and also mom and in lesser extent Dad, he is more average NW Germanic) I/we get higher NE-Euro (near Scandic) scores (not due to Steppe but to HG). Radboud mostly gets more (Isles like) W-Euro scores.

    Somehow there must underlying factors, otherwise this kind of stuff is indeed totally l'art pour l'art. Is G25 indeed so 'amateuristic'????

  14. #19
    Moderator
    Posts
    5,410
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA
    U152>L2>Z367
    mtDNA
    H5a1

    Normandie Netherlands Friesland Finland Orkney
    Scientific reliability ideal has nothing to see with "l'Art pour l'Art". G25 per se is smartpca (highly professional tool) on a very serious spreadsheet (elaborated by someone who strictly speaking is an amateur, but with highly reliable skills). Furthermore what I say about David could be said about Ger Huijbregt and his nMonte: amateur strictly speaking, but with skills and awareness that can not be criticized. The problem is not G25 per se, not nMonte per se, but what happens when amateurs like Sorcelow or myself use G25 and nMonte with restricted spreadsheets. It's of course not a crime, and I don't want to deny that such an exercise can provide a lot of fun, and even can make a very useful pedagogic job. That said, we should not confuse such "calculators" with professional tools like formal stats or fastIBD, etc. I only say that very fine phenomena ask for very fine analysis tools, otherwise you risk to think the moon is made of green cheese. Amateurs we are, and we should never forget what we are. Some issues are too difficult for us and our tools. Is it really impossible to admit this simple fact?
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  15. The Following 3 Users Say Thank You to anglesqueville For This Useful Post:

     fabrice E (04-07-2019),  Finn (04-07-2019),  JMcB (04-07-2019)

  16. #20
    Registered Users
    Posts
    1,040
    Sex
    Omitted

    Quote Originally Posted by anglesqueville View Post
    Well, in fact, there is here nothing really surprising. The only directing parameter on the HG continuum (between extreme western WHG, say Loschbour or La Brana, and extreme eastern EHG, say the Karelian one) is the affinity to ANE. In other words, between Euro-HG and Baltic-HG, there is also a terrible overlapping, and therefore a terrible hidden fluctuation in the way nMonte chooses the references.
    I know that my opinion on these issues is not very popular, but I repeat that if your goal is to reveal very fine phenomena, these friendly amateur calculators are not reliable tools.
    There is another factor here. Baltic_HG has a firm tad Magdalenian, as shown in Extended Data figure 7 in the Dzudzuana preprint. Loschbour has more, La Brana has even more than Loschbour.

  17. The Following 3 Users Say Thank You to epoch For This Useful Post:

     anglesqueville (04-07-2019),  Drewcastle (04-07-2019),  JMcB (04-07-2019)

Page 2 of 4 FirstFirst 1234 LastLast

Similar Threads

  1. Replies: 21
    Last Post: 07-21-2018, 12:31 PM
  2. Haplogroup R1a distribution in Iberia
    By jeanL in forum R1a General
    Replies: 36
    Last Post: 05-17-2018, 07:11 PM
  3. Britain and Iberia- E-BY5219
    By firemonkey in forum E1b-M215
    Replies: 0
    Last Post: 11-30-2017, 09:28 PM
  4. DNA Testing (or lack of) in Iberia
    By curiousII in forum General
    Replies: 2
    Last Post: 09-01-2017, 04:31 PM
  5. Y-dna distribution in Iberia
    By Il Papà in forum General
    Replies: 6
    Last Post: 08-31-2014, 11:26 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
  •