Page 114 of 120 FirstFirst ... 1464104112113114115116 ... LastLast
Results 1,131 to 1,140 of 1197

Thread: The genetic prehistory of the Greater Caucasus[preprint Harvard/Jena]

  1. #1131
    Moderator
    Posts
    5,212
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA
    U152>L2>Z367
    mtDNA
    H5a1

    Normandie Netherlands Friesland Finland Orkney
    Quote Originally Posted by Eterne View Post
    On another note on added elements to the published paper over pre-print, I noticed that their published Fig. 4, based on Supplementary Table 14 (fitting Steppe samples as EHG, CHG, WHG and Anatolian), features a fit for Eneolithic_Steppe and its p-value (which is not in Supplementary Table 14 for whatever reason).

    It seems like Eneolithic_Steppe actually seems to marginally fail to reach the p>0.05 level, as EHG+CHG, with value of only 0.0302.

    This could well just be qpAdm behaving kookily again, but that seems a bit surprising (since EHG+CHG cline looks good in most respects), and I'm not sure why their reviewers didn't question that.
    The only part of the paper that seems to formally fit Eneolithic_Steppe is actually their qpGraph* with some fairly limited outgroups, and they don't seem to have constructed a high p value model for Eneolithic_Steppe in qpAdm with lots of outgroups.



    *"Our fitted qpGraph model recapitulates the genetic separation between the Caucasus and Steppe groups with the Eneolithic steppe individuals deriving more than 60% of ancestry from EHG and the remainder from a CHG-related basal lineage, whereas the Maykop group received about 86.4% from CHG, 9.6% Anatolian farming related ancestry, and 4% from EHG. The Yamnaya individuals from the Caucasus derived the majority of their ancestry from Eneolithic steppe individuals, but also received about 16% from Globular Amphora-related farmers"
    I worked a lot on this issue, without being able to find for my Steppe_EneolithicW ( the Prog2) a model in term of a 2-way CHG+EHG with a satisfying tail_prob. Now, look on this:

    left pops:
    Steppe_EneolithicW
    CHG_KK1
    EHG_Samara
    AfontovaGora3

    right pops:
    Mota
    Ust_Ishim
    Kostenki14
    MA1
    WHG
    Iron_Gates_HG
    Iran_N
    Barcin
    LBK_Austria
    West_Siberia_N
    EHG

    numsnps used: 82371

    best coefficients: 0.476 0.377 0.147

    Jackknife mean: 0.475732219 0.377234822 0.147032959
    std. errors: 0.025 0.039 0.029


    fixed pat wt dof chisq tail prob
    000 0 8 10.649 0.22242 0.476 0.377 0.147
    001 1 9 33.484 0.000109978 0.474 0.526 -0.000
    010 1 9 92.824 4.40901e-16 0.663 0.000 0.337
    100 1 9 244.022 0 -0.000 1.162 -0.162 infeasible
    011 2 10 259.143 0 1.000 0.000 0.000
    101 2 10 252.982 0 0.000 1.000 -0.000
    110 2 10 557.430 0 0.000 0.000 1.000
    best pat: 000 0.22242 - -
    best pat: 001 0.000109978 chi(nested): 22.835 p-value for nested model: 1.76511e-06
    best pat: 011 0 chi(nested): 225.659 p-value for nested model: 5.27326e-51
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

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

     Bas (02-12-2019),  Camulogène Rix (02-11-2019),  Eterne (02-11-2019),  ffoucart (02-11-2019),  Helgenes50 (02-11-2019),  Megalophias (02-11-2019),  Onur Dincer (02-13-2019),  Ryukendo (02-11-2019),  traject (02-11-2019)

  3. #1132
    Registered Users
    Posts
    1,145
    Sex
    Omitted

    Maybe give Botai a shot in the left? Or West_Siberia_N? Instead of AG3.
    Collection of 14,000 d-stats: Hidden Content Part 2: Hidden Content Part 3: Hidden Content PM me for d-stats, qpadm, or f3-outgroup nmonte models. Looking for: KEB-IAM-TOR and Baikal_EN in plink/eigenstrat 1240k panel

  4. The Following 2 Users Say Thank You to Kale For This Useful Post:

     ffoucart (02-11-2019),  Ryukendo (02-11-2019)

  5. #1133
    Registered Users
    Posts
    280
    Sex

    anglesqueville, any information on what's driving the worst fit in the models? E.g. too basal Eurasian / not enough in relative sharing with CHG, etc? Anything that could form a basis on what looks like it may not be quite right.

    generalissimo's qpAdm fit of Vonyuchka Eneolithic is here - https://drive.google.com/file/d/1ZgE...HRvEoVb8U/view

    Best fitting model seems more CHG than you have found here for the progress pair - 58.7% vs 47.6% (which would average out 51%, and so closer to the paper's estimate of 53% EHG vs Yamnaya 49% EHG and 1% WHG). p-value looks higher, but different outgroups and SNPs - 182257.

    That one shows worst f4 statistics which show mainly Ganj_Dareh_N, West_Siberia_N, Kostenki14, MA-1 is closer to real sample than the fit relative to non-West Eurasian outgroups. If I read correctly.

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

     ffoucart (02-11-2019),  Megalophias (02-11-2019),  Ryukendo (02-11-2019)

  7. #1134
    Registered Users
    Posts
    816
    Sex

    Quote Originally Posted by Eterne View Post
    anglesqueville, any information on what's driving the worst fit in the models? E.g. too basal Eurasian / not enough in relative sharing with CHG, etc? Anything that could form a basis on what looks like it may not be quite right.

    generalissimo's qpAdm fit of Vonyuchka Eneolithic is here - https://drive.google.com/file/d/1ZgE...HRvEoVb8U/view

    Best fitting model seems more CHG than you have found here for the progress pair - 58.7% vs 47.6% (which would average out 51%, and so closer to the paper's estimate of 53% EHG vs Yamnaya 49% EHG and 1% WHG). p-value looks higher, but different outgroups and SNPs - 182257.

    That one shows worst f4 statistics which show mainly Ganj_Dareh_N, West_Siberia_N, Kostenki14, MA-1 is closer to real sample than the fit relative to non-West Eurasian outgroups. If I read correctly.
    Quote Originally Posted by Kale View Post
    Maybe give Botai a shot in the left? Or West_Siberia_N? Instead of AG3.
    Seconding these
    Quoted from this Forum:

    "Which superman haplogroup is the toughest - R1a or R1b? And which SNP mutation spoke Indo-European first? There's only one way for us to find out ... fight!"

  8. The Following User Says Thank You to Ryukendo For This Useful Post:

     ffoucart (02-11-2019)

  9. #1135
    Moderator
    Posts
    5,212
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA
    U152>L2>Z367
    mtDNA
    H5a1

    Normandie Netherlands Friesland Finland Orkney
    Ruykendo, Eterne: with my pops files West_Siberia_N and Botai don't give better adjustments than AG3. My perhaps naive or simplistic interpretation is that the constitution of Prog2 makes intervene a population rich in mesolithic-Caucasus and ANE. Compare the model published for Prog2, with this one (where Steppe_Eneo is for KhvalynskII):
    left pops:
    Steppe_Eneolithic
    CHG_KK1
    EHG_Samara
    AfontovaGora3

    right pops:
    Mota
    Ust_Ishim
    Kostenki14
    MA1
    WHG
    Iron_Gates_HG
    Iran_N
    Barcin
    LBK_Austria
    West_Siberia_N
    EHG


    numsnps used: 73491


    best coefficients: 0.116 0.818 0.065
    Jackknife mean: 0.116668037 0.817168272 0.066163690
    std. errors: 0.032 0.050 0.041

    fixed pat wt dof chisq tail prob
    000 0 8 7.618 0.471651 0.116 0.818 0.065

    Hopefully, when Generalissimo will launch his own analyses we'll get more solid models, with more SNPs. These ones were only quick trials.
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  10. The Following 8 Users Say Thank You to anglesqueville For This Useful Post:

     Agamemnon (02-14-2019),  Eterne (02-11-2019),  ffoucart (02-11-2019),  Megalophias (02-11-2019),  Onur Dincer (02-13-2019),  Psynome (02-12-2019),  Tomenable (02-13-2019),  traject (02-11-2019)

  11. #1136
    Registered Users
    Posts
    339
    Sex
    Ethnicity
    Celto-Germanic, Near East
    Nationality
    Murcan
    Y-DNA
    R1b-U152 (Alsace)
    mtDNA
    T2a1a

    United States of America Palestine Germany Ireland
    So..... are we talking about a genuinely "Native American" type strain of ancestry mediating Progress2's excess ANE admixture???

  12. #1137
    Registered Users
    Posts
    280
    Sex

    @K33, doesn't look like it, I don't think. Compare fits of Steppe_Maykop, when SM are fit as only Eneolithic_Piedmont_Steppe and AG3 in paper (without Kennewick), where they get strong Z=-4.679 for D(Steppe_Maykop, fitted_Steppe_Maykop, Mbuti, Han) and Z=-4.032 for D(Steppe_Maykop, fitted_Steppe_Maykop, Mbuti, Karitiana).

    Despite Steppe_Maykop:SA6001 in model from last page probably about 61% Progress, 34% AG3, 5% Kennewick, which would go to about 61% Progress, 35.7% ANE, 3.3% East Asian* (following typical sort of 33:66 ANE:ENA models for Native Americans). (And paper model for the Steppe_Maykop group is similar).

    Point not being the level of East Asian ancestry, but that there seems to be a fairly clear signal in Z with even that level, and can't see that here? Fits for the Progress pair and Vonyuchka so far don't look to have any problem of East Asian relatedness falling short. But I'm not totally sure about any of this.

    *Scaling up to about 9% ENA and 91% AG3, roughly similar to models for West_Siberia_N, although Narasimhan preprint also found that they fit best with some EHG ancestry rather than just AG3 if I remember correctly.

  13. #1138
    Registered Users
    Posts
    816
    Sex

    OK, some are gonna find this heretical for sure, how about a shot of Ganj_Dareh_N? Also, any way to update on the Gendstats regarding poor f4-stats that are driving the divergence?

    That's judging from the gendstats for the model in the paper, where there is insufficient closeness to Iran_N and Siberia_N and too much closeness to Native Americans (I'm guessing the high EHG is what does it?), which I expect the ANE will worsen?
    Quoted from this Forum:

    "Which superman haplogroup is the toughest - R1a or R1b? And which SNP mutation spoke Indo-European first? There's only one way for us to find out ... fight!"

  14. The Following User Says Thank You to Ryukendo For This Useful Post:

     ffoucart (02-12-2019)

  15. #1139
    Registered Users
    Posts
    2,939
    Sex

    I have all the high coverage samples, done on the same conversion pipeline as in the paper. Testing them now.

  16. The Following 21 Users Say Thank You to Generalissimo For This Useful Post:

     Просигој (02-12-2019),  Agamemnon (02-14-2019),  anglesqueville (02-12-2019),  Camulogène Rix (02-12-2019),  Eterne (02-12-2019),  fabrice E (02-13-2019),  ffoucart (02-12-2019),  J Man (02-12-2019),  JMcB (02-14-2019),  K33 (02-12-2019),  Kale (02-12-2019),  Michalis Moriopoulos (02-12-2019),  Onur Dincer (02-13-2019),  parasar (02-12-2019),  ph2ter (02-12-2019),  Psynome (02-12-2019),  rozenfeld (02-12-2019),  Ruderico (02-12-2019),  Ryukendo (02-12-2019),  Tomenable (02-13-2019),  traject (02-12-2019)

  17. #1140
    Registered Users
    Posts
    2,939
    Sex

    These are looking really good.

    Let me just organize a couple more and then I'll share.

  18. The Following 17 Users Say Thank You to Generalissimo For This Useful Post:

     Просигој (02-13-2019),  Agamemnon (02-14-2019),  anglesqueville (02-13-2019),  Eterne (02-13-2019),  ffoucart (02-13-2019),  J Man (02-13-2019),  Jessie (02-14-2019),  JMcB (02-14-2019),  Kale (02-13-2019),  Michalis Moriopoulos (02-13-2019),  MitchellSince1893 (02-13-2019),  Onur Dincer (02-13-2019),  Psynome (02-13-2019),  R.Rocca (02-13-2019),  Ruderico (02-13-2019),  Ryukendo (02-13-2019),  Tomenable (02-13-2019)

Page 114 of 120 FirstFirst ... 1464104112113114115116 ... LastLast

Similar Threads

  1. Which has greater autosomal genetic diversity: Europe or China?
    By Tomenable in forum Autosomal (auDNA)
    Replies: 7
    Last Post: 07-19-2018, 03:17 AM
  2. Replies: 14
    Last Post: 01-28-2018, 02:39 PM
  3. Replies: 10
    Last Post: 01-22-2018, 03:10 PM
  4. Racial Purity,Harvard’s Eugenics Era
    By rock hunter in forum History (Modern)
    Replies: 0
    Last Post: 05-05-2016, 08:36 AM
  5. Replies: 55
    Last Post: 02-20-2016, 10:52 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
  •