Page 29 of 29 FirstFirst ... 19272829
Results 281 to 290 of 290

Thread: Paleolithic DNA from the Caucasus reveals core of West Eurasian ancestry

  1. #281
    Registered Users
    Posts
    1,193
    Sex
    Omitted

    Quote Originally Posted by Psynome View Post
    Where do you get Taforalt as ~60% pre-OOA? Do you include Basal Eurasian as "pre-OOA"?
    How basal Taforalt is, if you wanted to model Taforalt as say 'something without basal' + Mota, with your only outgroup being Ust_Ishim, you'd need 39.7% Mota.
    We know Mota is not a proximate source though, it harbors ancestry from a more divergent lineage than dampens its affinity to some Africans, whereas Taforalt does not.
    Therefore we need a ghost that is, overall, closer to Eurasians, which means a higher % contribution (>39.7%) needed from that source to create the basal effect.
    I estimate this source based on the trajectory from Anatolia_N > Levant_N, which involves 10.5% of ancestry from a population that would be F3: MbutiDG popX Ust_Ishim = 0.1540
    (In place of popX, Mota = 0.1007, Taforalt = 0.1787, Barcin_N = 0.2202, Levant_N = 0.2120, non-basal-Eurasians = ~0.2300)
    The contribution from popX would be ~67.5%. 32.5% remains, which is probably how much West-Eurasian Taforalt has (24% at the absolute minimum).
    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

  2. The Following 5 Users Say Thank You to Kale For This Useful Post:

     blackflash16 (12-20-2018),  Hando (12-21-2018),  Megalophias (12-20-2018),  Psynome (12-20-2018),  Shamayim (12-25-2018)

  3. #282
    Registered Users
    Posts
    128

    Quote Originally Posted by Kale View Post
    How basal Taforalt is, if you wanted to model Taforalt as say 'something without basal' + Mota, with your only outgroup being Ust_Ishim, you'd need 39.7% Mota.
    We know Mota is not a proximate source though, it harbors ancestry from a more divergent lineage than dampens its affinity to some Africans, whereas Taforalt does not.
    Therefore we need a ghost that is, overall, closer to Eurasians, which means a higher % contribution (>39.7%) needed from that source to create the basal effect.
    I estimate this source based on the trajectory from Anatolia_N > Levant_N, which involves 10.5% of ancestry from a population that would be F3: MbutiDG popX Ust_Ishim = 0.1540
    (In place of popX, Mota = 0.1007, Taforalt = 0.1787, Barcin_N = 0.2202, Levant_N = 0.2120, non-basal-Eurasians = ~0.2300)
    The contribution from popX would be ~67.5%. 32.5% remains, which is probably how much West-Eurasian Taforalt has (24% at the absolute minimum).
    Thanks for describing your model. So by "Pre-OOA" you mean anything more basal than the common Eurasian ancestry found in Ust' Ishim. That would exclude Basal Eurasian.

    Your estimate of 32.5% West Eurasian lines up pretty well with the proportion given in the graph model in the paper, which has (.72)(.55)=39.6% West Eurasian.

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

     bmoney (12-22-2018),  Hando (12-21-2018)

  5. #283
    Registered Users
    Posts
    1,193
    Sex
    Omitted

    Honestly that's coincidental. The qpgraph models in that paper leave out crucial populations and as such are basically worthless.
    At this point I'd advocate that Taforalt is ~32.5% paleo-european-related + 67.5% of something that was probably the last stop before the main Eurasian/OOA bottleneck.
    The West Eurasian stuff in Taforalt is definitely different than that in Natufians.

    MbutiDG Natufian Vestonice16 Kostenki14 -0.0062 -0.911 15672 15868 309785
    MbutiDG Taforalt Vestonice16 Kostenki14 -0.0186 -3.851 32419 33646 678742
    MbutiDG Natufian GoyetQ116-1 Kostenki14 -0.0098 -1.364 16723 17055 325656
    MbutiDG Taforalt GoyetQ116-1 Kostenki14 -0.015 -3.175 34983 36051 707696
    MbutiDG Natufian ElMiron GoyetQ116-1 -0.0173 -2.291 12140 12567 255688
    MbutiDG Taforalt ElMiron GoyetQ116-1 -0.0012 -0.231 23931 23989 515690
    MbutiDG Natufian Bichon Vestonice16 -0.0217 -3.181 16518 17250 327386
    MbutiDG Taforalt Bichon Vestonice16 0.0005 0.099 35780 35747 726247

    If you still want the flow to be Natufian > Taforalt though, adding Oase1 as a third contributor to Taforalt for some weird reason works.

    I'm hoping Dzudzuana helps to resolve some of this back-and-forth modeling...

    Taforalt = Natufian + Oase1 + African
    Natufian = Anatolia_N + paleo-european + Taforalt

    Anatolia_N = Levant_N + CHG + WHG
    Levant_N = Anatolia_N + Taforalt
    WHG = Vestonice + ElMiron + EHG + Anatolia_N
    Last edited by Kale; 12-25-2018 at 03:27 AM.
    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

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

     bmoney (12-27-2018),  epoch (12-26-2018),  Hando (12-25-2018),  Psynome (12-25-2018)

  7. #284
    Registered Users
    Posts
    128

    Quote Originally Posted by Kale View Post
    Honestly that's coincidental. The qpgraph models in that paper leave out crucial populations and as such are basically worthless.
    At this point I'd advocate that Taforalt is ~32.5% paleo-european-related + 67.5% of something that was probably the last stop before the main Eurasian/OOA bottleneck.
    The West Eurasian stuff in Taforalt is definitely different than that in Natufians.

    MbutiDG Natufian Vestonice16 Kostenki14 -0.0062 -0.911 15672 15868 309785
    MbutiDG Taforalt Vestonice16 Kostenki14 -0.0186 -3.851 32419 33646 678742
    MbutiDG Natufian GoyetQ116-1 Kostenki14 -0.0098 -1.364 16723 17055 325656
    MbutiDG Taforalt GoyetQ116-1 Kostenki14 -0.015 -3.175 34983 36051 707696
    MbutiDG Natufian ElMiron GoyetQ116-1 -0.0173 -2.291 12140 12567 255688
    MbutiDG Taforalt ElMiron GoyetQ116-1 -0.0012 -0.231 23931 23989 515690
    MbutiDG Natufian Bichon Vestonice16 -0.0217 -3.181 16518 17250 327386
    MbutiDG Taforalt Bichon Vestonice16 0.0005 0.099 35780 35747 726247

    If you still want the flow to be Natufian > Taforalt though, adding Oase1 as a third contributor to Taforalt for some weird reason works.

    I'm hoping Dzudzuana helps to resolve some of this back-and-forth modeling...

    Taforalt = Natufian + Oase1 + African
    Natufian = Anatolia_N + paleo-european + Taforalt

    Anatolia_N = Levant_N + CHG + WHG
    Levant_N = Anatolia_N + Taforalt
    WHG = Vestonice + ElMiron + EHG + Anatolia_N
    Thanks for clearing that up. It will be nice to see what people come up with when the data is released!

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

     Hando (12-26-2018)

  9. #285
    Registered Users
    Posts
    1,193
    Sex
    Omitted

    Looking back, it's kind of interesting that instead of East Eurasian > Goyet to explain the weirdness with that sample, the authors try Goyet > East Eurasian.
    If that sort of thing is archaeologically plausible, is Basal Eurasian even really needed?
    Instead there could just be a split between proto-East and proto-West Eurasians, and the closer phylogenically a West-Eurasian is to that Goyet-like input into East-Eurasians, the less 'basal' they would appear to have.
    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

  10. The Following User Says Thank You to Kale For This Useful Post:

     Hando (02-02-2019)

  11. #286
    Registered Users
    Posts
    2,323
    Sex
    Y-DNA
    R1b-DF27 FGC17112
    mtDNA
    H5a1

    Goyet has the least Basal Eurasian. No flow with Asians. African outgroups don't work either. Got a post coming soon with an explanation.

    Edit:Natufians show no more relationship to Africans than Ganj_Dareh. Any flow from Torafalt to Natufians would not allow that. Flow is other direction. Torafalt is something deeper than Basal Eurasian as we know it, something like Natufians, and something more like post-Oase1 Europeans.
    Last edited by Chad Rohlfsen; 01-21-2019 at 01:17 AM.

  12. The Following 4 Users Say Thank You to Chad Rohlfsen For This Useful Post:

     Hando (02-02-2019),  K33 (01-21-2019),  Megalophias (01-21-2019),  Pribislav (01-21-2019)

  13. #287
    Registered Users
    Posts
    1,193
    Sex
    Omitted

    Quote Originally Posted by Chad Rohlfsen View Post
    Goyet has the least Basal Eurasian. No flow with Asians. African outgroups don't work either. Got a post coming soon with an explanation.
    What do you mean here by 'African outgroups don't work either'? Missing some context.

    Edit:Natufians show no more relationship to Africans than Ganj_Dareh. Any flow from Torafalt to Natufians would not allow that. Flow is other direction. Torafalt is something deeper than Basal Eurasian as we know it, something like Natufians, and something more like post-Oase1 Europeans.
    If you plot F3 outgroup values for Barcin_N on one axis, and Levant_N on the other axis, the R^2 correlation is a near perfect 0.98, and that is with every Eurasian population excluding Natufians.
    The simplest explanation is that Barcin_N and Levant_N are effectively the same thing, except Levant_N has a bit extra basal.
    The slope of that correlation is 0.89-0.90, meaning 10-11% extra basal ancestry. F3: Barcin_N Ust_Ishim MbutiDG = 0.2202. F3: Levant_N Ust_Ishim MbutiDG = 0.2120.
    If 10-11% basal knocks off 0.0082 Ust_Ishim affinity, then a population that is 100% basal would be (using East Eurasians as a 0% baseline at ~0.2320) F3: Basal Ust_Ishim MbutiDG = 0.1500 < > 0.1575
    This is in fact more basal than Taforalt at 0.1787
    Most likely explanation I see is a population in NE Africa splits one way into Near East, and another into NW Africa, mixing with paleo-Euro-like group to become Taforalt, then Taforalt-like group contributes to West Africa.

    IIRC Taforalt shows no more relationship to Africans than Iran_N as well, excluding West Africans (even then, the Z-score is only around 4)
    The current Taforalt samples we have are almost as bottlenecked as WHG, with F3 outgroup values (MbutiDG being outgroup) for (non-related) Taforalt individuals being ~0.330.
    The Taforalt-like input into Natufians would be less drifted, sharing around 0.285 with Taforalt.
    If Taforalt is a more proximate input into West Africans, and Natufians only have ~28% Taforalt-like input, we should only expect a Z-score of around 1 in stats with Natufians involving West Africans, so it's insignificant.

    Qpadm for Natufian, columns fully loaded (including Mota, MbutiDG, South_Africa_2000BP)
     

    parameter file: qpadmpar
    ### THE INPUT PARAMETERS
    ##PARAMETER NAME: VALUE
    genotypename: dataset16.geno
    snpname: dataset16.snp
    indivname: dataset16.ind
    popleft: qpadmleft
    popright: qpadmright
    details: YES
    ## qpAdm version: 200

    left pops:
    Natufian
    Mentese_N
    Vestonice16
    Taforalt

    right pops:
    Barcin_N
    Iran_N
    Kotias
    MA1
    ANA
    GoyetQ116-1
    Kostenki14
    Iron_Gates_HG
    Ukraine_Mesolithic
    Tianyuan
    Ust_Ishim
    Mota
    MbutiDG
    South_Africa_2000BP

    0 Natufian 1
    1 Mentese_N 4
    2 Vestonice16 1
    3 Taforalt 5
    4 Barcin_N 5
    5 Iran_N 7
    6 Kotias 1
    7 MA1 1
    8 ANA 2
    9 GoyetQ116-1 1
    10 Kostenki14 1
    11 Iron_Gates_HG 5
    12 Ukraine_Mesolithic 3
    13 Tianyuan 1
    14 Ust_Ishim 1
    15 Mota 1
    16 MbutiDG 3
    17 South_Africa_2000BP 2
    jackknife block size: 0.050
    snps: 1149844 indivs: 45
    number of blocks for block jackknife: 713
    dof (jackknife): 633.987
    numsnps used: 161193
    codimension 1
    f4info:
    f4rank: 2 dof: 11 chisq: 6.518 tail: 0.836652871 dofdiff: 13 chisqdiff: -6.518 taildiff: 1
    B:
    scale 1.000 1.000
    Iran_N 0.556 -0.072
    Kotias 0.440 0.052
    MA1 1.052 0.783
    ANA 1.005 0.325
    GoyetQ116-1 1.078 0.995
    Kostenki14 1.173 1.276
    Iron_Gates_HG 0.825 1.057
    Ukraine_Mesolithic 0.791 0.802
    Tianyuan 1.042 0.116
    Ust_Ishim 1.120 0.264
    Mota 1.150 -1.624
    MbutiDG 1.216 -1.688
    South_Africa_2000BP 1.181 -1.525
    A:
    scale 161.766 365.905
    Mentese_N -0.718 0.287
    Vestonice16 1.129 1.283
    Taforalt 1.100 -1.127


    full rank 1
    f4info:
    f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 11 chisqdiff: 6.518 taildiff: 0.836652871
    B:
    scale 1.000 1.000 1.000
    Iran_N 0.583 -0.045 1.102
    Kotias 0.471 0.070 1.085
    MA1 1.044 0.782 -0.361
    ANA 1.060 0.377 2.303
    GoyetQ116-1 1.050 0.977 -1.039
    Kostenki14 1.159 1.287 -0.637
    Iron_Gates_HG 0.828 1.074 0.165
    Ukraine_Mesolithic 0.809 0.830 0.648
    Tianyuan 1.048 0.128 0.130
    Ust_Ishim 1.078 0.222 -1.491
    Mota 1.160 -1.598 0.478
    MbutiDG 1.198 -1.692 -0.879
    South_Africa_2000BP 1.182 -1.519 -0.065
    A:
    scale 160.028 366.859 3544.408
    Mentese_N -0.667 0.279 1.574
    Vestonice16 1.142 1.276 0.257
    Taforalt 1.119 -1.137 0.675


    best coefficients: 0.607 0.112 0.282
    ssres:
    0.000503359 0.000468682 0.000108377 0.001003882 -0.000118910 0.000032932 0.000225368 0.000386291 0.000287017 -0.000244854 0.000467427 0.000029206 0.000291099
    1.236533707 1.151346968 0.266234704 2.466100356 -0.292111081 0.080899720 0.553630418 0.948949388 0.705076426 -0.601498590 1.148265586 0.071745239 0.715103406

    Jackknife mean: 0.606743547 0.111718320 0.281538133
    std. errors: 0.030 0.031 0.027

    error covariance (* 1000000)
    928 -586 -342
    -586 982 -397
    -342 -397 739


    fixed pat wt dof chisq tail prob
    000 0 11 6.518 0.836653 0.607 0.112 0.282
    001 1 12 108.504 1.18239e-17 0.728 0.272 0.000
    010 1 12 18.888 0.0912682 0.675 0.000 0.325
    100 1 12 354.729 0 0.000 0.519 0.481
    011 2 13 174.263 2.50624e-30 1.000 0.000 0.000
    101 2 13 472.645 0 0.000 1.000 0.000
    110 2 13 530.741 0 0.000 0.000 1.000
    best pat: 000 0.836653 - -
    best pat: 010 0.0912682 chi(nested): 12.370 p-value for nested model: 0.00043634
    best pat: 011 2.50624e-30 chi(nested): 155.375 p-value for nested model: 1.15953e-35

    ## end of run
    Last edited by Kale; 01-21-2019 at 03:22 AM.
    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

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

     Hando (02-02-2019),  Pribislav (01-21-2019),  Ryukendo (01-21-2019)

  15. #288
    Registered Users
    Posts
    2,323
    Sex
    Y-DNA
    R1b-DF27 FGC17112
    mtDNA
    H5a1

    Completely incorrect and poor modeling. Why would Natufians be more like Vestonice than Anatolians? Anatolians are part Natufian too. Makes zero sense. Again, Africans. You're missing the issue. You're just gonna have to wait.

  16. The Following 2 Users Say Thank You to Chad Rohlfsen For This Useful Post:

     Hando (02-02-2019),  pegasus (01-21-2019)

  17. #289
    Registered Users
    Posts
    1,193
    Sex
    Omitted

    Quote Originally Posted by Chad Rohlfsen View Post
    Completely incorrect and poor modeling.
    I don't see any statistics in that statement.
    Why would Natufians be more like Vestonice than Anatolians?
    No idea, Levantine Aurignacian probably?
    Anatolians are part Natufian too.
    Yes, or at least Natufian-like. But that's not really a problem for such a model.
    Makes zero sense. Again, Africans. You're missing the issue.
    Tangibles please.
    You're just gonna have to wait.
    For what?
    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

  18. The Following 5 Users Say Thank You to Kale For This Useful Post:

     Eterne (01-21-2019),  Hando (02-02-2019),  Megalophias (01-21-2019),  pgbk87 (01-21-2019),  traject (01-21-2019)

  19. #290
    Registered Users
    Posts
    2,323
    Sex
    Y-DNA
    R1b-DF27 FGC17112
    mtDNA
    H5a1

    I can't go into the full details. It's a lot to type up. My upcoming blog post will go into it.

  20. The Following User Says Thank You to Chad Rohlfsen For This Useful Post:

     Hando (02-02-2019)

Page 29 of 29 FirstFirst ... 19272829

Similar Threads

  1. How much ancient West Eurasian ancestry do Amerindians have?
    By Tsakhur in forum Autosomal (auDNA)
    Replies: 4
    Last Post: 09-18-2017, 12:06 PM
  2. Replies: 0
    Last Post: 03-14-2016, 05:53 PM
  3. Replies: 18
    Last Post: 07-20-2015, 01:49 AM
  4. Replies: 52
    Last Post: 02-18-2015, 10:14 AM

Posting Permissions

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