Page 9 of 15 FirstFirst ... 7891011 ... LastLast
Results 81 to 90 of 147

Thread: 1st Look at Damgaard et al 2018 C Asian Samples

  1. #81
    Registered Users
    Posts
    4,870
    Sex

    Quote Originally Posted by khanabadoshi View Post
    When you get a chance can you try with SiS-BA1 for myself and the khandan? Also do you have Yamnaya_Afanasevo? In previous dstat runs by Generalissimo, they seemed the most significant for myself and co. Jam's Z-Score for Ottoman is definitely interesting.

    These are the run outputted via poi's nMonte algorithm and SiS-BA1+3 seems to be what does it. The fit is suspiciously good... too good. f3 might be able to validate/invalidate the method.
    ]
    No worrries, I'll try BA1 as well. I sitll need to do BA3 I believe for Reza. I'll also validate the f3s later by doing IBS Steppe-modern, using only signature alleles (steppe individual groupwise allele agreement) where the various steppe groups approach 100% MAF
    EurasianDNA.com - A study of the population history of West & South Asia.

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

     bmoney (05-17-2018), khanabadoshi (05-16-2018), kingjohn (05-16-2018), Varun R (05-16-2018)

  3. #82
    Registered Users
    Posts
    269
    Sex
    Ethnicity
    Celto-Germanic, Near East
    Nationality
    Murcan
    Y-DNA
    R1b-U152 (Alsatian)
    mtDNA
    T2a1a

    United States of America Palestine Germany Ireland
    Quote Originally Posted by Kurd View Post
    I have started looking at S Asians using Shahr-e-Sukhta BA2 as source1. S and E Indians are not showing much in the way of signals, but by contrast 2 NW S Asian members Dluffy and Sapporo are showing a ton. Almost every single row of the 75 row table is showing a signal of admixture. Not only for the Saka layer but also for the older Sintashta and Srubna layer !

    Sapporo tends to be a little more C Saka shifted as well as Srubna shifted, Dluffy a little more E Saka shifted and less Srubna shifted. Will post their tables tomorrow.
    That probably explains why Duffy was getting some of the worst fits among Punjabis in the Parsimonious nMonte Modeling thread. His Iron Age/Saka Steppe must have been too high to have been approximated by Steppe_MLBA, even in nMonte.



    @Kurd: Have you tried Uttar Pradesh Indians in these f3 stats yet? They were getting really, really crappy fits (over 4% even at pen=0) using the Iran_N/Chl, Steppe_MLBA_East, Shar_I_Sokhta_BA3, Onge/Kusunda/Mongola 6-way nMonte model I was using.

    I'm thinking maybe Uttar Pradesh might have some of the highest Saka/Scythian levels on the subcontinent...

  4. The Following 3 Users Say Thank You to K33 For This Useful Post:

     bmoney (05-17-2018), Kurd (05-16-2018), Observer (05-17-2018)

  5. #83
    Registered Users
    Posts
    4,870
    Sex

    Quote Originally Posted by K33 View Post
    That probably explains why Duffy was getting some of the worst fits among Punjabis in the Parsimonious nMonte Modeling thread. His Iron Age/Saka Steppe must have been too high to have been approximated by Steppe_MLBA, even in nMonte.



    @Kurd: Have you tried Uttar Pradesh Indians in these f3 stats yet? They were getting really, really crappy fits (over 4% even at pen=0) using the Iran_N/Chl, Steppe_MLBA_East, Shar_I_Sokhta_BA3, Onge/Kusunda/Mongola 6-way nMonte model I was using.

    I'm thinking maybe Uttar Pradesh might have some of the highest Saka/Scythian levels on the subcontinent...
    I can try UP if they are part of the Damgaard set. Yes, I agree about Dluffy. For NW Indians, AASI alone clearly cant be used to model all their E Eurasian. Han of course never occupied C or S Asia, Saka did, so they are the logical source. The Saka inferred E Eurasian was inferred IA and later by descendants of Saka
    EurasianDNA.com - A study of the population history of West & South Asia.

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

     bmoney (05-17-2018), kingjohn (05-16-2018), Observer (05-17-2018), Varun R (05-16-2018)

  7. #84
    Registered Users
    Posts
    1,446
    Sex
    Location
    Canada
    Ethnicity
    Punjabi
    Nationality
    Canadian
    Y-DNA
    R-M417
    mtDNA
    M3a2

    India Canada Azad Baluchistan Sikh Empire India Punjab Nishan Sahib
    Quote Originally Posted by K33 View Post
    That probably explains why Duffy was getting some of the worst fits among Punjabis in the Parsimonious nMonte Modeling thread. His Iron Age/Saka Steppe must have been too high to have been approximated by Steppe_MLBA, even in nMonte.



    @Kurd: Have you tried Uttar Pradesh Indians in these f3 stats yet? They were getting really, really crappy fits (over 4% even at pen=0) using the Iran_N/Chl, Steppe_MLBA_East, Shar_I_Sokhta_BA3, Onge/Kusunda/Mongola 6-way nMonte model I was using.

    I'm thinking maybe Uttar Pradesh might have some of the highest Saka/Scythian levels on the subcontinent...
    Does that mean my steppe was underestimated by nmonte hence crap fits?
    Deg Teg Fateh - Victory to Charity and Arms

  8. #85
    Registered Users
    Posts
    4,870
    Sex

    Quote Originally Posted by MonkeyDLuffy View Post
    Does that mean my steppe was underestimated by nmonte hence crap fits?

    Probably because they didn’t use Saka samples to model you


    Edit: The f3s show that Saka are better than MLBA for modelling the Steppe portion of ancestry of SC Asians as well as many S Asians, because they better account for the non-AASI E Eurasian in SC/S Asians than MLBA.

    This can clearly be seen when using the combination of an AASI rich sample like Shahr-i-Sukhteh BA3 in combination with Saka to model S Asians vs SIS BA3-MLBA


    Remember that S Asians have alot of E Eurasian, and not all the E Eurasian is AASI in NW S Asians or Brahmins
    Last edited by Kurd; 05-16-2018 at 08:49 PM.
    EurasianDNA.com - A study of the population history of West & South Asia.

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

     bmoney (05-17-2018), Sapporo (05-17-2018)

  10. #86
    Registered Users
    Posts
    4,870
    Sex

    Another thing the f3 models:

    1- Shahr-i-Sukhteh BA3 - Saka

    vs

    2- Shahr-i-Sukhteh BA3 - MLBA

    show for S Asian based on 1 above yielding better signals of admixture (fits) is Saka has something that MLBA doesn’t which makes it a better model. It’s not AASI, because SIS BA3 has plenty, it’s Non-AASI E Eurasian which both SIS BA3 and MLBA lack
    EurasianDNA.com - A study of the population history of West & South Asia.

  11. The Following 4 Users Say Thank You to Kurd For This Useful Post:

     bmoney (05-17-2018), khanabadoshi (05-16-2018), Observer (05-17-2018), Varun R (05-16-2018)

  12. #87
    Registered Users
    Posts
    269
    Sex
    Ethnicity
    Celto-Germanic, Near East
    Nationality
    Murcan
    Y-DNA
    R1b-U152 (Alsatian)
    mtDNA
    T2a1a

    United States of America Palestine Germany Ireland
    Quote Originally Posted by Kurd View Post
    Probably because they didn’t use Saka samples to model you


    Edit: The f3s show that Saka are better than MLBA for modelling the Steppe portion of ancestry of SC Asians as well as many S Asians, because they better account for the non-AASI E Eurasian in SC/S Asians than MLBA.
    Is this true only for SintashtaMLBA, or did you try Zevakinskiy_LBA as well?

    Sintashta was only ~10% WSHG-related, Zevakinskiy was ~40% WSHG-related. But I agree it does look like WSHG has some kind of deep affinity with AASI anyway; the East Eurasian in the Saka samples is maybe more Baikal_HG-related or something?

  13. The Following User Says Thank You to K33 For This Useful Post:

     Observer (05-17-2018)

  14. #88
    Registered Users
    Posts
    4,870
    Sex

    Quote Originally Posted by K33 View Post
    Is this true only for SintashtaMLBA, or did you try Zevakinskiy_LBA as well?

    Sintashta was only ~10% WSHG-related, Zevakinskiy was ~40% WSHG-related. But I agree it does look like WSHG has some kind of deep affinity with AASI anyway; the East Eurasian in the Saka samples is maybe more Baikal_HG-related or something?
    If you look at the PCA I posted in the opening post youll notice Saka/Scythians are on a cline from ENF to Pazyryk and Kazakh and Okunevo, so it looks like E Eurasian of the E/NE Asian type
    EurasianDNA.com - A study of the population history of West & South Asia.

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

     bmoney (05-17-2018), K33 (05-16-2018), Observer (05-17-2018)

  16. #89
    Registered Users
    Posts
    1,446
    Sex
    Location
    Canada
    Ethnicity
    Punjabi
    Nationality
    Canadian
    Y-DNA
    R-M417
    mtDNA
    M3a2

    India Canada Azad Baluchistan Sikh Empire India Punjab Nishan Sahib
    Quote Originally Posted by Kurd View Post
    If you look at the PCA I posted in the opening post youll notice Saka/Scythians are on a cline from ENF to Pazyryk and Kazakh and Okunevo, so it looks like E Eurasian of the E/NE Asian type
    Okay, that helped big time, I ran my nmonte using Saka central and magically my fit is amazing, like its so close.

    [1] "distance%=2.2485 / distance=0.022485"


    Punjabi_Ramgarhia
    "Shahr_I_Sokhta_BA3" 47.65
    "Saka_central" 28.25
    "Shahr_I_Sokhta_BA2" 24.1
    "Yakut" 0


    Also ran Sapporo, who got way better fit as well using saka central:

    [1] "distance%=1.5728 / distance=0.015728"


    Sapporo
    "Shahr_I_Sokhta_BA2" 45.5
    "Shahr_I_Sokhta_BA3" 28.7
    "Saka_central" 25.8
    "Yakut" 0
    Last edited by MonkeyDLuffy; 05-17-2018 at 12:12 AM.
    Deg Teg Fateh - Victory to Charity and Arms

  17. The Following 4 Users Say Thank You to MonkeyDLuffy For This Useful Post:

     jortita (05-17-2018), K33 (05-21-2018), Kurd (05-16-2018), Observer (05-17-2018)

  18. #90
    Registered Users
    Posts
    4,870
    Sex

    Quote Originally Posted by MonkeyDLuffy View Post
    Okay, that helped big time, I ran my nmonte using Saka central and magically my fit is amazing, like its so close.

    [1] "distance%=2.2485 / distance=0.022485"


    Punjabi_Ramgarhia
    "Shahr_I_Sokhta_BA3" 47.65
    "Saka_central" 28.25
    "Shahr_I_Sokhta_BA2" 24.1
    "Yakut" 0


    Also ran Sapporo, who got way better fit as well using saka central:

    [1] "distance%=1.5728 / distance=0.015728"


    Sapporo
    "Shahr_I_Sokhta_BA2" 45.5
    "Shahr_I_Sokhta_BA3" 28.7
    "Saka_central" 25.8
    "Yakut" 0
    There you go... that’s your non-AASI E Eurasian and circumstantial EVIDENCE that Saka are important to the population history of SC/S Asia (along with SIS BA3)...Anyone who thinks they can ACCURATELY model S Asians such as you and Sapporo by simply using only AASI , MLBA , Iranian farmer related is living in a bubble and doesn’t understand SC and S Asian demography as your previous poor fits showed
    EurasianDNA.com - A study of the population history of West & South Asia.

  19. The Following 4 Users Say Thank You to Kurd For This Useful Post:

     bmoney (05-17-2018), jortita (05-17-2018), khanabadoshi (05-17-2018), MonkeyDLuffy (05-17-2018)

Page 9 of 15 FirstFirst ... 7891011 ... LastLast

Similar Threads

  1. Replies: 45
    Last Post: 04-29-2018, 04:21 AM
  2. Replies: 3
    Last Post: 02-23-2018, 06:36 AM
  3. Notes on South Asian genetics, 2018
    By firemonkey in forum Autosomal (auDNA)
    Replies: 57
    Last Post: 01-27-2018, 03:44 PM
  4. Replies: 0
    Last Post: 07-11-2017, 09:47 PM
  5. Replies: 0
    Last Post: 07-11-2017, 09:41 PM

Tags for this Thread

Posting Permissions

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