Results 1 to 10 of 10

Thread: Who are the closest modern descendants of the various steppe groups, based on formal

  1. #1
    Banned
    Posts
    2,951
    Sex
    Location
    America
    Ethnicity
    Telgu/marathi

    Siberian Tatars Tajikistan

    Who are the closest modern descendants of the various steppe groups, based on formal

    Yamnaya, Corded Ware/Sintashta, EHG, West Siberian HG?

    Corded Ware/Sintashta I’m guessing it’s Eastern Europeans though I’m not entirely sure because due to drift eastern euros appear more distant to them than Scandos based on G25.

    The rest I have no idea. I’d guess it to be Uralics but they also appear quite diverged on nmonte.

    Any relevant data?

  2. The Following User Says Thank You to Censored For This Useful Post:

     Hando (02-15-2020)

  3. #2
    Registered Users
    Posts
    4,612
    Sex

    Formal stats are crap for this sort of thing. What you want are Fst genetic distances.

    https://eurogenes.blogspot.com/2018/...-distance.html

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

     agent_lime (02-14-2020),  Dorkymon (02-14-2020),  Hando (02-15-2020),  JonikW (02-14-2020),  michal3141 (02-14-2020),  Mis (02-14-2020)

  5. #3
    Banned
    Posts
    2,951
    Sex
    Location
    America
    Ethnicity
    Telgu/marathi

    Siberian Tatars Tajikistan
    Quote Originally Posted by Generalissimo View Post
    Formal stats are crap for this sort of thing. What you want are Fst genetic distances.

    https://eurogenes.blogspot.com/2018/...-distance.html
    Strongly appreciate it.

    For Sintashta, it appears Russian central are the closest at a FST of only .01 while the rest of NW/N/NE Euros like Irish, Germans, Belarusian go from .011-.013.

    West Siberian_N: Lowest I found was for Komi at .055
    EHG: Komi also, something like .44
    Yamna: Russian central at .17

    Central Russian-CWC
    CWC Baltic-.003
    CWC Czech- .006




    It seems Central Russians are the closest descendants of these MLBA steppe groups. For two populations to be considered "identical", what is the minimum FST?

  6. The Following 2 Users Say Thank You to Censored For This Useful Post:

     Hando (02-15-2020),  michal3141 (02-14-2020)

  7. #4
    Registered Users
    Posts
    928
    Sex
    Location
    Krakow
    Ethnicity
    Polish
    Nationality
    Pole
    Y-DNA (P)
    R1a > M458 > Y23110
    mtDNA (M)
    J1c1

    Poland
    Quote Originally Posted by Generalissimo View Post
    Formal stats are crap for this sort of thing. What you want are Fst genetic distances.

    https://eurogenes.blogspot.com/2018/...-distance.html
    Dear Generalissimo,

    That's very interesting data.

    Could you generate something like this for your current dataset (if it doesn't take too much time, resources and hassle)?

    EDIT: I had a look at the FST distance matrix you posted. Not sure why there are some negative distances. For example, FST distances to Polish (top 50) and there is -0.001 distance between Polish and Russian_Central :

    Code:
    Distances to Polish:
    Russian_Central           -0.001
    Ukrainian_West             0.000
    Polish                     0.000
    Ukrainian_East             0.000
    Belarusian                 0.001
    Slovenian                  0.001
    Lithuanian                 0.001
    Slovakian                  0.001
    Russian_West               0.001
    German                     0.001
    Serbian                    0.002
    Latvian                    0.002
    Estonian                   0.002
    Croatian                   0.002
    Bulgarian                  0.003
    Norwegian                  0.003
    Swedish                    0.003
    Romanian                   0.003
    English                    0.003
    French                     0.003
    CWC_Baltic                 0.004
    England_Anglo-Saxon        0.004
    England_Roman              0.004
    Irish                      0.004
    Scottish                   0.004
    Montenegrin                0.004
    Mordovian                  0.004
    Tatar_Kryashen             0.004
    England_IA                 0.005
    Spanish                    0.005
    Russian_North              0.005
    Finnish                    0.006
    Tatar                      0.006
    Ingrian                    0.006
    Italian_Bergamo            0.006
    Italian_Tuscan             0.006
    Greek                      0.007
    Italian_EastSicilian       0.008
    Karelian                   0.008
    Beaker_Southern_France     0.008
    Komi                       0.008
    Vepsa                      0.008
    Italian_Abruzzo            0.008
    Poland_BA                  0.008
    Unetice_Germany            0.008
    Lisakovskiy_MLBA_Alakul    0.009
    Scotland_MBA               0.009
    Albanian                   0.009
    Italian_WestSicilian       0.009
    Ukraine_Eneolithic         0.010
    Last edited by michal3141; 02-14-2020 at 11:23 PM.

  8. The Following User Says Thank You to michal3141 For This Useful Post:

     Hando (02-15-2020)

  9. #5
    Registered Users
    Posts
    66
    Sex
    Y-DNA (P)
    J2a - L70
    mtDNA (M)
    N1a1a1a2

    Quote Originally Posted by Generalissimo View Post
    Formal stats are crap for this sort of thing. What you want are Fst genetic distances.

    https://eurogenes.blogspot.com/2018/...-distance.html
    Is there a source or link one can follow to better understand this distinction? I.e Why/When to use formal stats instead of Fst distances, what can formal stats tell us that Fst cannot & vice versa?
    Last edited by Nebuchadnezzar II; 02-14-2020 at 11:12 PM.

  10. The Following 2 Users Say Thank You to Nebuchadnezzar II For This Useful Post:

     Hando (02-15-2020),  michal3141 (02-14-2020)

  11. #6
    Registered Users
    Posts
    4,612
    Sex

    Quote Originally Posted by Nebuchadnezzar II View Post
    Is there a source or link one can follow to better understand this distinction? I.e Why/When to use formal stats instead of Fst distances, what can formal stats tell us that Fst cannot & vice versa?
    No, this is all very new.

    But formal stats generally focus on deep ancestry, and they also have a property that causes samples with less complex ancestry from the extremes of mixture clines to be favored ahead of the rest.

    Fst measures more recent relationships because it takes into account recent drift, but the potential problem is that highly drifted populations will look like outliers, even though they may not differ from their neighbors in terms of deep ancestry.

  12. The Following 5 Users Say Thank You to Generalissimo For This Useful Post:

     Hando (02-15-2020),  Helen (02-15-2020),  JMcB (02-15-2020),  JonikW (02-15-2020),  Nebuchadnezzar II (02-15-2020)

  13. #7
    Registered Users
    Posts
    2,689
    Sex
    Y-DNA (P)
    R1b-DF27 FGC17112
    mtDNA (M)
    H5a1

    This is not new and is rife with all kinds of problems. This conversation goes back to 2012 in the community.

    Not really true about formal stats either. It doesn't favor extremes. That only comes into play in the case where a certain subset of one's ancestry doesn't totally cover the variation there. For instance, it may pull you towards, West Siberia N, then into Iran, or vice versa. This just means that the Eastern ancestry can't and shouldn't be constrained by one source. If you're finding two things, rather than one, it means that neither by itself is the perfect source. This is what is so great about qpGraph. You don't have to settle for specific pops, and struggle with outgroups like qpAdm. We also know it doesn't just go to extremes from looking at things like D-stats, where a younger branch that contributes ancestry is the more significant number and also gets favored in qpGraph. I'm not sure where you're getting this from. It just isn't the case. FST is about as useless as Admixture and PCA. FST is highly dependent on the SNP subset and also highly affected by ROH, which really affects ancient samples.

    But the whole ancestry extremes thing is wrong. It goes to the drift paths, f2 and not just to the extreme. Look at CHG in Caucasus Eneolithic, for example. It goes there, instead of Iran N. For EHG, it is a different story. It is kind of significantly closer to CHG and Ganj_Dareh, but this also has to deal with the huge amount of Anatolia_HG-like stuff in CHG, as well as some small reciprocal with EHG. The fact that the stat isn't huge says that the most likely donor to EHG isn't really much closer to either Ganj Dareh or CHG. It is a bit of a toss up.

    That doesn't happen for something like Caucasus Eneolithic or Piedmont for a reason. It isn't about extremes. You can see it with Native Americans too. Why not go to the Yana or Goyet extreme, rather than Afontova Gora or MA1? This takes some thinking to understand your results. If you get a population that you aren't expecting, perhaps the stats you were expected were confounded by something else and you need to investigate it more. That is good science. Throwing it into the trash heap and reverting to methods way back in 2013 just doesn't make sense. They're so weak that they can't tell you the difference between CHG and Caucasus Eneolithic. qpGraph would fail miserably because it is so much better. This isn't so hard to see.

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

     Hando (02-15-2020),  Nebuchadnezzar II (02-15-2020),  ThaYamamoto (02-28-2020)

  15. #8
    Banned
    Posts
    2,951
    Sex
    Location
    America
    Ethnicity
    Telgu/marathi

    Siberian Tatars Tajikistan
    It seems some of the distances on there are questionable. The difference between Pathans and Brahmin Tamil is only .004, which seems way too close.

  16. The Following User Says Thank You to Censored For This Useful Post:

     Hando (02-15-2020)

  17. #9
    Moderator
    Posts
    7,126
    Sex
    Location
    Normandy
    Ethnicity
    northwesterner
    Y-DNA (P)
    R-BY3604-Z275
    mtDNA (M)
    H5a1

    Normandie Orkney Netherlands Friesland East Frisia Finland
    For info, general issues with Fst have been adressed in https://genome.cshlp.org/content/ear...1.113.full.pdf (2013).
    Issues with using Fst between modern and ancient genomes are well exposed in https://www.biorxiv.org/content/10.1...053v2.full.pdf
    En North alom, de North venom
    En North fum naiz, en North manom

    (Roman de Rou, Wace, 1160-1170)

  18. The Following 6 Users Say Thank You to anglesqueville For This Useful Post:

     grumpydaddybear (02-15-2020),  Hando (02-15-2020),  JMcB (02-15-2020),  Kulin (02-16-2020),  michal3141 (02-15-2020),  Nebuchadnezzar II (02-15-2020)

  19. #10


    Fig. S4. Outgroup f3 statistics. (A) Outgroup f3-statistics calculated in the form of f3(Yoruba, Srubnaya-Alakulskaya_individual, Modern_population) projected on a geographical map.

    Srubnaya-Alakulskaya individuals from the eastern Pontic-Caspian steppe exhibited genetic affinity to northern and northeastern present-day Europeans (Krzewińska et al. 2018). In Fig. S4, warmer colors show more genetic drift shared between Srubnaya-Alakulskaya individuals and modern populations. Outgroup f3 statistics show modern Norwegians are the closest match to Srubnaya-Alakulskaya, followed by Lithuanians and Finns.

    Outgroup Pop1 Pop2 f3 std.err Z SNPs
    Yoruba chy001 Norwegian 0.155436 0.001837 84.622 121054
    Yoruba chy001 Lithuanian 0.154615 0.00184 84.045 121030
    Yoruba chy001 Finnish 0.15354 0.001884 81.486 121015
    Yoruba chy001 Estonian 0.15347 0.001832 83.788 121060
    Yoruba chy001 English 0.153355 0.001804 84.994 121046
    Yoruba chy001 Ukrainian_East 0.153211 0.001939 79.023 120941
    Yoruba chy001 Belarusian 0.153125 0.001783 85.9 121032
    Yoruba chy001 Icelandic 0.153081 0.001771 86.422 121080
    Yoruba chy001 Mordovian 0.153047 0.001813 84.433 121057
    Yoruba chy001 Orcadian 0.153036 0.001814 84.343 121086
    Yoruba chy001 Czech 0.152836 0.001827 83.645 121062
    Yoruba chy001 Croatian 0.152733 0.001828 83.549 121032
    Yoruba chy001 Hungarian 0.152677 0.001753 87.09 121134
    Yoruba chy001 Russian 0.152634 0.001741 87.656 121132
    Yoruba chy001 Lezgin 0.151885 0.001801 84.325 121029
    Yoruba chy001 Scottish 0.151829 0.002028 74.873 120825
    Yoruba chy001 French 0.15146 0.001735 87.285 121140
    Yoruba chy001 Basque_French 0.151323 0.001815 83.378 121109
    Yoruba chy001 Basque_Spanish 0.151083 0.001899 79.549 120994
    Yoruba chy001 Chechen 0.150699 0.001769 85.195 121023
    Yoruba chy001 Bergamo 0.150488 0.001797 83.74 121052
    Yoruba chy001 French_South 0.150372 0.001854 81.12 120952
    Yoruba chy001 Albanian 0.150273 0.001858 80.898 120922
    Yoruba chy001 Tuscan 0.150131 0.001807 83.084 120985
    Yoruba chy001 Bulgarian 0.1501 0.001815 82.709 121034
    Yoruba chy001 Ukrainian_West 0.149952 0.002121 70.71 120809
    Yoruba chy001 Chuvash 0.149316 0.001775 84.111 121038

    Ancient genomes suggest the eastern Pontic-Caspian steppe as the source of western Iron Age nomads

    https://advances.sciencemag.org/content/4/10/eaat4457

    SUPPLEMENTARY MATERIALS

    https://advances.sciencemag.org/cont...aat4457_SM.pdf
    Last edited by ThirdTerm; 02-16-2020 at 06:45 AM.
    Давайте вместе снова сделаем мир великий!

  20. The Following 2 Users Say Thank You to ThirdTerm For This Useful Post:

     Hando (02-16-2020),  palamede (02-16-2020)

Similar Threads

  1. Post your closest 100 Global 25 modern populations
    By Molfish in forum Autosomal (auDNA)
    Replies: 309
    Last Post: 10-28-2020, 08:39 PM
  2. Replies: 4
    Last Post: 12-01-2019, 03:59 AM
  3. Replies: 282
    Last Post: 08-01-2017, 09:58 PM
  4. Replies: 52
    Last Post: 03-30-2017, 09:34 PM
  5. Replies: 12
    Last Post: 07-26-2016, 10:31 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
  •