PDA

View Full Version : The genetic history of the Southern Arc: A bridge between West Asia and Europe



Liquid
09-10-2022, 09:36 AM
Can be found here https://www.science.org/doi/10.1126/science.abm4247

There's been an explosion of the number of sources we have, gone is the pre-Turkic Anatolian proxies found in Rome, replaced by the real thing, e.g.

Target: Liquid_scaled
Distance: 1.3729% / 0.01372943 | R3P
59.0 TUR_Marmara_Balikesir_Byz
27.0 RUS_Maykop_Novosvobodnaya
14.0 RUS_Late_Xiongnu_Han

Has any knowledgeable created a more recent calculator specific for us Turks using the new samples? Is Karluk DA 230 still the best proxy we have for the Turks who entered Turkey?

chum
09-10-2022, 10:05 AM
Best proxy for the Turks who entered Turkey should be like Turkmens, there was some Persian-type ancestry in the Turcomans by the time they would go to Anatolia, until there are samples from Medieval Turkmens (from around Turkmenistan) one could use Turkmens for this, I made some ghost coordinates subtracting Turkmen ancestry from Azerbaijanis and the Turkish and the results look right to represent what would be present in their respective regions before the Turkic era:


Azerbaijani_-TM,0.09431984,0.12530476,-0.06089248,-0.05014942,-0.03499198,-0.00513775,0.00361024,-0.00592517,-0.03097814,-0.00942902,0.00021378,0.0001297,0.00066829,-0.00077064,0.00141252,0.00427857,-0.00286901,0.00153212,0.00273249,-0.00540471,0.0016595,-0.00076508,-0.00029933,-0.00255124,0.00029858
Azerbaijani_Dagestan_-TM,0.1024737,0.12167912,-0.04662607,-0.0269497,-0.03487358,0.00276814,0.00587176,-0.00619781,-0.04355794,-0.01953948,-0.0008195,0.00451008,-0.0039396,-0.00062118,0.00807469,-0.00096343,-0.00446357,0.00008611,-0.00241289,-0.00419632,0.00175313,-0.00045844,0.00157207,0.00029574,0.00028111
Azerbaijani_Iran_-TM,0.09512786,0.12342341,-0.0615538,-0.04682997,-0.03581202,-0.00538432,0.00418033,-0.00605453,-0.03129407,-0.01132236,0.0019775,-0.00001562,0.00105493,-0.00006383,0.00092797,0.00628782,-0.00216164,0.00146293,0.00283039,-0.00587792,0.00263806,-0.00263988,-0.00111256,-0.00259795,0.00148965
Azerbaijani_Turkey_-TM,0.0972102,0.1284824,-0.0570909,-0.04544859,-0.03548146,-0.00712048,0.00372142,-0.00717919,-0.02735719,-0.00749627,-0.00102235,0.00047891,-0.00055804,0.0042993,-0.00001996,-0.00924786,-0.01524824,0.00691777,0.00691215,-0.00627089,0.00192911,0.00120037,-0.00163756,-0.00315238,0.002197
Turkish_Bulgaria_N_-TM,0.11838813,0.14115883,0.02613636,0.00122828,0.0 2331506,0.00290079,0.00460392,0.00721252,-0.00346323,0.00140149,0.0003465,-0.00402527,0.00355819,0.01104825,-0.01139359,-0.00577074,0.00341487,0.00223885,0.00307736,-0.00231459,-0.00429347,-0.00087488,-0.0003154,0.00126115,-0.00098469
Turkish_Bulgaria_S_-TM,0.1216158,0.14147719,0.02334202,-0.00629292,0.02707149,-0.00577211,-0.00052971,0.00091347,0.00050252,0.00787786,-0.00178811,0.00277443,0.00183636,0.00430741,-0.01110425,0.00806896,0.01364746,-0.00253941,0.00520794,0.00401912,-0.00481778,0.00209825,0.00608306,0.00229013,-0.00146721
Turkish_C_-TM,0.1033198,0.13733331,-0.04074908,-0.05575698,-0.01675969,-0.01602717,0.00221702,-0.00395707,-0.02339886,0.00192233,0.00211773,0.00302628,-0.00383253,0.00420445,-0.0069112,-0.00300254,0.00199736,0.0009304,0.0019235,-0.00238634,0.0003817,0.0028896,-0.00188768,-0.00144761,-0.00093749
Turkish_Cyprus_-TM,0.08566194,0.14430484,-0.03573901,-0.06725138,-0.00230742,-0.02588767,0.00273865,-0.00485424,-0.0009047,0.01403756,0.00428555,0.00022242,-0.00011611,0.00584296,-0.00931656,-0.00185039,0.00545902,0.0006006,0.00642116,-0.00264641,0.00032945,0.00165917,0.00147171,-0.00154892,0.0004467
Turkish_Dodecanese_-TM,0.10332936,0.14571805,-0.02561893,-0.05825267,0.00343348,-0.02228933,0.00179748,-0.00373114,-0.00213377,0.01412291,0.00422708,-0.00025032,0.00347478,0.00773499,-0.01990986,-0.0068583,-0.01006047,0.00046051,0.00097947,-0.00782329,0.00230601,-0.00461297,0.0005666,-0.00363418,0.00186521
Turkish_E_-TM,0.10255123,0.1278654,-0.05367218,-0.04302342,-0.03480949,-0.00865314,0.00364549,-0.00495998,-0.03201918,-0.00498501,0.00401476,0.00234626,-0.00478626,-0.00150623,-0.00210685,0.00125092,0.00361556,0.0020569,0.00093 425,-0.00380778,0.00285992,-0.00027284,-0.00185782,-0.00435929,-0.0036181
Turkish_Greece_N_-TM,0.11463754,0.14167086,0.01555865,-0.01151723,0.01268148,-0.0089838,0.0001825,0.0031354,0.00075116,0.0076079 8,0.00559154,0.00105429,-0.00040952,0.00046904,-0.01715291,0.01063314,0.01437586,0.00013522,0.0094 1402,0.00409244,-0.00406126,0.00351531,0.00059046,0.00601192,-0.00067131
Turkish_N_-TM,0.11280346,0.14500145,-0.03206984,-0.06042983,-0.01049881,-0.02465068,0.00491407,-0.00461797,-0.02510308,0.00671027,0.0043217,0.00639702,-0.01552594,0.00549609,-0.01176921,-0.0117558,0.00273286,0.00005633,0.00112246,-0.00030261,-0.00387689,0.00496734,-0.00271085,0.00046386,-0.00209683
Turkish_NW_-TM,0.11090009,0.14518009,-0.013359,-0.04649766,-0.00012795,-0.01973231,0.00530054,-0.00093556,-0.01615288,0.01026825,0.00498493,0.00496439,-0.00667103,0.009053,-0.00446346,-0.01009349,0.0021214,0.00189251,0.00470318,0.00131 978,-0.00340106,0.00373483,0.00175538,0.00164584,-0.00266924
Turkish_S_-TM,0.10159109,0.13919461,-0.03860815,-0.0563002,-0.02086816,-0.01856603,0.00538591,-0.00772107,-0.020555,0.00721339,0.0044583,0.00601009,-0.00584578,0.00700432,-0.00564883,-0.00212036,0.00235636,0.00327942,0.00425236,0.0016 2405,0.00181055,0.00376172,-0.00168848,0.00537775,0.00133672
Turkish_SE_-TM,0.09108069,0.12948629,-0.04497226,-0.05085889,-0.02889069,-0.00718961,0.00370053,-0.00477942,-0.02165782,-0.00609664,0.00815327,0.00660428,0.00052764,-0.00333635,-0.00165024,0.00911649,0.00711159,0.0039464,-0.00287459,-0.01106533,0.00454366,-0.00277613,0.00252937,0.00814904,-0.00070172
Turkish_SW_-TM,0.10361009,0.14322262,-0.02245039,-0.05655691,-0.00626521,-0.02145582,0.0022355,-0.00419245,-0.01639522,0.00891485,0.00401909,0.00234399,-0.00829464,0.00528304,-0.01206612,0.00369822,0.01253488,-0.00033562,0.00376553,0.00051466,-0.00035499,-0.00128646,-0.00256007,-0.00058901,-0.00414144

bovefex
09-10-2022, 10:37 AM
Best proxy for the Turks who entered Turkey should be like Turkmens, there was some Persian-type ancestry in the Turcomans by the time they would go to Anatolia, until there are samples from Medieval Turkmens (from around Turkmenistan) one could use Turkmens for this, I made some ghost coordinates subtracting Turkmen ancestry from Azerbaijanis and the Turkish and the results look right to represent what would be present in their respective regions before the Turkic era:


Azerbaijani_-TM,0.09431984,0.12530476,-0.06089248,-0.05014942,-0.03499198,-0.00513775,0.00361024,-0.00592517,-0.03097814,-0.00942902,0.00021378,0.0001297,0.00066829,-0.00077064,0.00141252,0.00427857,-0.00286901,0.00153212,0.00273249,-0.00540471,0.0016595,-0.00076508,-0.00029933,-0.00255124,0.00029858
Azerbaijani_Dagestan_-TM,0.1024737,0.12167912,-0.04662607,-0.0269497,-0.03487358,0.00276814,0.00587176,-0.00619781,-0.04355794,-0.01953948,-0.0008195,0.00451008,-0.0039396,-0.00062118,0.00807469,-0.00096343,-0.00446357,0.00008611,-0.00241289,-0.00419632,0.00175313,-0.00045844,0.00157207,0.00029574,0.00028111
Azerbaijani_Iran_-TM,0.09512786,0.12342341,-0.0615538,-0.04682997,-0.03581202,-0.00538432,0.00418033,-0.00605453,-0.03129407,-0.01132236,0.0019775,-0.00001562,0.00105493,-0.00006383,0.00092797,0.00628782,-0.00216164,0.00146293,0.00283039,-0.00587792,0.00263806,-0.00263988,-0.00111256,-0.00259795,0.00148965
Azerbaijani_Turkey_-TM,0.0972102,0.1284824,-0.0570909,-0.04544859,-0.03548146,-0.00712048,0.00372142,-0.00717919,-0.02735719,-0.00749627,-0.00102235,0.00047891,-0.00055804,0.0042993,-0.00001996,-0.00924786,-0.01524824,0.00691777,0.00691215,-0.00627089,0.00192911,0.00120037,-0.00163756,-0.00315238,0.002197
Turkish_Bulgaria_N_-TM,0.11838813,0.14115883,0.02613636,0.00122828,0.0 2331506,0.00290079,0.00460392,0.00721252,-0.00346323,0.00140149,0.0003465,-0.00402527,0.00355819,0.01104825,-0.01139359,-0.00577074,0.00341487,0.00223885,0.00307736,-0.00231459,-0.00429347,-0.00087488,-0.0003154,0.00126115,-0.00098469
Turkish_Bulgaria_S_-TM,0.1216158,0.14147719,0.02334202,-0.00629292,0.02707149,-0.00577211,-0.00052971,0.00091347,0.00050252,0.00787786,-0.00178811,0.00277443,0.00183636,0.00430741,-0.01110425,0.00806896,0.01364746,-0.00253941,0.00520794,0.00401912,-0.00481778,0.00209825,0.00608306,0.00229013,-0.00146721
Turkish_C_-TM,0.1033198,0.13733331,-0.04074908,-0.05575698,-0.01675969,-0.01602717,0.00221702,-0.00395707,-0.02339886,0.00192233,0.00211773,0.00302628,-0.00383253,0.00420445,-0.0069112,-0.00300254,0.00199736,0.0009304,0.0019235,-0.00238634,0.0003817,0.0028896,-0.00188768,-0.00144761,-0.00093749
Turkish_Cyprus_-TM,0.08566194,0.14430484,-0.03573901,-0.06725138,-0.00230742,-0.02588767,0.00273865,-0.00485424,-0.0009047,0.01403756,0.00428555,0.00022242,-0.00011611,0.00584296,-0.00931656,-0.00185039,0.00545902,0.0006006,0.00642116,-0.00264641,0.00032945,0.00165917,0.00147171,-0.00154892,0.0004467
Turkish_Dodecanese_-TM,0.10332936,0.14571805,-0.02561893,-0.05825267,0.00343348,-0.02228933,0.00179748,-0.00373114,-0.00213377,0.01412291,0.00422708,-0.00025032,0.00347478,0.00773499,-0.01990986,-0.0068583,-0.01006047,0.00046051,0.00097947,-0.00782329,0.00230601,-0.00461297,0.0005666,-0.00363418,0.00186521
Turkish_E_-TM,0.10255123,0.1278654,-0.05367218,-0.04302342,-0.03480949,-0.00865314,0.00364549,-0.00495998,-0.03201918,-0.00498501,0.00401476,0.00234626,-0.00478626,-0.00150623,-0.00210685,0.00125092,0.00361556,0.0020569,0.00093 425,-0.00380778,0.00285992,-0.00027284,-0.00185782,-0.00435929,-0.0036181
Turkish_Greece_N_-TM,0.11463754,0.14167086,0.01555865,-0.01151723,0.01268148,-0.0089838,0.0001825,0.0031354,0.00075116,0.0076079 8,0.00559154,0.00105429,-0.00040952,0.00046904,-0.01715291,0.01063314,0.01437586,0.00013522,0.0094 1402,0.00409244,-0.00406126,0.00351531,0.00059046,0.00601192,-0.00067131
Turkish_N_-TM,0.11280346,0.14500145,-0.03206984,-0.06042983,-0.01049881,-0.02465068,0.00491407,-0.00461797,-0.02510308,0.00671027,0.0043217,0.00639702,-0.01552594,0.00549609,-0.01176921,-0.0117558,0.00273286,0.00005633,0.00112246,-0.00030261,-0.00387689,0.00496734,-0.00271085,0.00046386,-0.00209683
Turkish_NW_-TM,0.11090009,0.14518009,-0.013359,-0.04649766,-0.00012795,-0.01973231,0.00530054,-0.00093556,-0.01615288,0.01026825,0.00498493,0.00496439,-0.00667103,0.009053,-0.00446346,-0.01009349,0.0021214,0.00189251,0.00470318,0.00131 978,-0.00340106,0.00373483,0.00175538,0.00164584,-0.00266924
Turkish_S_-TM,0.10159109,0.13919461,-0.03860815,-0.0563002,-0.02086816,-0.01856603,0.00538591,-0.00772107,-0.020555,0.00721339,0.0044583,0.00601009,-0.00584578,0.00700432,-0.00564883,-0.00212036,0.00235636,0.00327942,0.00425236,0.0016 2405,0.00181055,0.00376172,-0.00168848,0.00537775,0.00133672
Turkish_SE_-TM,0.09108069,0.12948629,-0.04497226,-0.05085889,-0.02889069,-0.00718961,0.00370053,-0.00477942,-0.02165782,-0.00609664,0.00815327,0.00660428,0.00052764,-0.00333635,-0.00165024,0.00911649,0.00711159,0.0039464,-0.00287459,-0.01106533,0.00454366,-0.00277613,0.00252937,0.00814904,-0.00070172
Turkish_SW_-TM,0.10361009,0.14322262,-0.02245039,-0.05655691,-0.00626521,-0.02145582,0.0022355,-0.00419245,-0.01639522,0.00891485,0.00401909,0.00234399,-0.00829464,0.00528304,-0.01206612,0.00369822,0.01253488,-0.00033562,0.00376553,0.00051466,-0.00035499,-0.00128646,-0.00256007,-0.00058901,-0.00414144

It still hasn't been proven that the Turks who first entered Anatolia had Persian-like ancestry though. One of the two samples from the Ottomans looked almost exatly like Karluks/Karakhanids.

chum
09-10-2022, 11:21 AM
It still hasn't been proven that the Turks who first entered Anatolia had Persian-like ancestry though. One of the two samples from the Ottomans looked almost exatly like Karluks/Karakhanids.

That sample is near identical to a Turkmen from Uzbekistan, the other Ottoman sample has some Persian and local ancestry, they varied like modern Turkmens, Seljuk era Turkmens on average being similar to Modern Turkmens is a safe assumption IMO

bovefex
09-10-2022, 11:37 AM
That sample is near identical to a Turkmen from Uzbekistan, the other Ottoman sample has some Persian and local ancestry, they varied like modern Turkmens, Seljuk era Turkmens on average being similar to Modern Turkmens is a safe assumption IMO

Distance to: TUR_Ottoman:MA2195
0.04180672 KAZ_Karakhanid
0.04580906 KAZ_Karluk
0.05843827 Uygur
0.06699981 Uzbek
0.10497331 Turkmen_Uzbekistan
0.11363438 Turkmen

Distance to: TUR_Ottoman:MA2196
0.04926603 Turkmen
0.05962663 Turkmen_Uzbekistan
0.11261817 Uzbek
0.15558670 KAZ_Karluk
0.16029605 Uygur
0.17697458 KAZ_Karakhanid

The first one is much more similar to KAZ_Karakhanid and KAZ_Karluk, so it isn't near identical to Turkmen_Uzbekistan.

Target / Distance / Greek_Central_Anatolia/ Iranian_Fars /KAZ_Karluk
TUR_Ottoman:MA2195 /0.04556980 /0.0 / 1.8 / 98.2
TUR_Ottoman:MA2196 /0.03626992 /37.0 / 13.2 / 49.8
Turkmen / 0.01882160 /14.6 / 26.0 / 59.4
Turkmen_Uzbekistan / 0.01967988 /10.8 / 26.6 / 62.6
Average / 0.03008530 /15.6 / 16.9 / 67.5

Obviously the model I used is awful, but it is still obvious that the non-Turkic part of MA196's ancestry is much more shifted towards Greek_Central_Anatolia, while that of the Turkmen samples are much more shifted towards Iranian_Fars (likely bad proxy, but enough in this case imo).


Target: TUR_Ottoman:MA2196
Distance: 3.6270% / 0.03626992
49.8 KAZ_Karluk
37.0 Greek_Central_Anatolia
13.2 Iranian_Fars


Target: TUR_Ottoman:MA2196
Distance: 3.7302% / 0.03730182
52.4 KAZ_Karluk
47.6 Greek_Central_Anatolia

When Iranian_Fars is removed, the distance barely changes, which in my opinion points to the % of the Persian-like proxy being due to noise or something like that.

Alkaevli
09-10-2022, 11:43 AM
It still hasn't been proven that the Turks who first entered Anatolia had Persian-like ancestry though. One of the two samples from the Ottomans looked almost exatly like Karluks/Karakhanids.
What makes you think that they were homogeneous? The main migrations associated with the Oghuz lasted roughly two centuries (not including smaller-scaled later migrations such as that of Afshar tribes from Iran), and admixture dates provided by this paper clearly indicate that the admixture took place within a few centuries after the conquest.

An admixture date estimate of 12.2 ± 1.4 generations before their time using Roman-Byzantine and Central Asian sources suggests that the admixture occurred in the period surrounding the 11th century arrival and expansion of Seljuk Turks to Anatolia. Present-day Turkish individuals have an admixture date estimate of 30.6 ± 1.9 generations and thus from the same early centuries of the 1000s CE, which coincided with the transfer of control of Anatolia from the Romans to the Seljuks and eventually the Ottomans.

There are roughly 200 years between the first Oghuz who entered Anatolia and those who entered the peninsula following the Mongol Invasion of Khwarezm. It is quite unlikely that the Oghuz in Khwarezm and its surrounding regions remained genetically intact while their kin in Roman Anatolia were mixing.

Have you checked the difference between Byzantine Muğla (AD 650-1300) and Menteşe-Ottoman Muğla (AD 1300-1650)? I have.
https://abload.de/img/adsize7evl.png

For instance, this woman (I20325, AD 1300-1650) doesn't seem like a mixture of Byzantine Muğla and Karluk-profile, her non-Turkic admixture is clearly more eastern (Iran-shifted?) than the Byzantine population of Muğla.
https://abload.de/img/andnneh6.png

Straboo
09-10-2022, 12:00 PM
That sample is near identical to a Turkmen from Uzbekistan, the other Ottoman sample has some Persian and local ancestry, they varied like modern Turkmens, Seljuk era Turkmens on average being similar to Modern Turkmens is a safe assumption IMO

Whats the difference between Turkish and Turkmens? Turkmens didnt mix with locals as much and stayed more "pure" central asian Turkic?

Also wont using models from Turkmenistan show that the turkic invaders in anatolia were local west central asians instead of north east central asians? Or are the medieval turkmenistan samples expected to show proper north east asian turkic ancestry? In terms of genetics it pretty well established now Turkmenistan people have relatively recently switched language and culture, and their east asian components are not higher than 30% iirc

Alkaevli
09-10-2022, 12:02 PM
This perhaps doesn't indicate that the Turkic population as a whole was Turkmen-like (and I think many tribes were genetically intact/had no recent admix), but I think it indicates the admixture process was more complex than we thought.

Byzantine SW Anatolian pop + Karluk-like pop is not the best scenario here for sure.
https://abload.de/img/adsizk8fv6.png

chum
09-10-2022, 12:02 PM
Distance to: TUR_Ottoman:MA2195
0.04180672 KAZ_Karakhanid
0.04580906 KAZ_Karluk
0.05843827 Uygur
0.06699981 Uzbek
0.10497331 Turkmen_Uzbekistan
0.11363438 Turkmen

Distance to: TUR_Ottoman:MA2196
0.04926603 Turkmen
0.05962663 Turkmen_Uzbekistan
0.11261817 Uzbek
0.15558670 KAZ_Karluk
0.16029605 Uygur
0.17697458 KAZ_Karakhanid

The first one is much more similar to KAZ_Karakhanid and KAZ_Karluk, so it isn't near identical to Turkmen_Uzbekistan.

Target / Distance / Greek_Central_Anatolia/ Iranian_Fars /KAZ_Karluk
TUR_Ottoman:MA2195 /0.04556980 /0.0 / 1.8 / 98.2
TUR_Ottoman:MA2196 /0.03626992 /37.0 / 13.2 / 49.8
Turkmen / 0.01882160 /14.6 / 26.0 / 59.4
Turkmen_Uzbekistan / 0.01967988 /10.8 / 26.6 / 62.6
Average / 0.03008530 /15.6 / 16.9 / 67.5

Obviously the model I used is awful, but it is still obvious that the non-Turkic part of MA196's ancestry is much more shifted towards Greek_Central_Anatolia, while that of the Turkmen samples are much more shifted towards Iranian_Fars (likely bad proxy, but enough in this case imo).


Target: TUR_Ottoman:MA2196
Distance: 3.6270% / 0.03626992
49.8 KAZ_Karluk
37.0 Greek_Central_Anatolia
13.2 Iranian_Fars


Target: TUR_Ottoman:MA2196
Distance: 3.7302% / 0.03730182
52.4 KAZ_Karluk
47.6 Greek_Central_Anatolia

When Iranian_Fars is removed, the distance barely changes, which in my opinion points to the % of the Persian-like proxy being due to noise or something like that.

Try with individuals, I meant the UZB105 sample for Turkmen_Uzbekistan

Also MA2196's coverage is 18.63%, so the distance thing might be due to that

Also check the ghosts I made by subtracting Modern Turkmens from Azerbaijanis and Turkish, they're geographically consistent

chum
09-10-2022, 12:08 PM
Whats the difference between Turkish and Turkmens? Turkmens didnt mix with locals as much and stayed more "pure" central asian Turkic?

Also wont using models from Turkmenistan show that the turkic invaders in anatolia were local west central asians instead of north east central asians? Or are the medieval turkmenistan samples expected to show proper north east asian turkic ancestry? In terms of genetics it pretty well established now Turkmenistan people have relatively recently switched language and culture, and their east asian components are not higher than 30% iirc

Yeah, Turkish are a mix of Turkmens (who were mixed with Western Iranians, not entirely local, Turkmens vary from being around 20% iirc to 45% East Eurasian, around 25-30% on average) and local groups in West Asia basically

chum
09-10-2022, 12:11 PM
This perhaps doesn't indicate that the Turkic population as a whole was Turkmen-like (and I think many tribes were genetically intact/had no recent admix), but I think it indicates the admixture process was more complex than we thought.

Byzantine SW Anatolian pop + Karluk-like pop is not the best scenario here for sure.
https://abload.de/img/adsizk8fv6.png

I wouldn't be surprised if people from like Muğla or Bolu had a more East Eurasian source of Turkic ancestry, the Turkmen average isn't East Eurasian enough for them

Straboo
09-10-2022, 12:22 PM
Yeah, Turkish are a mix of Turkmens (who were mixed with Western Iranians, not entirely local, Turkmens vary from being around 20% iirc to 45% East Eurasian, around 25-30% on average) and local groups in West Asia basically

Im confused by the labels 'Turkmens' and 'Turkmenistani'

By turkmens I meant turkmens/turkomans in turkey/syria/iraq etc

45% east asian too high for a national average of the citizens of Turkmenistan the country. And they are/were east iranians (not west iranians) deriving most ancestry from BMAC

So Turkmens in anatolia/Syria/Iraq would be east asian + east iranian (turanian), then + west iranian with minor local ancestry then normal turkish people?

Wheras normal turkish (not kurdish or Alevi) are east asian + Turanian + west iranian + roughly 60 / 50 % anatolia?

Just how much "east asian" without any west iranian or Turanian do anatolian turks have? I mean like proper kazakh altai proto turk ancestry

Alkaevli
09-10-2022, 12:29 PM
In terms of genetics it pretty well established now Turkmenistan people have relatively recently switched language and culture, and their east asian components are not higher than 30% iirc
I'm seriously tired of the habit of summarizing the ethnogenesis of every Turkic people as language shift.

Those guys tried modeling modern Turkmen as a mixture of modern Tajiks and Golden_Horde_Asian (100% East Eurasian, even more than generic medieval Mongols) and found that modern Turkmens are 8-16% Turkic or something like that. I really hope that they are just idiots. People with this level of bias should not be publishing articles.

Modern Turkmens can roughly be modeled as a mixture of medieval Turkic samples from KAZ and the populations of the Iranian plateau (not Tajiks), which is also historically accurate.

Target: Turkmen
Distance: 1.1951% / 0.01195087
37.4 Iranian_Fars
31.4 KAZ_Turk
20.4 KAZ_Karakhanid
7.2 KAZ_Karluk
3.6 KAZ_Kimak

Target: Turkmen_Uzbekistan
Distance: 1.5048% / 0.01504767
36.0 Iranian_Fars
25.6 KAZ_Karakhanid
23.6 KAZ_Turk
13.8 KAZ_Karluk
1.0 KAZ_Kimak

Modern Turkmens descend for the most part from Turkic-speaking populations of medieval Kazakhstan. The medieval Turkic populations of Kazakhstan are not genetically proto-Turkic-like of course, they represent a mixture between preceding Iranic-speaking populations and Turkic-speakers from the east. But to say modern Turkmens have recently switched language is to speak ignorantly of them.

Alkaevli
09-10-2022, 12:33 PM
Wheras normal turkish (not kurdish or Alevi) are east asian + Turanian + west iranian + roughly 60 / 50 % anatolia?

Alevism is merely a sect. Turkish Alevis from Anatolia proper (west of Euphrates) cluster with their Sunni Turkish counterparts. Kurdish/Zaza Alevis on the other hand cluster with Sunni Kurds/Zazas.



Just how much "east asian" without any west iranian or Turanian do anatolian turks have? I mean like proper kazakh altai proto turk ancestry
You can try modeling Turkish averages as a mix of some relevant West Eurasian pops + Slab Grave. It depends on the region. Turks from Eastern Black Sea (Trabzon and Rize) are indistinguishable from Greeks of the same region for instance.

East Eurasian admixture in Anatolian Turks and other pops. Dodecad K12b is an old admixture calculator, but it still works well when it comes to estimating the amount of East Eurasian admixture.
https://docs.google.com/spreadsheets/d/1rNnXVwnlj-dOnb6bfVWgd2GPxvlQyGPL9fE9NZklvvs/htmlview#gid=0

chum
09-10-2022, 12:45 PM
Im confused by the labels 'Turkmens' and 'Turkmenistani'

By turkmens I meant turkmens/turkomans in turkey/syria/iraq etc

45% east asian too high for a national average of the citizens of Turkmenistan the country. And they are/were east iranians (not west iranians) deriving most ancestry from BMAC

So Turkmens in anatolia/Syria/Iraq would be east asian + east iranian (turanian), then + west iranian with minor local ancestry then normal turkish people?

Wheras normal turkish (not kurdish or Alevi) are east asian + Turanian + west iranian + roughly 60 / 50 % anatolia?

Just how much "east asian" without any west iranian or Turanian do anatolian turks have? I mean like proper kazakh altai proto turk ancestry

45% is not average, just saying it can go up to that much

Turkmens in Anatolia are different to Turkmens in Syria/Iraq, they're closer to the Turkish than to Syrian/Iraqi Turkmens, I can't say for Syrian Turkmens but Iraqi Turkmens look like a mix of Mesopotamian/Mesopotamian Arab + Azerbaijani

Local ancestry in Turkish is about 55%-75% usually, outliers being in the East and Northeast

Turks usually have around 10%-15% East Eurasian, but some can go to about 20% (Muğla, Bolu) and some can be undetectable (East, Northeast)

Straboo
09-10-2022, 01:06 PM
I'm seriously tired of the habit of summarizing the ethnogenesis of every Turkic people as language shift.

Those guys tried modeling modern Turkmen as a mixture of modern Tajiks and Golden_Horde_Asian (100% East Eurasian, even more than generic medieval Mongols) and found that modern Turkmens are 16% Turkic or something like that. I really hope that they are just idiots. People with this level of bias should not be publishing articles.

Modern Turkmens can roughly be modeled as a mixture of medieval Turkic samples from KAZ and the populations of the Iranian plateau (not Tajiks), which is also historically accurate.

Target: Turkmen
Distance: 1.1951% / 0.01195087
37.4 Iranian_Fars
31.4 KAZ_Turk
20.4 KAZ_Karakhanid
7.2 KAZ_Karluk
3.6 KAZ_Kimak

Target: Turkmen_Uzbekistan
Distance: 1.5048% / 0.01504767
36.0 Iranian_Fars
25.6 KAZ_Karakhanid
23.6 KAZ_Turk
13.8 KAZ_Karluk
1.0 KAZ_Kimak

Modern Turkmens descend for the most part from Turkic-speaking populations of medieval Kazakhstan. The medieval Turkic populations of Kazakhstan are not genetically proto-Turkic-like of course, they represent a mixture between preceding Iranic-speaking populations and Turkic-speakers from the east. But to say modern Turkmens have recently switched language is to speak ignorantly of them.

I was reffering to the population of Turkmenistan. Not pan national Turkmens. Maybe its correct to call Turkmenistan people Turkmens, but I call them Turkmenistanis to avoid confusion.

It could equally be said that you are biased towards to a pro turkic outcome, and your research is not peer reviewed either.

I can only rely on published articles that everyone has access to and can review, and in 2 recent publications that dealt with central asia, it was explicitly stated that the population of Turkmenistan is considerably less "Turkic" than Kyrgzistan, Uzbekistan and Kazakhstan, and harbours a dominant east iranic ancestry (BMAC + middle steppe iirc). It seems that country in particular was more of a transit zone for turks and probay became turkic in culture and language much later than anatolia, probably because of turkish persian syncretism.

I dont think using Iranian_Fars to model (former) east iranians makes any sense. West and east iranian are 2 different genetic "provinces" iirc.

Straboo
09-10-2022, 01:07 PM
Alevism is merely a sect. Turkish Alevis from Anatolia proper (west of Euphrates) cluster with their Sunni Turkish counterparts. Kurdish/Zaza Alevis on the other hand cluster with Sunni Kurds/Zazas.


You can try modeling Turkish averages as a mix of some relevant West Eurasian pops + Slab Grave. It depends on the region. Turks from Eastern Black Sea (Trabzon and Rize) are indistinguishable from Greeks of the same region for instance.

East Eurasian admixture in Anatolian Turks and other pops. Dodecad K12b is an old admixture calculator, but it still works well when it comes to estimating the amount of East Eurasian admixture.
https://docs.google.com/spreadsheets/d/1rNnXVwnlj-dOnb6bfVWgd2GPxvlQyGPL9fE9NZklvvs/htmlview#gid=0

Yes thats correct sorry. I meant kurds and Zazas

Straboo
09-10-2022, 01:12 PM
45% is not average, just saying it can go up to that much

Turkmens in Anatolia are different to Turkmens in Syria/Iraq, they're closer to the Turkish than to Syrian/Iraqi Turkmens, I can't say for Syrian Turkmens but Iraqi Turkmens look like a mix of Mesopotamian/Mesopotamian Arab + Azerbaijani

Local ancestry in Turkish is about 55%-75% usually, outliers being in the East and Northeast

Turks usually have around 10%-15% East Eurasian, but some can go to about 20% (Muğla, Bolu) and some can be undetectable (East, Northeast)

Would you say, if we left aside Kurdish/Zaza and Pontic-Laz areas, and focussed solely on classical anatolia, that local ancestry would drop even below 50% on average? I know the west has the highest "Turk" input in turkey

chum
09-10-2022, 01:23 PM
Central Anatolia and Central Blacksea would still be around 75% local, but the rest (Western Blacksea, Anatolian side of Marmara, Aegean, most of the Mediterranean, Northern Giresun) are generally 55% local with only some being below 50%

Alkaevli
09-10-2022, 02:12 PM
I was reffering to the population of Turkmenistan. Not pan national Turkmens. Maybe its correct to call Turkmenistan people Turkmens, but I call them Turkmenistanis to avoid confusion.
Turkmen in G25 represents Turkmens from Turkmenistan.



I can only rely on published articles that everyone has access to and can review, and in 2 recent publications that dealt with central asia, it was explicitly stated that the population of Turkmenistan is considerably less "Turkic" than Kyrgzistan, Uzbekistan and Kazakhstan and harbours a dominant east iranic ancestry (BMAC + middle steppe iirc). It seems that country in particular was more of a transit zone for turks and probay became turkic in culture and language much later than anatolia.
I dont think using Iranian_Fars to model (former) east iranians makes any sense. West and east iranian are 2 different genetic "provinces" iirc.

Some papers can be so awful that you can't give them a second look. Only those who know nothing about the region's history and medieval genetic profile can think modeling Turkmens as a mixture of modern Tajiks + Golden_Horde_Asian is a good idea.

1. Turkmens don't descend from Kazakhs who emerged as a nation in the 15th century from an amalgam of Turkic and Mongol tribes. They don't descend from the Golden Horde either, the Golden Horde postdates the first recorded presence of the Oghuz in the region by 5-6 centuries. They descend from the medieval Oghuz population who lived between the Caspian and Aral seas (AD 750-1050) and most likely resembled other Turkic populations (Karluk, Kimak etc.) of medieval Kazakhstan. So comparing Turkmens with modern Kazakhs or Kyrgyz is meaningless.

2. Modern Tajiks don't represent the pre-Turkic Iranic populations of the region. We have samples related to Sogdia (some "Uigur" individuals from Mongolia and Late Antiquity Otyrar), they have negligible levels of East Eurasian ancestry and resemble Yaghnobis, which means modern Tajiks have Turkic and/or Mongol ancestry in varying degrees.

3. Modern Turkmens shift away from the medieval Turkic cluster (Kazakhstan, AD 650-1200) in the direction of Iranian plateau populations, not Tajiks. It isn't rocket science, it is quite clear that they are West Iranian-shifted rather than East Iranian (including West Iranian-speaking Tajiks) when medieval Turkic populations of KAZ are used as proxy.
https://abload.de/img/adsizl2fyy.png

The medieval Turkic populations of Kazakhstan can be modeled as East Iranic + Early Turkic, so Turkmens have East Iranic ancestry from those medieval Turkic populations. Their post-medieval non-Turkic ancestry is mainly Iranian plateau-like. You can trust those published papers of course and think Turkmens are Tajik + Golden_Horde_Asian. Yeah sure.

Now, if you spare us of your classic "X Turkic population is language shifted Y" stories, we want to discuss the overall genetic profile of the Turkic population that contributed to the gene pool of medieval Anatolia.

Liquid
09-10-2022, 08:25 PM
These samples are dated to the medieval period, which spans the pre and post-Turkic periods, I expect they belong to the post-Turkic period, what strikes me is that these people are not very iranic at all, their Ganj Dareh levels are fairly typical for the Byzantine period and before. We might need to rethink the idea that the Turks who first arrived were like the Turkmens found in Iran today, i.e. heavily admixed with the local Persians, but were just orientals influenced by the culture of Iran.

Perhaps they were almost purely Far Eastern, like this statue from the Seljuk period found in Isfahan Iran or the Islamic hadiths, describing Turks with flat faces?

51162

51163

Just for a rudimentary pre and post comparison only:

Target: TUR_Aegean_Mugla_Capalibag_Mdv
Distance: 1.4834% / 0.01483419 | R5P
37.8 TUR_Tepecik_Ciftlik_N
20.6 Pontic_Steppe_Yamnaya_Pastoralist_Yamnaya_RUS_Sama ra
17.0 Kura-Araxes_ARM_Kaps
16.0 Amur_Basin_Hunter-Gatherer_RUS_Devils_Gate_Cave_N
8.6 Iran_Neolithic_Farmer_IRN_Ganj_Dareh_N

Target: TUR_Aegean_Mugla_Stratonikeia_Byz
Distance: 1.3567% / 0.01356742 | R5P
51.2 TUR_Tepecik_Ciftlik_N
19.2 Kura-Araxes_ARM_Kaps
13.0 Pontic_Steppe_Yamnaya_Pastoralist_Yamnaya_RUS_Sama ra
10.6 Levant_Neolithic_Farmer_Levant_PPNB
6.0 Iran_Neolithic_Farmer_IRN_Ganj_Dareh_N

At first glance it looks like a Steppe-Caucasian population might be responsible for the difference here.

J Man
09-10-2022, 08:34 PM
Turkmen in G25 represents Turkmens from Turkmenistan.



Some papers can be so awful that you can't give them a second look. Only those who know nothing about the region's history and medieval genetic profile can think modeling Turkmens as a mixture of modern Tajiks + Golden_Horde_Asian is a good idea.

1. Turkmens DON'T descend from Kazakhs who emerged as a nation in the 15th century from an amalgam of Turkic and Mongol tribes. They don't descend from the Golden Horde either, the Golden Horde postdates the first recorded presence of the Oghuz in the region by 5-6 centuries. They descend from the medieval Oghuz population who lived between the Caspian and Aral seas (AD 750-1050) and most likely resembled other Turkic populations (Karluk, Kimak etc.) of medieval Kazakhstan. So comparing Turkmens with modern Kazakhs or Kyrgyz is meaningless.

2. Modern Tajiks don't represent the pre-Turkic Iranic populations of the region. We have samples related to Sogdia (some "Uigur" individuals from Mongolia and Late Antiquity Otyrar), they have negligible levels of East Eurasian ancestry and resemble Yaghnobis, which means modern Tajiks have Turkic and/or Mongol ancestry in varying degrees.

3. Modern Turkmens shift away from the medieval Turkic cluster (Kazakhstan, AD 650-1200) in the direction of Iranian plateau populations, not Tajiks. It isn't rocket science, it is quite clear that they are West Iranian-shifted rather than East Iranian (including West Iranian-speaking Tajiks) when medieval Turkic populations of KAZ are used as proxy.
https://abload.de/img/adsizl2fyy.png

The medieval Turkic populations of Kazakhstan can be modeled as East Iranic + Early Turkic, so Turkmens have East Iranic ancestry from those medieval Turkic populations. Their post-medieval non-Turkic ancestry is mainly Iranian plateau-like. You can trust those published papers of course and think Turkmens are Tajik + Golden_Horde_Asian. Yeah sure.

Now, if you spare us of your classic "X Turkic population is language shifted Y" stories, we want to discuss the overall genetic profile of the Turkic population that contributed to the gene pool of medieval Anatolia.

Is there any archeological information out there about the TUR_Aegean_Mugla_Capalibag_Mdv samples?

Liquid
09-10-2022, 08:43 PM
Ok, there you go:

Target: TUR_Aegean_Mugla_Capalibag_Mdv
Distance: 1.2112% / 0.01211212 | R4P
56.0 TUR_Aegean_Mugla_Stratonikeia_Byz
19.2 RUS_Late_Xiongnu
14.2 Kura-Araxes_ARM_Kaps
10.6 Yamnaya_RUS_Caucasus

Liquid
09-10-2022, 09:09 PM
I removed TUR_Aegean_Mugla_Stratonikeia_Byz from TUR_Aegean_Mugla_Capalibag_Mdv, this is what I get:

Distance to: Byzantine_removed_medieval_mugla
0.03686673 MNG_Early_Med_Uigur
0.04698201 KAZ_Turk
0.04838349 KGZ_Chilpek_IA
0.04856051 TUR_Ottoman
0.05906535 KGZ_Alai_Nura_Antiquity_2
0.05942868 RUS_Nomad_MA
0.06094798 RUS_Zevakino_Chilikta_IA
0.06230625 KAZ_Kipchak
0.06632401 CHN_Gongnaisi_IA2
0.06679163 KAZ_Karluk <----------------------

Distance to: Byzantine_removed_medieval_mugla
0.04280909 Uzbek
0.05796501 Turkmen_Uzbekistan
0.06416230 Turkmen
0.07334432 Hazara_Afghanistan
0.07540351 Bashkir
0.08097134 Hazara
0.08104042 Uygur
0.08343630 Nogai
0.08399158 Tatar_Crimean_steppe
0.08584935 Tatar_Siberian

So more like Modern Uzbek and Early Medieval Uyghur rather than Kaz_Karluk?

Byzantine_removed_medieval_mugla,0.073793955,-0.0929967,0.011927382,0.007206873,-0.050485291,0.000454291,0.0129959,0.002916936,-0.025824418,-0.021201973,-0.010627364,0.004457045,-0.004611227,-0.005293127,0.013697636,-0.00941316,-0.014838091,0.003819664,-0.00160550,-0.000679455,-0.002764345,-0.000114109,0.001345391,-0.009562727,0.000364427

Tujue19
09-10-2022, 09:41 PM
I removed TUR_Aegean_Mugla_Stratonikeia_Byz from TUR_Aegean_Mugla_Capalibag_Mdv, this is what I get:

Distance to: Byzantine_removed_medieval_mugla
0.03686673 MNG_Early_Med_Uigur
0.04698201 KAZ_Turk
0.04838349 KGZ_Chilpek_IA
0.04856051 TUR_Ottoman
0.05906535 KGZ_Alai_Nura_Antiquity_2
0.05942868 RUS_Nomad_MA
0.06094798 RUS_Zevakino_Chilikta_IA
0.06230625 KAZ_Kipchak
0.06632401 CHN_Gongnaisi_IA2
0.06679163 KAZ_Karluk <----------------------

Distance to: Byzantine_removed_medieval_mugla
0.04280909 Uzbek
0.05796501 Turkmen_Uzbekistan
0.06416230 Turkmen
0.07334432 Hazara_Afghanistan
0.07540351 Bashkir
0.08097134 Hazara
0.08104042 Uygur
0.08343630 Nogai
0.08399158 Tatar_Crimean_steppe
0.08584935 Tatar_Siberian

So more like Modern Uzbek and Early Medieval Uyghur rather than Kaz_Karluk?

Byzantine_removed_medieval_mugla,0.073793955,-0.0929967,0.011927382,0.007206873,-0.050485291,0.000454291,0.0129959,0.002916936,-0.025824418,-0.021201973,-0.010627364,0.004457045,-0.004611227,-0.005293127,0.013697636,-0.00941316,-0.014838091,0.003819664,-0.00160550,-0.000679455,-0.002764345,-0.000114109,0.001345391,-0.009562727,0.000364427

How does this refute the fact that the Turkic nomads incoming to Anatolia could have Turkmen-like genetic make-up? After all, Turkmens and Uzbeks aren't that genetically distant from each other at all.

Distance to: Turkmen_Uzbekistan
0.01481804 Turkmen
0.06060090 Uzbek
0.06194977 Tatar_Crimean_steppe
0.07660501 Tajik_Hisor
0.07715037 Tajik_Ayni
0.08567239 Tatar_Lipka


Distance to: Turkmen
0.01481804 Turkmen_Uzbekistan
0.06046915 Tatar_Crimean_steppe
0.06810945 Tajik_Hisor
0.06835747 Tajik_Ayni
0.06974033 Uzbek
0.08030169 Tatar_Lipka

Alkaevli
09-10-2022, 09:50 PM
We might need to rethink the idea that the Turks who first arrived were like the Turkmens found in Iran today, i.e. heavily admixed with the local Persians, but were just orientals influenced by the culture of Iran.
They weren't, and we don't think they were Iranian Turkmen-like. Read the posts carefully. The migration lasted two centuries and later ones might have been eastern/Iranian-admixed, which can explain why modern Turkmens produce better fits in G25 models.


Perhaps they were almost purely Far Eastern, like this statue from the Seljuk period found in Isfahan Iran or the Islamic hadiths, describing Turks with flat faces?
So, you mean the Oghuz were an island of 100% East Eurasian population in the sea of West Eurasian-admixed neighboring Turkic populations (Pechenegs, Kipchaks, Kimaks, Karluks/Karakhanids). Brilliant indeed.
https://abload.de/img/downloadtaeuo.png

40-50% East Eurasian admixture (generic medieval Kazakhstan Turkic genetic profile, 800-1200 CE) is enough to make them "flat faced" in the eyes of Arabs with protruding West Eurasian facial features. Astrakhan Nogais with ~50% EA admixture are pretty "flat faced" as well from the West Eurasian perspective. Early Medieval Western Khaganate samples (DA89, DA93, DA94 etc.) and the Karakaba Turkic samples near the Altai Mountains show the same pattern, West-East Eurasian mix.

16th century Ottoman miniatures depicting the battles between the Ottomans and the Mamluks. They must have been purely East Eurasian too.
https://abload.de/img/21pepm.jpg
https://abload.de/img/1vcd9a.png



These samples are dated to the medieval period, which spans the pre and post-Turkic periods, I expect they belong to the post-Turkic period, what strikes me is that these people are not very iranic at all, their Ganj Dareh levels are fairly typical for the Byzantine period and before.
51162


Why are you using Kura-Araxes if you want to compare their Iran_N levels? Kura-Araxes itself has Iran_N-related admixture.

Comparison of Byzantine Muğla and Menteşe-Ottoman (Turkic) Muğla

Eastern component (CHG+Iran_N) slightly increases
Southern (Levant) component decreases
Anatolian (Barcin-related) component decreases
Northern (EHG) component increases
Far Eastern component increases
https://abload.de/img/adsizpaee1.png

Anyone with basic math skills can see from this table alone that the Turkic population that contributed to the gene pool of Muğla cannot be fully East Eurasian.

When we extract Byzantine Muğla from Turkic Muğla we get Central Asian populations, not East Asian ones.
https://abload.de/img/adsizu7cf7.png

Tujue19
09-10-2022, 10:03 PM
So, you mean the Oghuz were an island of 100% East Eurasian population in the sea of West Eurasian-admixed neighboring Turkic populations (Pechenegs, Kipchaks, Kimaks, Karluks/Karakhanids). Brilliant indeed.
https://abload.de/img/downloadtaeuo.png

That cliché claim is very much connected to another nonsense I often come across on Quora, YouTube comments and various Internet forums. It is that Yakuts are the original, unmixed, purest Turkic people and that the Turkics are indigenous to Yakutia or Northeast Asia in general. This misconception stems from the debunked Altaic hypothesis which puts the Turkic Urheimat at somewhere inbetween Manchuria and Yakutia, with the implication that the Proto-Turkic speakers must be a 100% Mongoloid people from the Far East. The indigenous ethnic groups of Yakutia were various Paleo-Siberian peoples most of which went extinct due to being assimilated by Yakuts and Evenks. The one that hasn’t gone extinct yet is Yukaghir people whose language is claimed by some linguists to be Para-Uralic. They’ve been living there since the Neolithic Age.
The Yukaghirs: A nomadic Siberian tribe on the brink of extinction (https://www.rbth.com/society/2014/01/17/the_yukaghirs_a_nomadic_siberian_tribe_on_the_brin g_of_extinction_33149.html)
Most of Yukaghirs had been gradually assimilated by Evenks and Yakuts. Tungusic Evenks arrived in Yakutia before the Turkics did. The name Yakut is an Evenki word meaning “alien”. The culture, clothing, language, faces of the newly-arrived Turkics were so alien to the medieval Evenks that they called the Turkics “yakut”. Yakut people don’t call themselves Yakut of course, they refer to themselves as Sakha. It’s a well-established fact that the Turkic Proto-Yakuts weren’t native to Yakutia. Yakut legends put their homeland near the western shores of Lake Baikal in south-central Siberia, an area now occupied by the Mongolic Buryats. The two peoples must have interacted extensively, as roughly one-third (33%) of the Yakut vocabulary is of Mongolic origin. Relations were not always cordial; Yakuts tell stories of their ancestors being driven into the northern forests by the Buryats. Modern Yakuts are quite far from being “the original Turks” since they absorbed many non-Turkic tribes of Mongolic, Tungusic and indigenous Paleo-Siberian origin as evidenced by the cultural, linguistic and genetic studies on Yakuts.
The Yakut (Sakha) Migration to Central Siberia (http://www.geocurrents.info/place/russia-ukraine-and-caucasus/siberia/the-yakut-sakha-migration-to-central-siberia)
(The title of this article is erroneous though. As explained in the article itself, Proto-Yakuts migrated from southern central Siberia to Yakutia which is in northeastern Siberia.)
Yakut language has many influences from Yukaghir, Tungusic and especially from Mongolic. Among all Turkic languages, Yakut language is the most ‘Mongolicised’ one due to Yakuts having been heavily mixed with Mongolic peoples (especially Buryats).

Alkaevli
09-10-2022, 10:18 PM
That cliché claim is very much connected to another nonsense I often come across on Quora, YouTube comments and various Internet forums. It is that Yakuts are the original, unmixed, purest Turkic people and that the Turkics are indigenous to Yakutia or Northeast Asia in general. This misconception stems from the debunked Altaic hypothesis which puts the Turkic Urheimat at somewhere inbetween Manchuria and Yakutia, with the implication that the Proto-Turkic speakers must be a 100% Mongoloid people from the Far East. The indigenous ethnic groups of Yakutia were various Paleo-Siberian peoples most of which went extinct due to being assimilated by Yakuts and Evenks. The one that hasn’t gone extinct yet is Yukaghir people whose language is claimed by some linguists to be Para-Uralic. They’ve been living there since the Neolithic Age.
The Yukaghirs: A nomadic Siberian tribe on the brink of extinction (https://www.rbth.com/society/2014/01/17/the_yukaghirs_a_nomadic_siberian_tribe_on_the_brin g_of_extinction_33149.html)
Most of Yukaghirs had been gradually assimilated by Evenks and Yakuts. Tungusic Evenks arrived in Yakutia before the Turkics did. The name Yakut is an Evenki word meaning “alien”. The culture, clothing, language, faces of the newly-arrived Turkics were so alien to the medieval Evenks that they called the Turkics “yakut”. Yakut people don’t call themselves Yakut of course, they refer to themselves as Sakha. It’s a well-established fact that the Turkic Proto-Yakuts weren’t native to Yakutia. Yakut legends put their homeland near the western shores of Lake Baikal in south-central Siberia, an area now occupied by the Mongolic Buryats. The two peoples must have interacted extensively, as roughly one-third (33%) of the Yakut vocabulary is of Mongolic origin. Relations were not always cordial; Yakuts tell stories of their ancestors being driven into the northern forests by the Buryats. Modern Yakuts are quite far from being “the original Turks” since they absorbed many non-Turkic tribes of Mongolic, Tungusic and indigenous Paleo-Siberian origin as evidenced by the cultural, linguistic and genetic studies on Yakuts.
The Yakut (Sakha) Migration to Central Siberia (http://www.geocurrents.info/place/russia-ukraine-and-caucasus/siberia/the-yakut-sakha-migration-to-central-siberia)
(The title of this article is erroneous though. As explained in the article itself, Proto-Yakuts migrated from southern central Siberia to Yakutia which is in northeastern Siberia.)
Yakut language has many influences from Yukaghir, Tungusic and especially from Mongolic. Among all Turkic languages, Yakut language is the most ‘Mongolicised’ one due to Yakuts having been heavily mixed with Mongolic peoples (especially Buryats).
Yakuts are definitely far from representing Proto-Turks, but it doesn't mean Proto-Turkic wasn't East Eurasian. It is likely that they were Slab Grave-like. I personally think that Proto-Turkic was spoken by a population mixture of Altai_IA-like and Slab Grave-like components.

Tujue19
09-10-2022, 10:36 PM
Yakuts are definitely far from representing Proto-Turks, but it doesn't mean Proto-Turkic wasn't spoken by a Slab Grave-like population. It is highly likely. I personally think that Proto-Turkic was spoken by a population mixture of Altai_IA-like and Slab Grave-like components.

https://www.researchgate.net/publication/320552611_A_Comparative_Analysis_of_Chinese_Histor ical_Sources_and_Y-DNA_Studies_with_Regard_to_the_Early_and_Medieval_ Turkic_Peoples

Well I regard that unlikely, that Proto-Turkics were 100% Mongoloid a.k.a East-Eurasian.

In its conclusion, the authors state:
“The analysis of genetic survey data on the Turkic peoples also allows us to speculate on the Turkic Urheimat. We suggest that it was a geographical region where the carriers of haplogroups C2, N, Q and R1a1 could intermix, since these haplogroups are carried by various past and modern-day Turkic peoples in eastern Inner Asia and the Xiongnu. It has been suggested that the early Turkic peoples probably had contact with Indo-European, Uralic, Yeniseian, and Mongolic groups in their formative period (Golden 2006: 139). As non-linguists, we are unqualified to discuss the origin of the Turkic languages. However, drawing on the findings of studies, we are inclined to think that certain similarities that exist between the Turkic languages and the Mongolic, Tungusic and Uralic languages are at least partly associated with haplogroups C2 and N, among others. More specifically, we conjecture that the Turkic languages came into existence as a result of the fusion of Uralic groups (characterized by a high frequency of haplogroup N subclades) and Pre-Proto-Mongolic groups (characterized by a high frequency of haplogroup C2) who also merged with other linguistic groups, including Yeniseian speakers (characterized by a high frequency of haplogroup Q like the Kets) and Indo-European speakers (characterized by a high frequency of haplogroups R1a1). The best candidate for the TURKIC URHEIMAT would then be NORTHERN AND WESTERN MONGOLIA AND TUVA, where all these haplogroups could have intermingled, rather than eastern and southern Mongolia OR THE YENISEI RIVER AND THE ALTAI MOUNTAINS regions in Russia.”

The percentages of haplogroup O3 and C2 among Naimans (Mongolians) were 61.2% and 63% respectively. Both are too high compared to Turkic peoples. See Table 2 in the paper. Even the most Mongoloid-looking Turkic peoples have much lower percentages of O3 and C2.
51166

People are acting like the DNA analysis conducted on the skeletal remains of Köktürks, Karakhanids, early Kypchaks, medieval Yakuts don’t exist? People have to explain how today’s most Mongoloid-looking Turkic peoples (the Siberian ones plus Kazakhs and Kyrgyzes) could drastically lost the high frequencies of O3 and C2 if the hypothetical 100% Mongoloid ancestral Turkics were genetically similar to East Asians/Mongolians. Even 100% Mongoloid Yugurs have less O3 and less C2 than Naimans.

51167

Besides that, after Donghu people ("Eastern Barbarians") were defeated by the Xiongnu around 209 BC, they split into two groups. The northern Donghu became the Wuhuan while the southern Donghu living around modern Liaoning became the Xianbei. According to the Book of the Later Han, “the language and culture of the Xianbei are the same as the Wuhuan”. They were the common ancestors of the Proto-Mongolic people and Para-Mongolic peoples such as the Tabgach, the Nine Tatars, the Thirty Tatars, the Khitan.

Dingling and Khyagas were the Chinese renditions of the names of two ancient Turkic peoples that are the Tegreg and the Yenisei Kyrgyz.

As shown on the map above, the Xiongnu were eastern/southeastern neighbours of ancient Turkic-speaking peoples so it’s safe to assume they were more Mongoloid than Turkic peoples were. However, even the Xiongnu had a significant Caucasoid admixture as demonstrated by a genetic analysis of 62 skeletons excavated from a 2000 year-old Xiongnu necropolis in Mongolia (https://pubmed.ncbi.nlm.nih.gov/12858290/).

So, the cliché claim that ancient Turkic peoples or Proto-Turkics were 100% Mongoloid doesn’t make any sense at all.

Alkaevli
09-10-2022, 10:55 PM
Comparison of Byzantine Muğla and Menteşe-Ottoman (Turkic) Muğla

Eastern components (CHG+Iran_N) slightly increases
Southern (Levant) component decreases
Anatolian (Barcin-related) component decreases
Northern (EHG) component increases
Far Eastern component increases
https://abload.de/img/adsizpaee1.png


Dodecad K12b comparison (Caucasus here=/=CHG)
https://abload.de/img/muglaadmixtureengwxcfe.png

bovefex
09-11-2022, 06:24 AM
They weren't, and we don't think they were Iranian Turkmen-like. Read the posts carefully. The migration lasted two centuries and later ones might have been eastern/Iranian-admixed that makes modern Turkmen average a better fit in G25 models.


So, you mean the Oghuz were an island of 100% East Eurasian population in the sea of West Eurasian-admixed neighboring Turkic populations (Pechenegs, Kipchaks, Kimaks, Karluks/Karakhanids). Brilliant indeed.
https://abload.de/img/downloadtaeuo.png

1. The hadiths predate the Seljuk period.

2. 40-50% East Eurasian admixture (generic medieval Kazakhstan Turkic genetic profile, 800-1200 CE) is enough to make them "flat faced" in the eyes of Arabs with protruding facial features. Astrakhan Nogais with ~50% EA admixture are pretty "flat faced" as well from the West Eurasian perspective. Early Medieval Western Khaganate samples (DA89, DA93, DA94 etc.) and the Karakaba Turkic samples near the Altai Mountains show the same pattern, West-East Eurasian mix.

16th century Ottoman miniatures depicting the battles between the Ottomans and the Mamluks. They must have been purely East Eurasian too.
https://abload.de/img/21pepm.jpg
https://abload.de/img/1vcd9a.png



Why do you use Kura-Araxes if you want to compare their Iran_N levels? Kura-Araxes itself has Iran_N-related admixture.

Comparison of Byzantine Muğla and Menteşe-Ottoman (Turkic) Muğla

Eastern components (CHG+Iran_N) slightly increases
Southern (Levant) component decreases
Anatolian (Barcin-related) component decreases
Northern (EHG) component increases
Far Eastern component increases
https://abload.de/img/adsizpaee1.png

Anyone with basic math skills can see from this table alone that the Turkic population that contributed to the gene pool of Muğla cannot be fully East Eurasian.

When we extract Byzantine Muğla from Turkic Muğla we get Central Asian populations, not East Asian ones.
https://abload.de/img/adsizu7cf7.png

How is it possible that CHG increases during the Turkic migrations? Is it an increase in Iran_N misinterpreted as CHG?

Alkaevli
09-11-2022, 07:39 AM
How is it possible that CHG increases during the Turkic migrations? Is it an increase in Iran_N misinterpreted as CHG?
Possible. Just in case, I refer to them as the eastern component.

chum
09-11-2022, 07:44 AM
How is it possible that CHG increases during the Turkic migrations? Is it an increase in Iran_N misinterpreted as CHG?

Could be via Steppe and/or more Inner Anatolian ancestry

bovefex
09-11-2022, 07:45 AM
Possible. Just in case, I refer to them as the eastern component.

I have also noticed that when a G25 ancients model doesn't the proper Turkic proxy, it tries to substitute for it with one East-Asian-rich pop and one (possibly Steppe-rich, don't remember) Caucasian pop. I was wondering whether some Caucasians could have migrated along with the Turks, but guess it's just funny behavior from G25. Maybe it is real, but guess we have to wait for more ancient samples.

bovefex
09-11-2022, 07:47 AM
Could be via Steppe and/or more Inner Anatolian ancestry

Don't think Inner Anatolians were that much more CHG-admixed compared to Western Anatolians that they were able to not only counteract the effects of the Turkic-admix, but even increase the CHG.

bovefex
09-11-2022, 07:48 AM
Possible. Just in case, I refer to them as the eastern component.

By the way, do you guys have a list of the samples from Anatolia (and environs) from Southern Arc you wouldn't mind sharing?

chum
09-11-2022, 07:53 AM
Don't think Inner Anatolians were that much more CHG-admixed compared to Western Anatolians that they were able to not only counteract the effects of the Turkic-admix, but even increase the CHG.

Then it's probably via Steppe, the EHG also increases

Alkaevli
09-11-2022, 07:53 AM
By the way, do you guys have a list of the samples from Anatolia (and environs) from Southern Arc you wouldn't mind sharing?
Do you mean their G25 coordinates? They have been added to David's spreadsheet. If you mean the supp info (archaelogical context, dates, haplogroups etc.) I'm still using the .xls file provided by the paper for that. But we are planning to create our own list with new labels.

Straboo
09-11-2022, 11:40 AM
Now, if you spare us of your classic "X Turkic population is language shifted Y" stories, we want to discuss the overall genetic profile of the Turkic population that contributed to the gene pool of medieval Anatolia.

Not very freindly are you

I just wanted to learn what happened. Its not my fault if according to you I consumed false information and nobody told me it was false, or explain it in a way that is easy to understand for newcomers.

Gentica277282
09-11-2022, 12:29 PM
These samples are dated to the medieval period, which spans the pre and post-Turkic periods, I expect they belong to the post-Turkic period, what strikes me is that these people are not very iranic at all, their Ganj Dareh levels are fairly typical for the Byzantine period and before. We might need to rethink the idea that the Turks who first arrived were like the Turkmens found in Iran today, i.e. heavily admixed with the local Persians, but were just orientals influenced by the culture of Iran.

Perhaps they were almost purely Far Eastern, like this statue from the Seljuk period found in Isfahan Iran or the Islamic hadiths, describing Turks with flat faces?

51162

51163

Just for a rudimentary pre and post comparison only:

Target: TUR_Aegean_Mugla_Capalibag_Mdv
Distance: 1.4834% / 0.01483419 | R5P
37.8 TUR_Tepecik_Ciftlik_N
20.6 Pontic_Steppe_Yamnaya_Pastoralist_Yamnaya_RUS_Sama ra
17.0 Kura-Araxes_ARM_Kaps
16.0 Amur_Basin_Hunter-Gatherer_RUS_Devils_Gate_Cave_N
8.6 Iran_Neolithic_Farmer_IRN_Ganj_Dareh_N

Target: TUR_Aegean_Mugla_Stratonikeia_Byz
Distance: 1.3567% / 0.01356742 | R5P
51.2 TUR_Tepecik_Ciftlik_N
19.2 Kura-Araxes_ARM_Kaps
13.0 Pontic_Steppe_Yamnaya_Pastoralist_Yamnaya_RUS_Sama ra
10.6 Levant_Neolithic_Farmer_Levant_PPNB
6.0 Iran_Neolithic_Farmer_IRN_Ganj_Dareh_N

At first glance it looks like a Steppe-Caucasian population might be responsible for the difference here.

Isn’t tepecik absorbing some of that Iranian. Can you try a model with barcin n

Kelmendasi
09-11-2022, 01:06 PM
As shown on the map above, the Xiongnu were eastern/southeastern neighbours of ancient Turkic-speaking peoples so it’s safe to assume they were more Mongoloid than Turkic peoples were. However, even the Xiongnu had a significant Caucasoid admixture as demonstrated by a genetic analysis of 62 skeletons excavated from a 2000 year-old Xiongnu necropolis in Mongolia (https://pubmed.ncbi.nlm.nih.gov/12858290/).

So, the cliché claim that ancient Turkic peoples or Proto-Turkics were 100% Mongoloid doesn’t make any sense at all.
I apologise for the off topic question, however, what is your opinion on the core or bulk of the Xiongnu being themselves Turkic-speakers?

Avraam Kyriakidis
09-11-2022, 01:42 PM
I have to agree here that Proto-Turkic were probably slab-grave like.Turks of Anatolia probably arrived kaz kharamanid like i would say but we need def samples about it.

Liquid
09-18-2022, 06:48 AM
Isn’t tepecik absorbing some of that Iranian. Can you try a model with barcin n

Tepecik was clearly already there in the early neolithic, right there in the middle of Turkey, I don't like to exclude it for Near Easterners.

Liquid
09-18-2022, 06:58 AM
Why are you using Kura-Araxes if you want to compare their Iran_N levels? Kura-Araxes itself has Iran_N-related admixture.

https://abload.de/img/adsizu7cf7.png

Because the Kura-Araxes people had like 4000 years to move around, obviously.




Anyone with basic math skills can see from this table alone that the Turkic population that contributed to the gene pool of Muğla cannot be fully East Eurasian.

https://abload.de/img/adsizu7cf7.png

Is it simple A-B when treating these coordinates? Because that's not what I did. And another question for your persona, do you think it's immediately obvious for the non-acquainted?

Thanks in advance.

Alkaevli
09-18-2022, 08:47 PM
Because the Kura-Araxes people had like 4000 years to move around, obviously.
They represent a relatively mixed population, and if you want to compare their Iran_N levels Kura_Araxes_Kaps must not be there.



Is it simple A-B when treating these coordinates? Because that's not what I did.
That doesn't give us the exact population but closest thing available. Muğla_Mdv without local Byzantine admixture is not 100% or predominantly East Eurasian-like.



And another question for your persona, do you think it's immediately obvious for the non-acquainted?

What do you mean?

Anyway, adding Alan_MA (or modern Chechens) significantly improves the models.

Sorted from best to worst.
https://abload.de/img/mdv8fex6.png

It works well with modern southwestern Turks too.
https://abload.de/img/mdv2qyd1m.png

Liquid
10-20-2022, 05:39 AM
Well these elements appear to work in a recent, non-modern calculator, for Turkey and its surroundings, like for when the Turks first entered the scene:

TUR_Marmara_Balikesir_Byz (Roman/Greek Byzantine)
TUR_SE_Gaziantep_Byz (Levantine/Armenian-like)
RUS_Maykop_Novosvobodnaya (Bronze Age/Georgian-like)
KAZ_Karluk (Middle Ages Turk)
IRN_Hasanlu_IA (Kurdish/Persian/Xerces... still a really good match for the modern variants)
RUS_Alan_MA

Adding RUS_Alan_MA helps slightly with some of the fits, no idea why. Georgians and Turkish_Trabzon like RUS_Alan_MA and RUS_Maykop_Novosvobodnaya together, but RUS_Maykop_Novosvobodnaya is a must for them so it should be included for the general Turkish populace. All these samples when used together produce good fits for all others in the region like Armenians, Assyrians, Cappadocians, Kurds etc. I haven't looked at how the Circassians could've contributed, perhaps the Maykop and Alans together...

Hmm, I don't know why Karluk has appeared there but they seem to be pretty Alan-like. Perhaps it needs something non-East Asian in Karluk that is lacking in the others:

Target: Circassian
Distance: 1.4258% / 0.01425771
59.4 RUS_Alan_MA
27.0 RUS_Maykop_Novosvobodnaya
10.0 KAZ_Karluk
3.6 TUR_Marmara_Balikesir_Byz

Are there any Byzantine or MA North Caucasians that can be used in lieau of the really early RUS_Maykop_Novosvobodnaya? Anyone?