Global25 automated nMonte for South/Central Asian members

[Coldmountains]

Anyone wanna offer up a model for pashtuns would be appreciated. Pakistani and Afghan. These guys can sometimes superficially appear like other NW South-Asians but as far as I know they often have different ancestral sources? The difference seems to be Seh Gabi oriented but I need yall input.

Target: Pashtun
Distance: 2.5058% / 0.02505807
38.2	KAZ_Kangju
37.2	PAK_Saidu_Sharif_H_o
24.6	IRN_Seh_Gabi_C


Target: Yusufzai
Distance: 2.7258% / 0.02725809
37.4	KAZ_Kangju
36.2	PAK_Saidu_Sharif_H_o
26.4	IRN_Seh_Gabi_C

Target: Tarkalani
Distance: 2.8948% / 0.02894760
39.2	KAZ_Kangju
33.0	PAK_Saidu_Sharif_H_o
27.8	IRN_Seh_Gabi_C

Target: Uthmankhel
Distance: 2.6889% / 0.02688939
40.2	KAZ_Kangju
33.2	PAK_Saidu_Sharif_H_o
26.6	IRN_Seh_Gabi_C

Target: Pashtun:Pashtun2_20Af
Distance: 2.8330% / 0.02832990
32.0	IRN_Seh_Gabi_C
24.6	Saka_Tian_Shan
23.8	PAK_Saidu_Sharif_H_o
19.6	KAZ_Kangju

I would not use Kangju who seem to be Sogdians and not match Pashtuns in terms of autosomal and uniparental markers (especially Y-DNA). So far we don't have really good IA references for Pashtuns. Swat_IA is too eastern for most Pashtuns, TKM_IA too northern and western and Kangju too late and too northwestern again. There is also significant variation among modern-day Pashtuns from almost Iranian Sistani-like Kandhari to strongly Swat_IA/Kalash/Dardic shifted East Afghan and KPK Pashtuns. In the north and even sometimes in regions like Kandahar some Pashtuns show extra Turkmen/Uzbek/Pamiri-like admix. Proto-Pashtuns would rather resemble Kandahari here minus the extra Turkic and Persian/Qizilbash admix some of them show. All of them definitely have Iran_Chalc-like on top of BMAC-like admix (especially Kandahari and Tajik-admixed Pashtuns) but it overlaps so much with their BMAC-like ancestry that it is hard to model the extra Iranian plateau ancestry after BMAC. So I used Syr_Ebla as an extreme reference for more "recent" West Asian ancestry

Target: Pashtun_Kandahar_AVG
Distance: 1.7057% / 0.01705691
42.6 TKM_Gonur1_BA
25.0 RUS_Sintashta_MLBA
19.4 IRN_Shahr_I_Sokhta_BA2
8.6 SYR_Ebla_EMBA
4.4 MNG_Early_Xiongnu_SKT007

Target: Iranian_Sistan
Distance: 2.2952% / 0.02295198
45.4 TKM_Gonur1_BA
19.6 IRN_Shahr_I_Sokhta_BA2
16.2 SYR_Ebla_EMBA
15.6 RUS_Sintashta_MLBA
3.2 MNG_Early_Xiongnu_SKT007

For comparison Yusufzai Pashtuns with rich Swat_IA-like admix. These kind of types can also be found in some regions of East Afghanistan like Nangarhar/Jalalabad for example.

Target: Yusufzai_AVG
Distance: 1.2210% / 0.01221015
33.8 IRN_Shahr_I_Sokhta_BA2
30.2 TKM_Gonur1_BA
21.8 RUS_Sintashta_MLBA
6.4 SYR_Ebla_EMBA
4.6 KGZ_Aigyrzhal_BA
3.2 MNG_Early_Xiongnu_SKT007

And for comparison some Kabuli and Kunduz Pashtuns with extra Pamiri-like and/or East Asian-like admix

Target: Kabul_scaled
Distance: 2.4036% / 0.02403591
34.4 TKM_Gonur1_BA
29.6 RUS_Sintashta_MLBA
22.8 IRN_Shahr_I_Sokhta_BA2
9.0 MNG_Early_Xiongnu_SKT007
4.2 SYR_Ebla_EMBA

Target: Pashtunashtun2_20Af
Distance: 2.3844% / 0.02384364
38.0 TKM_Gonur1_BA
23.2 RUS_Sintashta_MLBA
20.2 IRN_Shahr_I_Sokhta_BA2
10.2 MNG_Early_Xiongnu_SKT007
8.4 SYR_Ebla_EMBA

[pegasus]

I would not use Kangju who seem to be Sogdians and not match Pashtuns in terms of autosomal and uniparental markers (especially Y-DNA). So far we don't have really good IA references for Pashtuns. Swat_IA is too eastern for most Pashtuns, TKM_IA too northern and western and Kangju too late and too northwestern again. Netherless there is significant variation among modern-day Pashtuns from almost Iranian Sistani-like Kandhari to strongly Swat_IA/Kalash/Dardic shifted East Afghan and KPK Pashtuns. In the north and even sometimes in regions like Kandahar some Pashtuns show extra Turkmen/Uzbek/Pamiri-like admix. Proto-Pashtuns would rather resemble Kandahari here minus the extra Turkic and Persian/Qizilbash admix some of them show. All of them definitely have Iran_Chalc-like on top of BMAC-like admix (especially Kandahari and Tajik-admixed Pashtuns) but it overlaps so much with their BMAC-like ancestry that it is hard to model the extra Iranian plateau ancestry after BMAC. So I used Syr_Ebla as an extreme reference for more "recent" West Asian ancestry

Target: Pashtun_Kandahar_AVG
Distance: 1.7057% / 0.01705691
42.6 TKM_Gonur1_BA
25.0 RUS_Sintashta_MLBA
19.4 IRN_Shahr_I_Sokhta_BA2
8.6 SYR_Ebla_EMBA
4.4 MNG_Early_Xiongnu_SKT007

Target: Iranian_Sistan
Distance: 2.2952% / 0.02295198
45.4 TKM_Gonur1_BA
19.6 IRN_Shahr_I_Sokhta_BA2
16.2 SYR_Ebla_EMBA
15.6 RUS_Sintashta_MLBA
3.2 MNG_Early_Xiongnu_SKT007

For comparison Yusufzai Pashtuns with rich Swat_IA-like admix. These kind of types can also be found in some regions of East Afghanistan like Nangarhar/Jalalabad for example.

Target: Yusufzai_AVG
Distance: 1.2210% / 0.01221015
33.8 IRN_Shahr_I_Sokhta_BA2
30.2 TKM_Gonur1_BA
21.8 RUS_Sintashta_MLBA
6.4 SYR_Ebla_EMBA
4.6 KGZ_Aigyrzhal_BA
3.2 MNG_Early_Xiongnu_SKT007

And for comparison some Kabuli and Kunduz Pashtuns with extra Pamiri-like and/or East Asian-like admix

Target: Kabul_scaled
Distance: 2.4036% / 0.02403591
34.4 TKM_Gonur1_BA
29.6 RUS_Sintashta_MLBA
22.8 IRN_Shahr_I_Sokhta_BA2
9.0 MNG_Early_Xiongnu_SKT007
4.2 SYR_Ebla_EMBA

Target: Pashtunashtun2_20Af
Distance: 2.3844% / 0.02384364
38.0 TKM_Gonur1_BA
23.2 RUS_Sintashta_MLBA
20.2 IRN_Shahr_I_Sokhta_BA2
10.2 MNG_Early_Xiongnu_SKT007
8.4 SYR_Ebla_EMBA

Kangju are clearly not ancestral but thus far they are the only profile which seems to work and even in qpAdm, 121,206 in the subset work very well. Using 3-4 left pops, they seem to be the favorite Kangju + Bustan_o2/Loebanr+Ebla , though I think that Kwarzemian lady (TUK 001) would work well in place of Ebla, since she has that shift already.


Most insanely difficult modern population to model on qpAdm are Kalash they kind of point the way of an admixed base population already present , since SIS2 fails as a base, they too also prefer Bustan_o2.


Kalash.DG
Uzbekistan_Bustan_BA_o2 : 52.6 +/- 6.5%
Turkmenistan_Namazga_CA_o.SG: 30.7 +/- 6.5%
Russia_Mezhovskaya_MLBA.SG: 16.8 +/- 4.6%
Tail prob: 0.45

[Coldmountains]

Anyone wanna offer up a model for pashtuns would be appreciated. Pakistani and Afghan. These guys can sometimes superficially appear like other NW South-Asians but as far as I know they often have different ancestral sources? The difference seems to be Seh Gabi oriented but I need yall input.

Target: Pashtun
Distance: 2.5058% / 0.02505807
38.2	KAZ_Kangju
37.2	PAK_Saidu_Sharif_H_o
24.6	IRN_Seh_Gabi_C


Target: Yusufzai
Distance: 2.7258% / 0.02725809
37.4	KAZ_Kangju
36.2	PAK_Saidu_Sharif_H_o
26.4	IRN_Seh_Gabi_C

Target: Tarkalani
Distance: 2.8948% / 0.02894760
39.2	KAZ_Kangju
33.0	PAK_Saidu_Sharif_H_o
27.8	IRN_Seh_Gabi_C

Target: Uthmankhel
Distance: 2.6889% / 0.02688939
40.2	KAZ_Kangju
33.2	PAK_Saidu_Sharif_H_o
26.6	IRN_Seh_Gabi_C

Target: Pashtun:Pashtun2_20Af
Distance: 2.8330% / 0.02832990
32.0	IRN_Seh_Gabi_C
24.6	Saka_Tian_Shan
23.8	PAK_Saidu_Sharif_H_o
19.6	KAZ_Kangju

Also about the similarity with NW South Asians i agree very much with Pegasus here. The basic components (BMAC, IVC, Steppe MLBA+ Central MLBA) are very much shared with NW South Asians and in the end most Afghan Pashtuns are on a cline between Sokhta_BA3+Iran_N and Pamiri with some heavily East Asian admixed Pashtuns only significantly outside of it. The same is to a large extent also true for Iranian Sistani who differ from Kandahari mainly because of less Steppe admix and more Iran_Chalc/West Asian admix but IVC-like admix seems to be very similar. Emotional discussions about if Afghanistan is genetically South Asian, Central Asian or West Asian are in my eyes really useless and redundant because Afghanistan shows and showed strong influences from all these regions and was always a quite unique and super-fascinating mix of all these influences. Historically many parts of it were genetically very South Asian-like (Sokhta samples which are IVC-like, Buddhist/Kushan era Indo-Aryan civilization of East Afghanistan, modern day East Afghans very strongly shifted towards Swat_IA, all Afghan showing IVC-like admix). Other parts like Herat for example were/are historically and genetically more linked/shifted to West Asia/Iran but even here you find strong IVC-like admix.

[ThaYamamoto]

Thank you guys, much appreciated. Pashtuns have been confusing me for a while so thanks for breaking it down. Some new ancients from the region would go down a treat. You both mentioned some extra west-Asian/PPNB/ANF in some Pashtuns - any idea where this is coming from?

[Coldmountains]

Thank you guys, much appreciated. Pashtuns have been confusing me for a while so thanks for breaking it down. Some new ancients from the region would go down a treat. You both mentioned some extra west-Asian/PPNB/ANF in some Pashtuns - any idea where this is coming from?

Likely from various periods and starting with the Achaemenid period if not earlier (late Iran_N/BMAC groups richer in Iran_Chalc?) and intensifying with Sassanid and finally Islamo-Persian expansion into Afghanistan. If i correctly remember Arachosia (Proto-Pashtun homeland) showed during the Achaemenid period already a lot of archaeological influences from more western Iranian regions. After that, you had Sassanid settlers like one of the Kushan period samples from Tajikistan which clustered with West Iranics. Such outliner types were likely also present in many parts of Sassanid-ruled Afghanistan and finally absorbed by locals.

With the arrival of Islam you surely had a lot of Persian and to a lesser extent Arabic settlers coming to the region. Most of them settled in urbanized regions of Khorasan and Bactria than in Pashtun core areas of South and East Afghanistan. Still there was historically and even recently significant immigration of Persian and other West Asians to Kandahar and Kabul so Pashtuns admixed with urban Persian-speaking people of Kabul or Kandhar can also can show recent West Asian input.

In terms of Y-DNA Pashtuns so far show not much of direct matches with Persians or even more western West Asians. Arabic Y-DNA seems to be absent or extremely rare so far ( I dont remember a Pashtun with Arabic Y-DNA unlike some Tajiks and even Pamiri, who show it in rare cases). Persian Y-DNA is harder to differentiate from Pashtun Y-DNA, because they show some overlap from shared BA/Neolithic ancestry (Steppe, BMAC, Iran_N,..) . But if correctly remember neither recent Persian Y-DNA seems to be present among Pashtuns at least not in above trace amounts. So the admixture seems to be mainly female mediated, what makes sense for an originally pastoralist and very tribal group.

[MonkeyDLuffy]

So the Sikh Kamboh that were clustering with Eastern Pashtuns were just superficial similarity?

[1938helerich]

Hi, I just got my G25 coordinates from David. What do I do to get a breakdown of components such as simulated AASI, Iran farmer, etc?


PrV_scaled,0.045529,-0.052808,-0.152357,0.087533,-0.084939,0.048248,0.010105,0.016153,0.026793,0.014 761,-0.004384,0.00015,-0.003271,-0.003578,-0.004479,-0.005436,0.002999,-0.00038,-0.001257,-0.012006,-0.003119,-0.007666,-0.005669,-0.001325,-0.002395

PrV,0.004,-0.0052,-0.0404,0.0271,-0.0276,0.0173,0.0043,0.007,0.0131,0.0081,-0.0027,0.0001,-0.0022,-0.0026,-0.0033,-0.0041,0.0023,-0.0003,-0.001,-0.0096,-0.0025,-0.0062,-0.0046,-0.0011,-0.002


Target: PrV_scaled
Distance: 2.6796% / 0.02679566
75.6 IRN_Shahr_I_Sokhta_BA2
9.6 TJK_Sarazm_En
7.0 Anatolia_Tepecik_Ciftlik_N
2.4 Anatolia_Barcin_N
2.0 Han
1.4 Nganassan
1.0 IRN_Ganj_Dareh_N
1.0 MAR_Iberomaurusian

[ESPLover4]

Hi, I just got my G25 coordinates from David. What do I do to get a breakdown of components such as simulated AASI, Iran farmer, etc?


PrV_scaled,0.045529,-0.052808,-0.152357,0.087533,-0.084939,0.048248,0.010105,0.016153,0.026793,0.014 761,-0.004384,0.00015,-0.003271,-0.003578,-0.004479,-0.005436,0.002999,-0.00038,-0.001257,-0.012006,-0.003119,-0.007666,-0.005669,-0.001325,-0.002395

PrV,0.004,-0.0052,-0.0404,0.0271,-0.0276,0.0173,0.0043,0.007,0.0131,0.0081,-0.0027,0.0001,-0.0022,-0.0026,-0.0033,-0.0041,0.0023,-0.0003,-0.001,-0.0096,-0.0025,-0.0062,-0.0046,-0.0011,-0.002


Target: PrV_scaled
Distance: 2.6796% / 0.02679566
75.6 IRN_Shahr_I_Sokhta_BA2
9.6 TJK_Sarazm_En
7.0 Anatolia_Tepecik_Ciftlik_N
2.4 Anatolia_Barcin_N
2.0 Han
1.4 Nganassan
1.0 IRN_Ganj_Dareh_N
1.0 MAR_Iberomaurusian

Here's an elemental run that I did

Target: PrV_scaled
Distance: 2.3199% / 0.02319922
43.4 IRN_Ganj_Dareh_N
19.0 S_AASI
16.2 AASI_NW
9.6 RUS_Sosonivoy_HG
9.4 TUR_Barcin_N
1.6 NPL_Mebrak_2125BP
0.8 MAR_Taforalt

[1938helerich]

Here's an elemental run that I did

Target: PrV_scaled
Distance: 2.3199% / 0.02319922
43.4 IRN_Ganj_Dareh_N
19.0 S_AASI
16.2 AASI_NW
9.6 RUS_Sosonivoy_HG
9.4 TUR_Barcin_N
1.6 NPL_Mebrak_2125BP
0.8 MAR_Taforalt

Thanks! I appreciate it.

[a_kayani987]

Can someone help me with my g25 co ordinates

a_kayani987_scaled,0.066017,-0.005078,-0.107102,0.076874,-0.064935,0.04769,-0.002115,0.004384,0.003886,0.000364,-0.01088,-0.002847,0.002527,-0.003853,0.00285,-0.00053,-0.022165,-0.005068,0.003645,-0.008879,-0.003369,-0.002349,0.003821,-0.009519,0.006945

a_kayani987,0.0058,-0.0005,-0.0284,0.0238,-0.0211,0.0171,-0.0009,0.0019,0.0019,0.0002,-0.0067,-0.0019,0.0017,-0.0028,0.0021,-0.0004,-0.017,-0.004,0.0029,-0.0071,-0.0027,-0.0019,0.0031,-0.0079,0.0058