Global25 automated nMonte for South/Central Asian members

[MonkeyDLuffy]

How about some Jat Hindus? Here's mine.

Target: hoodakush
Distance: 1.3372% / 0.01337192
54.2 IRN_Shahr_I_Sokhta_BA2
37.8 RUS_Sintashta_MLBA
5.4 TKM_Gonur1_BA
2.6 NPL_Chokhopani_2700BP


Distance to: hoodakush
0.01330421 Ror
0.01940982 Jatt_Pathak
0.02957210 Punjabi_Jatt
0.02983139 Pashtun_Kurram
0.03060586 Yusufzai
0.03061970 Kamboj
0.03263400 Uthmankhel
0.03604726 Tarkalani
0.03743795 Punjabi_Sikh_India
0.03865305 Khatri
0.04018296 Kalash
0.04046180 Kho_Singanali
0.04394162 Kohistani
0.05044378 Sindhi
0.05087782 Gujar_Pakistan
0.05102764 Punjabi_Hindu_India
0.05131362 Kashmiri_Pandit
0.05303763 Pashtun_North_Afghanistan
0.05539617 Kashmiri_Pakistan
0.05809693 Tajik_Ishkashim
0.06026444 Kashmiri_Pakistan_o
0.06332192 Punjabi_Muslim_India
0.06466599 Sarikoli_China
0.06469247 Burusho
0.06491075 Gujar_India

Which part of Haryana your family's from?

[pegasus]

Distance to: hoodakush
0.04362101 PAK_Barikot_H
0.04426696 PAK_Singoor_MA
0.04569176 PAK_Udegram_MA_Ghaznavid
0.04707867 PAK_Barikot_IA
0.04947433 PAK_Udegram_IA
0.05274858 PAK_Saidu_Sharif_H
0.05367692 PAK_Katelai_IA
0.05424520 PAK_Gogdara_IA
0.05442458 PAK_Loebanr_IA
0.05646980 PAK_Butkara_IA
0.05921188 CHN_Junmachanyilian_HE2
0.06262930 PAK_Aligrama_H
0.06366908 PAK_Loebanr_IA_o
0.06581827 PAK_Barikot_MA
0.06808884 PAK_Arkotkila_IA
0.07050025 UZB_Bustan_BA_o2
0.08073411 PAK_Butkara_H
0.08383557 CHN_Jirzankal_IA3
0.08450401 CHN_Junmachanyilian_HE1
0.09329416 CHN_Jirentaigoukou_IA2
0.09537033 KAZ_Otyrar_Antiquity
0.09623281 CHN_Liushui_IA2
0.09658902 PAK_Aligrama_IA
0.09784094 UZB_Bustan_BA_o1
0.10130825 TJK_Ksirov_H_Kushan

Can you share your coordinates , you should not be getting 2.6 chokopani at all, thats excess AASI.

[pegasus]

I've tried modeling w Onge + Chokopani, here are the results for other Jats/Rors

Target: UP_Jatt:jattsheoran_AGUser_scaled
Distance: 2.7818% / 0.02781774
54.6 IRN_Shahr_I_Sokhta_BA2
41.8 RUS_Sintashta_MLBA
2.4 NPL_Chokhopani_2700BP
1.2 TKM_Gonur1_BA



Target: kringzy123t:Jatt_Malik_West_UP_scaled
Distance: 2.9420% / 0.02942025
53.4 IRN_Shahr_I_Sokhta_BA2
38.2 RUS_Sintashta_MLBA
4.8 TKM_Gonur1_BA
3.6 NPL_Chokhopani_2700BP



Target: UP_Jatt:mathura_haga_AGUser_scaled
Distance: 2.4463% / 0.02446334
65.0 IRN_Shahr_I_Sokhta_BA2
28.6 RUS_Sintashta_MLBA
4.6 Onge
1.8 NPL_Chokhopani_2700BP



Target: Jatt_Pathak
Distance: 2.0636% / 0.02063608
54.2 IRN_Shahr_I_Sokhta_BA2
40.0 RUS_Sintashta_MLBA
3.0 NPL_Chokhopani_2700BP
2.8 TKM_Gonur1_BA



Target: hoodakush (mine)
Distance: 1.3372% / 0.01337192
54.2 IRN_Shahr_I_Sokhta_BA2
37.8 RUS_Sintashta_MLBA
5.4 TKM_Gonur1_BA
2.6 NPL_Chokhopani_2700BP


Target: Ror
Distance: 1.3702% / 0.01370185
56.2 IRN_Shahr_I_Sokhta_BA2
39.4 RUS_Sintashta_MLBA
2.2 NPL_Chokhopani_2700BP
2.2 TKM_Gonur1_BA


Target: Punjabi_Jatt
Distance: 1.3996% / 0.01399580
63.0 IRN_Shahr_I_Sokhta_BA2
28.8 RUS_Sintashta_MLBA
6.2 TKM_Gonur1_BA
2.0 NPL_Chokhopani_2700BP


It definitely looks like some Chokopani/Tibetan is needed. Do you think it's excess east eurasian being counted as Chokopani but its actually AASI?

I have posted qpAdm models on Rors extensively, they have excess WSHG affinity, not East Asian affinity, also they don't model well at all with the SIS2 cluster , that has too much Iran_N for them, they need a Zagrosian related source which has ample CHG to the exclusion of what is in Steppe MLBA but a local substrate which is rich in AASI esp at their admixture date around 1100-1000 BC.


sample: Ror:Average
distance: 1.026
Telugu_GBR • HG03785.SG: 37
Srubnaya_Alakul_MLBA • Average: 36
Dzharkutan1_BA • Average: 24
Tarim_EMBA1 • Average: 3

[Kaazi]

Nagar Brahmin sims

Sample	Fit	IRN Shahr I Sokhta BA3	RUS Srubnaya Alakul MLBA	UZB Dzharkutan1 BA	NPL Chokhopani 2700BP
Nagar Brahmin Gujarat Rana2-HW-sim (Kaazi)	1.81	61.2	28.4	10.4	0
Nagar Brahmin Gujarat Nanavati-HW-sim (Kaazi)	1.84	57.6	31.6	10.8	0
Nagar Brahmin Gujarat Rana1-HW-sim (Kaazi)	1.76	57.2	26.4	15.4	1
Nagar Brahmin Median (Kaazi)	1.43	57	28.2	14.2	0.6
Nagar Brahmin Average (Kaazi)	1.34	56.4	28.2	14.8	0.6
Nagar Brahmin Gujarat Buch-HW-sim (Kaazi)	2.53	56	26.8	16.2	1
Nagar Brahmin Gujarat Vaishnav-HW-sim (Kaazi)	1.64	54.4	27.4	16.6	1.6
Nagar Brahmin Nanavati 2-HW-sim (Kaazi)	1.58	54	29.2	16.4	0.4
Nagar Brahmin Gujarat Parghi-HW-sim (Kaazi)	1.49	54	27.6	18.2	0.2

Sample:	Kaazi ► Nagar Brahmin [email protected]
Fit:	1.34
Results:	IRN Shahr I Sokhta BA3	56.4
	RUS Srubnaya Alakul MLBA	28.2
	UZB Dzharkutan1 BA	14.8
	NPL Chokhopani 2700BP	0.6

Nagar Brahmin HW Average (7kit)

 

S-Indian: 33.82%
Baloch: 36.44%
Caucasian: 8.85%
NE-Euro: 14.29%
SE-Asian: 0.05%
Siberian: 0.84%
NE-Asian: 0.27%
Papuan: 0.31%
American: 0.68%
Beringian: 0.75%
Mediterranean: 2.54%
SW-Asian: 0.84%
San: 0.04%
E-African: 0.18%
Pygmy: 0.10%
W-African: 0.00%

I realized that AG Bahun average is slightly skewed towards Steppe while Nepalese A seems to represent the overall Bahun average better. Nagars seem close to these proto-Bahuns/NepaleseA.

Sample:	Kaazi ► NepaleseA ghost
Fit:	1.53
Results:	IRN Shahr I Sokhta BA3	58.4
	RUS Srubnaya Alakul MLBA	27.4
	UZB Dzharkutan1 BA	14.2

 

Sample:	Nepalese A ► Average
Fit:	1.42
Results:	IRN Shahr I Sokhta BA3	54.2
	RUS Srubnaya Alakul MLBA	25.4
	UZB Dzharkutan1 BA	13.4
	NPL Chokhopani 2700BP	7

[MonkeyDLuffy]

Thanks, that's very interesting. Why do you think they don't model well with Sis2? Can you share some of your qpadm models? What p value did you get? I can PM you my coordinates too.

Because BA2/3 are Iran N rich samples, and high steppe groups like Rors and Hindu Jats score low Iran N. Most probably the steppe carrying ancestors of these groups mixed up with low Iran N groups (similar to Saidu Sharif outlier) of the region when they arrived.

[dante911]

I have posted qpAdm models on Rors extensively, they have excess WSHG affinity, not East Asian affinity, also they don't model well at all with the SIS2 cluster , that has too much Iran_N for them, they need a Zagrosian related source which has ample CHG to the exclusion of what is in Steppe MLBA but a local substrate which is rich in AASI esp at their admixture date around 1100-1000 BC.


sample: Ror:Average
distance: 1.026
Telugu_GBR • HG03785.SG: 37
Srubnaya_Alakul_MLBA • Average: 36
Dzharkutan1_BA • Average: 24
Tarim_EMBA1 • Average: 3

Can you post the rors qpADM models along with the right populations and allSNPS setting used ? I ask this because i know some folks who have been trying to model rors on qpADM and they have failed.
Can you also please post the ALDER data for estimate admixture in rors as well ?

Thanks.
(As far as I know, sometimes, using the incorrect or feeble right pop set can make a qpADM model pass even with same source pops used but that doesn't make it correct)

[pegasus]

Thanks, that's very interesting. Why do you think they don't model well with Sis2? Can you share some of your qpadm models? What p value did you get? I can PM you my coordinates too.

Because they are relatively depressed for Zagrosian related ancestry, have plenty of Steppe related ancestry but significant AASI levels. Hypothetically if they did work with the main SIS2 cluster they would be like Steppe shifted Kalash or Pashtun and get sterling p values, they don't.

[Kapisa]

https://anthrogenica.com/showthread....l=1#post865283

This Sindhi Hindu scores one of the highest amount of BMAC.

Sample	Fit	IRN Shahr I Sokhta BA3	UZB Dzharkutan1 BA	RUS Srubnaya Alakul MLBA	NPL Chokhopani 2700BP
Average (Sindhi)	1.62	45.6	36.4	16.6	1.4
Average (Khatri)	1.18	44.8	32	21.4	1.8
Sindhi Hindu (Kaazi)	3.15	44.2	38	17.8	0
Arain AVG(13) (Samand)	1.43	41.8	38.8	18.2	1.2
Average (Kamboj)	1.37	38.8	35	24	2.2

Sample	Fit	Iran N	AASI	Eastern HG	Anatolia N	West Siberia HG	Caucasus HG
Sindhi Hindu (Kaazi)	7.64	55	20	17	6.2	1.8	0
Average (Sindhi)	7.10	50.8	24.2	8.6	10.6	5.8	0
Arain AVG(13) (Samand)	6.96	47.8	21.4	10	9	7.2	4.6
Average (Khatri)	7.06	47.6	22.4	11.8	10.4	6.2	1.6
Average (Kamboj)	6.42	43.2	21	11.2	12.6	7.4	4.6

Sample:	Kaazi ► Sindhi Hindu
Fit:	7.64
Results:	Iran N	55
	AASI	20
	Eastern HG	17
	Anatolia N	6.2
	West Siberia HG	1.8

He's 55% IranN related that's sick number, exceeds all these Indo-Aryan groups. Do this guy carry Gedrosian flow rather than BMAC as he's low on ANF? @Samand, can you share his Harappaworld?

Sample:	Kapisa ► Sindhi Hindu
Fit:	2.8936
Results:	Bustan BA • Average	41.5
	Shahr I Sokhta BA3 • Average	41.5
	Ak Moustafa MLBA1 • Average	16
	Chokhopani 2700BP • Average	1

Sample:	Kapisa ► Sindhi Hindu
Fit:	3.2913
Results:	Tepe Hissar C • Average	36.5
	Ganj Dareh N • Average	25
	Simulated AASI by DMXX • Average	18.5
	Samara HG • Average	15.5
	IRL Megalithic • Average	4.5

[pegasus]

Can you post the rors qpADM models along with the right populations and allSNPS setting used ? I ask this because i know some folks who have been trying to model qpADM on rors and they have failed.
Can you also please post the ALDER data for estimate admixture in rors as well ?

Thanks.
(As far as I know, sometimes, using the incorrect or feeble right pop set can make a qpADM model pass even with same source pops used but that doesn't make it correct)

I posted it almost a year back , I have to look it up, or you can search in the posts. The right pops vary, but in rotating qpAdm, Kangju was highly preferred and I doubt that will change because there are no comparable samples yet, but Kaazi's ghost is probably extremely close and its shaping up that way looking at some of those Hasanlu and Megiddo outliers.

 

Ror
 
jmean:      0.000 std. err:    106.043 jmean (years):      0.000 std. err(years):   3075.247
 
Kalash
 
jmean:      0.000 std. err:    126.689 jmean (years):      0.000 std. err(years):   3673.981

GujaratiA
 
jmean:      0.000 std. err:    103.088 jmean (years):      0.000 std. err(years):   2989.552
 
 
Gujrati_Brahmins
 
jmean:      0.000 std. err:    105.536 jmean (years):      0.000 std. err(years):   3060.544

Here are the dates, I just realized GujaratiA are likely "Nagar" Brahmins their admixture dates coalesce with Rors so it is clear all these groups descend from very similar populations.

[dante911]

I posted it almost a year back , I have to look it up, or you can search in the posts. The right pops vary, but in rotating qpAdm, Kangju was highly preferred and I doubt that will change because there are no comparable samples yet, but Kaazi's ghost is probably extremely close and its shaping up that way looking at some of those Hasanlu and Megiddo outliers.

 

Ror
 
jmean:      0.000 std. err:    106.043 jmean (years):      0.000 std. err(years):   3075.247
 
Kalash
 
jmean:      0.000 std. err:    126.689 jmean (years):      0.000 std. err(years):   3673.981

GujaratiA
 
jmean:      0.000 std. err:    103.088 jmean (years):      0.000 std. err(years):   2989.552
 
 
Gujrati_Brahmins
 
jmean:      0.000 std. err:    105.536 jmean (years):      0.000 std. err(years):   3060.544

Here are the dates, I just realized GujaratiA are likely "Nagar" Brahmins their admixture dates coalesce with Rors so it is clear all these groups descend from very similar populations.

Thanks a lot. Two follow-up questions -

1. Do you remember by any chance the populations A and population B using which admixture dates for Rors was inferred ?

2. You talked about Hansanlu and Megiddo outliers. May I know which outliers to be more precise and what's their break-up like ?