PDA

View Full Version : European Component in South and Central Asians



fil
03-06-2015, 06:57 PM
Sorry if this has been answered before, I did a search and couldn't find a topic on this.

Nearly all South and Central Asians score significant European component (I haven't seen anyone else without this component except for me). What does the European component (or the lack of) in South and Central Asians indicate? Thanks!

newtoboard
03-10-2015, 12:57 AM
It represents Indo-Iranian related ancestry as well as some ancestry associated with R2/ANE. I suspect this component will be less European than expected by a lot of people. It will be predominantly ANE and ENF with just a little bit WG imo which is why Reich and Haak couldn't fit Samara Yamnaya to the population that brought IE languages to South Asia. My reasoning for low WHG is the distance away from the forest steppe and Balkans (and proximity to Central Asia) for Indo-Iranian speakers in the lower Don-Volga-Ural regions. And proxmity to the Caucasus and Maykop influence during the Poltavka period will increase ENF.

Anabasis
03-10-2015, 01:24 AM
Fil which calculator did you check? And whats the average of European companent in your population or geography?

Dr_McNinja
03-10-2015, 02:49 AM
The population that brought IE languages to India could've had WHG, but it just disappeared after subsequent migrations by peoples lacking in it. At its highest in the subcontinent, WHG reaches levels found in some Central Asian/Caucasian populations.

parasar
03-10-2015, 03:53 AM
Sorry if this has been answered before, I did a search and couldn't find a topic on this.

Nearly all South and Central Asians score significant European component (I haven't seen anyone else without this component except for me). What does the European component (or the lack of) in South and Central Asians indicate? Thanks!

Components derived from modern populations are virtually useless if you are looking to find ancestral components. Shared ancestry is another matter. The so called European component is shared ancestry - that's all.
Per Lazaridis et al. there are three ancestral components among Europeans - ANE, EEF, and WHG. Even though all three entered Europe from elsewhere - Africa ultimately, most people on this forum identify WHG as the most "European" among the three. We have a number of WHG samples now and on the Y side there is no WHG line that I think of that could have reached India from Europe. On the mtDNA side, that is also the case, though mtDNA U provides some semblance of a 40-50000 year old connection.

fil
03-10-2015, 04:15 AM
It represents Indo-Iranian related ancestry as well as some ancestry associated with R2/ANE. I suspect this component will be less European than expected by a lot of people. It will be predominantly ANE and ENF with just a little bit WG imo which is why Reich and Haak couldn't fit Samara Yamnaya to the population that brought IE languages to South Asia. My reasoning for low WHG is the distance away from the forest steppe and Balkans (and proximity to Central Asia) for Indo-Iranian speakers in the lower Don-Volga-Ural regions. And proxmity to the Caucasus and Maykop influence during the Poltavka period will increase ENF.

Making sure I understand, so the European component represents ancestries from Volga/Ural regions? Does this relate to the Indo-Aryan migration theory? Therefore would people lacking this component would be reflective of the population prior to this migration?


Fil which calculator did you check? And whats the average of European companent in your population or geography?

For Eurogenes K9 for instance, North Indians score ~12% and Afghans score higher around ~18%. Among South Indians, Varun scored ~10% and soulbrighter scored ~8%. I scored .64%. The European components seems to decline abruptly when moving west into Iran.

Morges
03-10-2015, 11:19 AM
You mean autosomally (like Eurogenes K13) or ANE-WHG-ENF?

Anabasis
03-10-2015, 11:22 PM
Making sure I understand, so the European component represents ancestries from Volga/Ural regions? Does this relate to the Indo-Aryan migration theory? Therefore would people lacking this component would be reflective of the population prior to this migration?
.

I think answer is either yes or no. Its yes if the population which lack of this companent is somewhere could be isolated and diverse from thier neighbours. Personal results does not mean anything as far as its not a member of "unique" population which is suppose to be diferent then other populations.

Anabasis
03-10-2015, 11:29 PM
BTW Eurogenes companents very unsuccesful for dedecting indo european expansion. K8 might be better for it but Davidski had made very big mistake. He didint separate Gedrosa and Anchestral North Indian Companents. He put both of them into same pocket, The ANE. That made Armenians semi ancestor of Proto IE. But most "ANE" in Armenians was related with Gedrosa companent which has deep connection with caucasia since paleotic times. MDLP k23 calculator i think its better. The anchestral Altaic companent may give some ideas about the expansion of IE.

Generalissimo
03-11-2015, 03:34 AM
BTW Eurogenes companents very unsuccesful for dedecting indo european expansion. K8 might be better for it but Davidski had made very big mistake. He didint separate Gedrosa and Anchestral North Indian Companents. He put both of them into same pocket, The ANE. That made Armenians semi ancestor of Proto IE. But most "ANE" in Armenians was related with Gedrosa companent which has deep connection with caucasia since paleotic times. MDLP k23 calculator i think its better. The anchestral Altaic companent may give some ideas about the expansion of IE.

"Gedrosia" was never a real population. It's just a cluster from an Admixture run not based on any ancient genomes.

And I seriously doubt that any Altaic component can represent the proto Indo-Europeans.

Anabasis
03-11-2015, 06:15 AM
"Gedrosia" was never a real population. It's just a cluster from an Admixture run not based on any ancient genomes.

And I seriously doubt that any Altaic component can represent the proto Indo-Europeans.

Yes its companent which refers wider geographical genetic character, and we are talking about companents here.

Dr_McNinja
03-11-2015, 07:00 PM
If we can distinguish between different types of ANE, we may be able to figure out more about this in South Asians.

To recap from previous posts:

With the Eurogenes K7/K8 calculators in conjunction with the typical modern component-based calculators, it becomes possible to discern a "basal" layer of 28-32% "ENF"-like admixture (what I called Southwest-ENF or SW-ENF because of where I presumed the source was from) in South Asians. This made a "leftover ENF" which corresponded to Gedrosian.

So that's two layers of ENF. A basal layer of ENF, and then a Gedrosian-like layer of ENF.

The problem is, we can't figure out how many layers of ANE there are and how much ANE goes in the different layers.

Fooling around with the numbers:

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE#gid=25

Scroll all the way over to the far right to the blue labels.

I assembled pseudo-components of South Asian (50% ASE, 25% ANE, 25% ENF), Gedrosian (70:30 Gedrosian-ENF to ANE), and K15-Eastern Euro (40% WHG, 24.8% ANE, 27.72% ENF). The South Asian is calculated off ASE, the Gedrosian from Gedrosian-ENF, and K15-Eastern-Euro from all of the WHG.

This leaves leftover ANE and ENF. The last column is the % of ANE for the "missing component", assuming we can lump all the leftover ANE and ENF together. Obviously, this won't be relevant for Afghans who probably have at least one extra Caucasian layer, but for South Asians, a surprisingly clustered range emerges. 36.18% ANE in a Lohana Sindh on one end to 46.4% ANE in a Haryana Jatt on the other end. The average of the South Asians east-of-Afghanistan is 40.6% ANE.

The extra bit missing from the total could be assumed to be ENA or Archaic/African-like.

I also put in the % ANE of the combined Gedrosian+East_Euro (if we presume both taken together represent actual West Eurasian/West Asian origin ancestry), and the numbers are a few points lower, around mid-30s average. What that represents is the ANE tied up in components associated with WHG and Gedrosian-ENF.

What this suggests, I'm not sure. There's a clear increasing trend towards Iran for the combined Gedrosian-European, the first time I've gotten that. And the closest South Asian is, surprisingly, a Nepal Brahmin (sort of makes sense if the East Asian in Nepal represents completely separate admixture layer, which it likely does... although can't say for sure because all the Arctic-like East Eurasian is lumped in with it). At the opposite end is a WHG-less Gedrosian at 30% ANE. Remember, this is only meaningful if we assume that the combined Gedrosian-European represents something real.

As for the "leftover ANE + leftover ENF", I don't know what this could represent. It'd be a population from the Near East that's 40% ANE. I don't think that could represent something real. I think this may suggest ANE's origin is closely related to South/South-Central Asia. There's just too much of it for a modern day population, across all layers (as opposed to Native Americans who didn't have as many overlapping admixture layers). The original South Asians would have had to be overwhelmingly ANE in composition it seems... Don't know, that's totally a guess.

EDIT: I mean, it's a pretty safe bet to assume a West Asian/Gedrosian component (70:30 ENF:ANE) would have to be older than the WHG in South Asia, right? The WHG spreads out in the Indo-Aryan pattern, and correct me if I'm wrong, I don't think that can be older than 3000-4000ybp.

Dr_McNinja
03-12-2015, 02:23 AM
I added some more columns, colored in red. There's South Asian (same as before, 50% ASE, 25% ENF, 25% ANE), Yamnaya (35% ANE, 35% WHG, 24% ENF, 5% ASE) and "Total Leftover" here comes out like Gedrosian (averaging 30-35% ANE, 63-70% ENF).

Which is interesting because it looks like mixing a population that's 31% ANE, ~6% ASE, ~48% ENF, and ~19% WHG with the Karelia/Samara Hunter Gatherers will give you Yamnaya. That sounds like an old South Central Asian population because it's one admixture step away from current Haryana Jatts. Mix it with 31% ANE, 25% ASE, 35% ENF, 0% WHG which is close to a current South Indian/Bengali population.

One possible scenario: That population (31 ANE, 6 ASE, 48 ENF, 19 WHG) no longer exists, if it ever did, but it would've been the north/easternly neighbor of a 30/70 to 40/60 ANE:ENF style West Asian population and probably the product of that kind of population mixed with whoever was occupying the steppe at the time (or just those Hunter Gatherers). Then it mixed with that ancestor population again to lead to today, where West Asian is just a component spread throughout Eurasia. It wouldn't even be a single population per say but probably an entire range of peoples from that population covering the border of SW-Asia and Anatolia/Asia, marking the first time early near eastern farmers mixed with ANE. This proportion of mixture probably ranged from 50/50 to every combination/range, but the signal just survives in its most "pure" form within an admixture layer stretching over Eurasia currently being called West Asian/Caucasus/Gedrosian.

I think ANE and ENF could've mixed first, possibly somewhere in Iran or South Central Asia, before moving up to the steppe and then mixing with Hunter Gatherers.

I have no idea how SC-Asians would be scoring substantial Yamnaya in Admixture without anywhere near enough WHG. The WHG numbers in that spreadsheet are K7, so using K8's would be even lower.

EDIT: This would imply 3 admixture events in South Asia (well, a high of 25% ASE in Southeast India already implied that). For simplicity's sake: 100% ASE mixes with 50% ANE/50% ENF from the West to make 50% ASE, 25% ANE, 25% ENF. Then mixed with another population that's 37 ANE, 45 ENF, maybe some WHG. This gives 31 ANE, 25 ASE, 35 ENF. Then mixed a final time with something similar to the 31 ANE, 48 ENF, 19 WHG Central/South Asian population that was an ancestor of Yamnaya. This gives something very close to Haryana Jatt admixture in North India. The second two events would be with similar populations or the same population twice after a long period of time.

Dr_McNinja
03-12-2015, 03:46 PM
You can use these in the 4-Ancestors Oracle program:

8 {Quantity of components}
ANE
ASE
WHG
ENF
East-Eurasian
Oceanian
Pygmy
SSA
EHG 37.71 3.61 52.40 0.02 4.99 0.75 0.40 0.13
Yamnaya 34.38 4.72 35.64 23.96 0.00 0.55 0.59 0.15
SC-Asia 31.06 5.84 18.88 44.22 0.00 0.00 0.00 0.00
Corded-Ware 24.69 2.11 40.57 30.63 0.32 0.99 0.59 0.10
Bell-Beaker 15.55 0.98 44.82 36.78 0.93 0.18 0.01 0.74
Euro-MN 0.00 0.00 43.71 54.81 0.02 1.46 0.00 0.00
Ancestral-South-Asia 25.00 50.00 0.00 25.00 0.00 0.00 0.00 0.00
Uygur 17.86 6.58 7.00 25.00 42.00 1.10 0.00 0.50

EHG is Karelia/Samara Hunter Gatherers. Euro-MN is Baalberge_MN, Esperstedt_MN, and Gokhem_MN. SC-Asia is the mystery population that could've mixed with EHG to produce Yamnaya and currently being picked up as West Asian/Gedrosian. Ancestral-South-Asia is the early Neolithic South Asia presuming a Near Eastern/West Asian population of 50 ANE, 50 ENF mixing with ASE in the subcontinent (an alternate scenario is ENF entering South Asia, mixing with ASE and later ANE coming in to produce 50 ANE, 25 ASE, 25 ENF). I added Uygur for some East Asian/Eastern Central Asian.

Using my dad's West Eurasia K8 results:


Least-squares method.

Using 1 population approximation:
1 SC-Asia @ 20.962151
2 Ancestral-South-Asia @ 37.942617
3 Yamnaya @ 39.765418
4 Corded-Ware @ 42.826192
5 Uygur @ 44.45964
6 Bell-Beaker @ 48.053646
7 Euro-MN @ 55.988426
8 EHG @ 66.35615
8 iterations.

Using 2 populations approximation:
1 SC-Asia+Ancestral-South-Asia @ 16.166446
2 SC-Asia+SC-Asia @ 20.962151
3 SC-Asia+Uygur @ 24.403526
4 Corded-Ware+Ancestral-South-Asia @ 25.371437
5 Yamnaya+Ancestral-South-Asia @ 25.648408
6 Bell-Beaker+Ancestral-South-Asia @ 26.719992
7 Ancestral-South-Asia+Uygur @ 27.762329
8 Euro-MN+Ancestral-South-Asia @ 28.382918
36 iterations.

Using 3 populations approximation:
1 50% SC-Asia +25% SC-Asia +25% Ancestral-South-Asia @ 13.447663
2 50% SC-Asia +25% Ancestral-South-Asia +25% Uygur @ 13.931692
3 50% SC-Asia +25% Ancestral-South-Asia +25% Ancestral-South-Asia @ 16.166446
4 50% SC-Asia +25% Yamnaya +25% Ancestral-South-Asia @ 18.293469
5 50% SC-Asia +25% Corded-Ware +25% Ancestral-South-Asia @ 19.023095
6 50% Ancestral-South-Asia +25% SC-Asia +25% Uygur @ 19.050579
7 50% SC-Asia +25% SC-Asia +25% Uygur @ 19.08785
8 50% SC-Asia +25% Bell-Beaker +25% Ancestral-South-Asia @ 20.086299
252 iterations.

Using 4 populations approximation:
1 SC-Asia+SC-Asia+SC-Asia+Ancestral-South-Asia @ 13.447663
2 SC-Asia+SC-Asia+Ancestral-South-Asia+Uygur @ 13.931692
3 SC-Asia+SC-Asia+Ancestral-South-Asia+Ancestral-South-Asia @ 16.166446
4 Yamnaya+SC-Asia+SC-Asia+Ancestral-South-Asia @ 18.293469
5 SC-Asia+SC-Asia+Corded-Ware+Ancestral-South-Asia @ 19.023095
6 SC-Asia+Ancestral-South-Asia+Ancestral-South-Asia+Uygur @ 19.050579
7 SC-Asia+SC-Asia+SC-Asia+Uygur @ 19.08785
8 Yamnaya+SC-Asia+Ancestral-South-Asia+Uygur @ 19.13617
9 SC-Asia+Corded-Ware+Ancestral-South-Asia+Uygur @ 19.686907
10 SC-Asia+SC-Asia+Bell-Beaker+Ancestral-South-Asia @ 20.086299
11 SC-Asia+Corded-Ware+Ancestral-South-Asia+Ancestral-South-Asia @ 20.201231
12 Yamnaya+SC-Asia+Ancestral-South-Asia+Ancestral-South-Asia @ 20.385655
13 SC-Asia+Euro-MN+Ancestral-South-Asia+Ancestral-South-Asia @ 20.624491
14 SC-Asia+Bell-Beaker+Ancestral-South-Asia+Ancestral-South-Asia @ 20.628416
15 SC-Asia+Bell-Beaker+Ancestral-South-Asia+Uygur @ 20.643573
16 SC-Asia+SC-Asia+Euro-MN+Ancestral-South-Asia @ 20.919019
330 iterations.


Haryana Jatt:


Least-squares method.

Using 1 population approximation:
1 SC-Asia @ 16.214864
2 Yamnaya @ 33.987877
3 Corded-Ware @ 37.437131
4 Ancestral-South-Asia @ 39.031263
5 Bell-Beaker @ 43.31671
6 Uygur @ 46.115334
7 Euro-MN @ 53.011973
8 EHG @ 61.169182
8 iterations.

Using 2 populations approximation:
1 SC-Asia+Ancestral-South-Asia @ 14.66314
2 SC-Asia+SC-Asia @ 16.214864
3 Corded-Ware+Ancestral-South-Asia @ 21.664176
4 Yamnaya+Ancestral-South-Asia @ 22.060896
5 Yamnaya+SC-Asia @ 23.085904
6 Bell-Beaker+Ancestral-South-Asia @ 23.226038
7 SC-Asia+Uygur @ 24.130257
8 SC-Asia+Corded-Ware @ 25.582448
36 iterations.

Using 3 populations approximation:
1 50% SC-Asia +25% SC-Asia +25% Ancestral-South-Asia @ 8.331957
2 50% SC-Asia +25% Yamnaya +25% Ancestral-South-Asia @ 12.681766
3 50% SC-Asia +25% Ancestral-South-Asia +25% Uygur @ 12.817469
4 50% SC-Asia +25% Corded-Ware +25% Ancestral-South-Asia @ 13.654542
5 50% SC-Asia +25% Ancestral-South-Asia +25% Ancestral-South-Asia @ 14.66314
6 50% SC-Asia +25% Bell-Beaker +25% Ancestral-South-Asia @ 15.098341
7 50% SC-Asia +25% SC-Asia +25% Uygur @ 16.415532
8 50% SC-Asia +25% Euro-MN +25% Ancestral-South-Asia @ 17.005914
252 iterations.

Using 4 populations approximation:
1 SC-Asia+SC-Asia+SC-Asia+Ancestral-South-Asia @ 8.331957
2 Yamnaya+SC-Asia+SC-Asia+Ancestral-South-Asia @ 12.681766
3 SC-Asia+SC-Asia+Ancestral-South-Asia+Uygur @ 12.817469
4 SC-Asia+SC-Asia+Corded-Ware+Ancestral-South-Asia @ 13.654542
5 SC-Asia+SC-Asia+Ancestral-South-Asia+Ancestral-South-Asia @ 14.66314
6 SC-Asia+SC-Asia+Bell-Beaker+Ancestral-South-Asia @ 15.098341
7 SC-Asia+SC-Asia+SC-Asia+SC-Asia @ 16.214864
8 SC-Asia+SC-Asia+SC-Asia+Uygur @ 16.415532
9 Yamnaya+SC-Asia+Ancestral-South-Asia+Uygur @ 16.552426
10 SC-Asia+SC-Asia+Euro-MN+Ancestral-South-Asia @ 17.005914
11 SC-Asia+Corded-Ware+Ancestral-South-Asia+Uygur @ 17.139319
12 SC-Asia+Corded-Ware+Ancestral-South-Asia+Ancestral-South-Asia @ 17.254853
13 Yamnaya+SC-Asia+Ancestral-South-Asia+Ancestral-South-Asia @ 17.516362
14 SC-Asia+Bell-Beaker+Ancestral-South-Asia+Ancestral-South-Asia @ 17.751165
15 SC-Asia+Bell-Beaker+Ancestral-South-Asia+Uygur @ 18.228307
16 Yamnaya+Yamnaya+SC-Asia+Ancestral-South-Asia @ 18.44662
330 iterations.

Dr_McNinja
03-13-2015, 02:33 AM
Messing around in Excel:

http://i.imgur.com/GaMwVCw.png

(This is ANE K7, not K8)

Dr_McNinja
03-13-2015, 01:26 PM
Harappa spreadsheet,

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE#gid=6

Scroll to the right to the orange ANE/ENF/WHG categories. I derived those from Harappa components assuming the following:

S-Indian: 25% ANE, 25% ENF
Baloch: 50% ANE, 50% ENF
Caucasian: 30% ANE, 70% ENF
NE-Euro: 18% ANE, 28% ENF, 50% WHG
Siberian: 17% ANE
American: 38% ANE
Beringian: 25% ANE
Mediterranean: 65% ENF, 35% WHG
SW-Asian: 89% ENF

On the bottom are two Yamnaya individuals. Their WHG seems underestimated while the rest of ours seem to be between Eurogenes ANE K7 and Eurogenes West Eurasia K8 (so, a little overestimated).

I don't know if anyone's run zombies of these components through K7/K8, but that was my best guess. I did the other components first, then tweaked Baloch to try and see where it would be. An absurdly high ANE number makes sense and it would explain how it varies so drastically. I think Parasar was saying this a while back too.

That also explains the high Arctic categories in the presence of East Eurasian and absence of WHG.

But it does show that in lieu of more fundamental components, Admixture can really fudge things together to make a best fit. For better (when using populations that are themselves hybrids of different groups, as is the case in South Asia) and worse (in order to bring everyone else into line).

EDIT: Added another 4 columns (ANE/ASE/WHG/ENF/East-Eurasian) using the actual Baloch->K7 breakdown from everest:


35.41% ANE
9.28% ASE
0.01% WHG-UHG
0.00% East_Eurasian
0.07% West_African
0.12% East_African
55.11% ENF

Adjusted S-Indian to 30% ANE, 25% ENF, 36% ASE, 9% ENA and took 4% ENA from leftover NE-Euro.

He'll send the breakdown of the other components soon. I think the actual one will have more ASE and less ENA in S-Indian. Too bad K7 doesn't have the Oceanian component.

tamilgangster
03-16-2015, 11:08 AM
Sorry if this has been answered before, I did a search and couldn't find a topic on this.

Nearly all South and Central Asians score significant European component (I haven't seen anyone else without this component except for me). What does the European component (or the lack of) in South and Central Asians indicate? Thanks!

The only south asians that score european component are north west indian populations and brahmins. Only north indian brahmins, Jatts, and khatris score greater than 10% northern european