Central and South Asian DNA Paper


I see, but there is no reason to believe that the earliest PIE speakers must have come from BMAC civilization just because it was very advanced and sophisticated for its time. The earliest PIE tribes were probably not even that developed and sophistiated compared to others, so I see no necessary cultural link between them. If we had seen a huge IE expansion based mostly on refined urban civilizations and luxury trades, I'd take this much more seriously, but that isn't what the archaeological and linguistic records indicate.

Besides, the article in the link you provided refers to the Bronze Age Oxus civilization of Turkmenistan. In the Bronze Age the fully developed and in fact already diverging IE tribes were already spreading from Central Europe to Central Asia, and Anatolian speakers were already in Anatolia forming their kingdoms. There is no use in investigating Bronze Age features of cultures in Turkmenistan when we are discussing about the earliest formation of PIE still in the Neolithic age, probably earlier than 4,500 or even 5,000 BC, before Yamna, CWC and any other seemingly IE-speaking culture. What happens in Iran or in Turkmenistan before the earliest introgression of CHG and R1b in the steppes probably interests us, but that was at the latest around 4,000 BC.
 
Are there any kazark Eneolithic and west siberian neolithic Y DNAs? why did they miss all of them? How come the yamna and afanasievo connection is to be explanined w/o their genes. They were all intermediate like okunevo.

All Botaya skulls are large, have a characteristic horizontal
Flatness in the front part, which is also noted in some ancient
Finds of Western Siberia (Protoka & Sopka-2), the steppe Urals (Gladunino-3),
Western Kazakhstan (Shoktybai, Kumsai, Zhirenkopa, Ishkinovka), the Eastern
Kazakhstan (Shiderty, Zhelezinka, Ust-Narymsky, Rough II), and the Northern
Turkmenistan (Tumek-Kichidzhik / Priaralye). Thus the Botany skulls
Represent a separate anthropological type, formed in the steppe
Part of Asia during the Eneolithic period - "Kazakh steppe type".
 
Last edited:
Other than the fact that there's no BMAC in Indians, and that they found the ANE heavy admixing population (West Siberian hunter-gatherers), not much surprising in terms of the ethnogenesis of the South Asian populations.

I guess Reich got tired of waiting for the Indian scientists to bite the bullet and accept the obvious.
 
Looks like someone on Anthrogenica got his hands on the data and [FONT=Helvetica, Arial, sans-serif]sure enough the Hajji Firuz Tepe individual is R1b-Z2103. Autosomally all samples from this site are very similar - i. e. a coalescence of Iran Neolithic and Anatolian Neolithic streams of ancestry, with an emphasis on the former. Lazaridis predicted that such a population would have been the source of the southern ancestry in the PC steppe. Wow.[/FONT]
 
Looks like someone on Anthrogenica got his hands on the data and sure enough the Hajji Firuz Tepe individual is R1b-Z2103. Autosomally all samples from this site are very similar - i. e. a coalescence of Iran Neolithic and Anatolian Neolithic streams of ancestry, with an emphasis on the former. Lazaridis predicted that such a population would have been the source of the southern ancestry in the PC steppe. Wow.

Wow indeed. So, the Anatolian Neolithic stream may have come from two directions, one from over the Caucasus and one from "Old Europe"?

If this is right, it also means our predictions as to dating are really off, yes? Wasn't R1b Z2103 supposed to be a lot younger than this?

I'm interested in those two pulses north too, one into Khvalynsk and the other into Yamnaya.

Overall "steppe" impact is not at all as high in South Asia as some would have had us believe. Maybe 10-20%, which would analogous perhaps to the situation in Greece? It's higher in Brahmins and princely castes, however, maybe from 20-30%.
 
Hajji Firuz autosomal data for these 6 samples. (Green: Iranian Farmer - Orange: Anatolian Farmer)
xd5kc5.png

1. R1b-M269
2. J2b
3. No Y-dna
4. J2b
5. No Y-dna
6. CT


A simpler overview of the Y-Hg's from the study:
ehaczq.png
 
Wow indeed. So, the Anatolian Neolithic stream may have come from two directions, one from over the Caucasus and one from "Old Europe"?

If this is right, it also means our predictions as to dating are really off, yes? Wasn't R1b Z2103 supposed to be a lot younger than this?

I'm interested in those two pulses north too, one into Khvalynsk and the other into Yamnaya.

Overall "steppe" impact is not at all as high in South Asia as some would have had us believe. Maybe 10-20%, which would analogous perhaps to the situation in Greece? It's higher in Brahmins and princely castes, however, maybe from 20-30%.

There was a paper last year that anticipated that age estimates based on ancient West Eurasian samples would be a bit off. Harris et al. "Rapid evolution of the human mutation spectrum". I think Reich referenced it a couple of times.

About South Asia: could the Siberian HGs have been the confounding factor? I have long wondered about groups like the Kalash, who in some models appeared to have steppe ancestry on par with north-eastern Europeans, but 'tribal' South Asian Y-DNA profiles. It's the mtdna that appears to be West Eurasian.
 
There was a paper last year that anticipated that age estimates based on ancient West Eurasian samples would be a bit off. Harris et al. "Rapid evolution of the human mutation spectrum". I think Reich referenced it a couple of times.

About South Asia: could the Siberian HGs have been the confounding factor? I have long wondered about groups like the Kalash, who in some models appeared to have steppe ancestry on par with north-eastern Europeans, but 'tribal' South Asian Y-DNA profiles. It's the mtdna that appears to be West Eurasian.

Well, having them as part of the equation certainly deflates the amount of steppe being proposed.

It's weird, as I said in one of these threads, that there's only one R1a among all these ancient samples, and that's from a relatively more recent time, a couple of centuries BC. I mean, who brought the darn steppe genes to the Kalash among others? Amazon type female migrants? :)

Maybe there are samples which the Reich Lab couldn't use because of the disputes with the Indian researchers?

It's a puzzle.
 
According to this study, Mal'ta Boy was R1b-M269 and one of Iron Age Pakistan samples was R1b-U106.

Something is wrong with some haplogroup assignments here. Unless there were Germanics in Pakistan.

Mal'ta Boy is also way too old (~24,000 kya) to be R1b-M269.
 
That may very well be. From what I hear, though, someone got the whole list of calls for this one sample and it does seem to be Z2103, and the archaeological setting seems pretty airtight.

Also, the analysis of the other two bodies in the grave matches that one autosomally. I mean, if he's an intrusive Bronze Age sample and he's Z2103, everyone would expect to see "steppe" in him, yes? The fact is he's just Iranian and Anatolian farmer, no steppe at all, and very similar to the two J2b samples with whom he's buried.

Of course, if there is a problem, everything would have to be re-thought.
 
Two immediate stand-outs:
No BMAC in Indians

R1b M269 6000 BC in Iran, along side J2b
I2327 K1a17a R1b1a1a2a2 Hajji_Firuz_C 5900-5500 BCE Iran

Maciamo was right, not that the usual suspects will acknowledge it.

Are you Maciamo's mom? ;)

Angela, Maciamo provides HG links for R1b in the steppe coming from Iran, but many more millenia ago: he suffers the same kind of steppitis as others, including blurred vision about steppe males getting by rape or import Caucasian brunettes to explain their CHG component. I remember when he said that Villabruna R1b was a steppe wanderer in Italy, now I'm waiting his opinion about other HG R1b wanderers in Germany, Lithuania, Romania... the same was true for R1b south of Caucasus, when I suggested that Yamna R1b-Z2103 was coming from the Caucasus as it was found a R1b guy in Kura-Araxes the answer was... that this guy came from... the steppe!

Game over for R1b from steppe! And now the new R1b guy from Iran, his date is allowing to his M269 bro a chance to participate with the EEF expansion from Anatolia...
 
Are you Maciamo's mom? ;)
Angela, Maciamo provides HG links for R1b in the steppe coming from Iran, but many more millenia ago: he suffers the same kind of steppitis as others, including blurred vision about steppe males getting by rape or import Caucasian brunettes to explain their CHG component. I remember when he said that Villabruna R1b was a steppe wanderer in Italy, now I'm waiting his opinion about other HG R1b wanderers in Germany, Lithuania, Romania... the same was true for R1b south of Caucasus, when I suggested that Yamna R1b-Z2103 was coming from the Caucasus as it was found a R1b guy in Kura-Araxes the answer was... that this guy came from... the steppe!
Game over for R1b from steppe! And now the new R1b guy from Iran, his date is allowing to his M269 bro a chance to participate with the EEF expansion from Anatolia...

Did you even read my work? In my history of R1b I postulated many years ago that R1b-M269 probably crossed over the Caucasus into the Pontic Steppe shortly before 5200 BCE because the first clearly Proto-Indo-European cultures were the Khvalynsk (5200-4500 BCE).

This migration map from 2009 shows R1b moving from the South Caucasus into the Steppe between 6000 and 5000 BCE. I wasn't sure where exactly was the source of R1b-M269 in the South Caucasus, so I added a ? next to Hassuna. No DNA from Hassuna has been tested yet, so it could still be related to Shulaveni-Shomu.

old_neolithic_map.gif


On the very first R1b migration map that I made in 2009 (see this link as I can't copy/paste Flash maps here) I wrote that R1b crossed the Caucasus c. 7000 ybp (5000 BCE), although my arrow shows that it crossed on the western side of the Caucasus toward Maykop, but then I changed my mind a few months later for the above map as I thought it would be more likely that R1b have crossed to the east along the Caspian (just a little detail, it doesn't change the course of history).

I also wrote that J2b2-L283 crossed the Caucasus at the same period and also become a PIE lineage. This was a simple deduction based on the fact that this haplogroup is about 6000 years old (according to Y-Full, but might be 7500 years old I think) and that it is found throughout Europe and South Asia, as well as in the South Caucasus from Eastern Anatolia to NW Iran (exactly the region where Neolithic R1b-M269 would have been found).
 
From Anthrogenica

Hajji Firuz (I2327) R1b calls:

R1b-1 derived call
R1b1-2 derived calls
R1b1a-4 derived calls
R1b1c-2 ancestral calls
R1b1a2-1 derived call
R1b1a1a(R1b-P297)- 5 derived calls
R1b1a2b-1 ancestral call
R1b1a1a2 (R1b-M269)-9 derived calls
R1b1a1a2a1-1 ancestral calls
R1b1a1a2a2 (R1b-Z2103) -3 derived calls
R1b1a1a2a1a-3 ancestral calls
R1b1a1a2a1a1-1 ancestral call
R1b1a1a2a1a2-1 ancestral call
R1b1a1a2a1a1d-1 ancestral call
R1b1a1a2a1a1e-1 ancestral call
R1b1a1a2a1a2c-1 ancestral call
R1b1a1a2a1a1c3-1 ancestral call
R1b1a1a2a1a1g2-1 ancestral call
R1b1a1a2a1a2a5-1 ancestral call
R1b1a1a2a1a2b1-1 ancestral call
R1b1a1a2a2c1a1-1 ancestral call
R1b1a1a2a1a2b1a-1 ancestral call
R1b1a1a2a1a2b1b-1 ancestral call
R1b1a1a2a1a2b3a-1 ancestral call
R1b1a1a2a1a2c1b-1 ancestral call
R1b1a1a2a1a2c1g-1 ancestral call
R1b1a1a2a1a2c1i-1 ancestral call
R1b1a1a2a1a2c1j-1 ancestral call
R1b1a1a2a1a2c1k-1 ancestral call
R1b1a2a1a1b1a1a-1 ancestral call
R1b1a1a2a1a1c2a1d-1 ancestral call
R1b1a1a2a1a1c2b1b-1 ancestral call
R1b1a1a2a1a1c2b2a-1 ancestral call
R1b1a1a2a1a1c2b2b-1 ancestral call
R1b1a1a2a1a2b1c1b-1 ancestral call
R1b1a1a2a1a2c1b2a 1 ancestral call
R1b1a1a2a1a2c1c1b 1 ancestral call
R1b1a1a2a1a2c1g3c 1 ancestral call
R1b1a1a2a1a1c2b1a1 1 ancestral call
R1b1a1a2a1a2a1a1a1 1 ancestral call
R1b1a1a2a1a2a1b1a1 1 ancestral call
R1b1a1a2a1a2b1c1b4 1 ancestral call
R1b1a1a2a1a2c1b1b3 1 ancestral call
R1b1a1a2a1a2c1e2b1 1 ancestral call
R1b1a1a2a1a2c1g1a1 1 ancestral call
R1b1a1a2a1a2c1g1b1 2 ancestral calls
R1b1a1a2a1a2c1g2a1 1 ancestral call
R1b1a1a2a1a2c1k1a2 1 ancestral call
R1b1a1a2a1a1c2b2b1a 2 ancestral calls
R1b1a1a2a1a2c1e2b3b 1 ancestral call
R1b1a1a2a1a2c1e2b4a 1 ancestral call
R1b1a1a2a1a2c1g1b1a 2 ancestral calls
R1b1a1a2a1a2c1k1a1a 1 ancestral call
R1b1a1a2a1a2b1c1b3a1 1 ancestral call
R1b1a1a2a1a2c1e2b2a1 1 ancestral call
R1b1a1a2a1a2c1e2b3a1 1 ancestral call
R1b1a1a2a1a2c1g1b1a1 2 ancestral calls
R1b1a1a2a1a1c2b2a1b1a 1 ancestral call
R1b1a1a2a1a2c1a1a1a1a 1 ancestral call
R1b1a1a2a1a2c1a1a1a1e 2 ancestral calls
R1b1a1a2a1a1c2b2a1b1a1 1 ancestral call
R1b1a1a2a1a2c1a1a1a1a1 1 ancestral call
R1b1a1a2a1a1c2b2a1b1a1b1 1 ancestral call
R1b1a1a2a1a2c1c1b1a2a1b1 1 ancestral call
R1b1a1a2a1a1c2b2a1b1a4a1a 1 ancestral call
R1b1a1a2a1a1c2b2b1a1a1e3a1 1 ancestrall call
 
I agree that there might be mistakes in assigning y dna. Mal'ta can't be that deep in the R1b tree.

The fundemental association between EHG or ANE in general to haplogroup R subclades seems valid to me. CHG has 20% EHG. Iran_N shares more alleles with EHG than WHG. This type of ancestry has a deep history in Iran and Caucasus. So it is possible for R subclades to exist without obvious EHG ancestry in this part of the world.
 
It's a great paper, very extensive. Most of the results are what I expected.

- BMAC was a mixture of J2a (main haplogroup), G2a, L1a, Q1b and R2a, just as I had predicted.

- R1b went south from the Caspian Steppe across Central Asia and settled in Turan/BMAC. Nowadays R1b is far higher than Turkmenistan, Uzbekistan and northern Afghanistan (Hazaras).

- R1a-Z93 went south from southern Siberia along the Tianshan to reach Pakistan and India. Once again that is expected as the R1a-Z93 concentration today are much higher along the Tianshan (Kyrgyzstan, Tajikistan, eastern Afghanistan). What I don't understand is why this paper insists on separating the admixture of those R1a-Z93 people and calling them 'West Siberian HG related' when they in fact originated in the northern forest-steppe of Europe (around Belarus and northern Ukraine).


Here are a few things no one could expect:

- E1b1a1a1c2b1 (aka Z6005) in Mesolithic Iran (12,000 to 8,000 BCE), a Sub-Saharan African clade now found mostly in the far western end of Africa (Gambia, Sierra Leone). Hunter-gatherers did travel a long way... Y-Full says that its parent clade CTS6649 formed 9600 ybp and has a TMRCA of only 6700 ybp, so that is a gross underestimation as it had time to travel over 7000 km and be in Iran about 12,000 years ago.

- E1b1a1a1c2c3c, another SSA lineage, and plenty of Y-DNA A, BT and CT in the BMAC during the Bronze Age. The A, BT and CT could be early Homo sapiens lineages that died out, but why again West African Y-DNA in Central Asia?

- Surprisingly lots of (Levantine?) E1b1b1b2 with also some A0, CT, DE and E in Early Iron Age northern Pakistan. Was there a massive Natufian migration to the region in the Early Neolithic? Or is that more Paleolithic or Mesolithic African hunter-gatherers?

- I2a2a2a in Neolithic Turkmenistan (5000-2000 BCE). Is that an offshoot of a Neolithic culture of Old Europe, a Mesolithic European tribe that ended up in Central Asia, or an early Steppe invader?

- As mentioned before, Mal'ta can't be R1b1a1a2 and it's extremely suspicious to find a Nordic Bronze Age clade of R1b-U106 (S21728, downstream of Z9) in Iron Age Pakistan. Either it's a typo or that sample was contaminated.
 
E1b1a ... amazing.

The fun thing is that no sub Saharan ancestry in Iran Neolithic has been detected. Using formal stats. The same with the Natufians who are E1b1b.

Maybe Basal Eurasian is somehow connected to Africa ?
 
Did you even read my work? In my history of R1b I postulated many years ago that R1b-M269 probably crossed over the Caucasus into the Pontic Steppe shortly before 5200 BCE because the first clearly Proto-Indo-European cultures were the Khvalynsk (5200-4500 BCE).
This migration map from 2009 shows R1b moving from the South Caucasus into the Steppe between 6000 and 5000 BCE. I wasn't sure where exactly was the source of R1b-M269 in the South Caucasus, so I added a ? next to Hassuna. No DNA from Hassuna has been tested yet, so it could still be related to Shulaveni-Shomu.
old_neolithic_map.gif

On the very first R1b migration map that I made in 2009 (see this link as I can't copy/paste Flash maps here) I wrote that R1b crossed the Caucasus c. 7000 ybp (5000 BCE), although my arrow shows that it crossed on the western side of the Caucasus toward Maykop, but then I changed my mind a few months later for the above map as I thought it would be more likely that R1b have crossed to the east along the Caspian (just a little detail, it doesn't change the course of history).
I also wrote that J2b2-L283 crossed the Caucasus at the same period and also become a PIE lineage. This was a simple deduction based on the fact that this haplogroup is about 6000 years old (according to Y-Full, but might be 7500 years old I think) and that it is found throughout Europe and South Asia, as well as in the South Caucasus from Eastern Anatolia to NW Iran (exactly the region where Neolithic R1b-M269 would have been found).
Then you were right the first time. They did cross via eastern black sea coast...to kuban river and up to samarra river.
Something dislodged the Shulaveri and that something came from south /south eastern.
They flee to west and northwest.
Btw, That Kum6 girl was clearly a shulaveri fleeing by 4700bc as I have been saying. So Pie went to north caucasus also to marmara sea region. At least.
Note: I always amed Hussuna as the culprits of attacking the Shulaveri but nowadays not so sure.
Also I think halaf might turn out R1b...
But, as in my "story", after 5000 bc all was defined by the Ubaid , Dalma and Uruk. The farmers , highly pastoral R1b were pretty much gone.
 
Thank you for sharing the full contents of a subject that helped me to better understand. At times each thread seems more important than any other. Yet the locating a spark makes the choices less difficult. Opening the door to discovery add a focus for supporting a point of entry.

Population genetic characterization of ancient individuals
Population Modeling Strategy

In this section, we describe our overall strategy to model the admixture history in our set of3251 newly reported samples described in Supplementary materials section 1, which takes 33252 ordered steps in increasing order of complexity, each drawing on an understanding developed3253 by the previous step.32543255 First line of analysis – PCA and ADMIXTURE3256 We examine qualitative differences in ancestry in the newly reported samples by observing3257 the position of the samples on the West Eurasian and All Eurasian PCA plot (Methods) and3258 the magnitude and number of major components for each sample on the ADMIXTURE plot3259 (Methods). We have chosen the orientation of the West Eurasian and All Eurasian PCA plots3260 to correlate with geography (the genetic patterns mirror geography to an extent), providing an3261 intuitive map of the population structure and to some extent the history
 
I used to give R1a the benefit of the doubt, but no R1a in Chalcolithic and Bronze Age Iran is a very strong indication that R1a is not even Indoeuropean.
R1a-Z93 only gained some relevance in IE Asia thousands of years after the spread of Indo-Iranian, Baltoslav(e)s must have picked up the IE language from their Scythian masters. R1b-M269 and J2 are the major IE markers.
wixAFlq.jpg
 

This thread has been viewed 286687 times.

Back
Top