March 01, 2015

Two observations on the ancestry of Armenians

I was thinking a bit on how to interpret the findings of the new Haber et al. preprint, and especially the idea that "29% of the Armenian ancestry may originate from an ancestral population best represented by Neolithic Europeans." I looked at the globe13 proportions, and strangely enough, I had estimated that the three Armenian samples (Armenian_D, Armenians, and Armenians_15_Y) have 28-29% of the Mediterranean component that is modal in Sardinians. This seems like a curious coincidence which has raised my confidence that Haber et al. is picking something real.

Looking back at my inferences of Armenian ancestry, it seems (according to globe13) to come completely from West_Asian, Mediterranean, and Southwest_Asian. The Mediterranean component seems real enough as it seems to match Sardinians/early European farmers well. I am not so sure about the Southwest Asian component which is modal in Yemen Jews and may represent population-specific drift in relatively recent Arabians. The West_Asian component is bimodal in Caucasus and Gedrosia, so it can't be the result of a very drifted population in either region (unless there is spooky action at a distance). 

Another curious finding is the lack of North_European in a latitudinal "column" of populations from the Yemen, through the Levant to the South Caucasus (Georgians and Armenians). It seems that North_European is the only one of the four major Caucasoid components that Armenians lack to any important degree. There is a rather abrupt change between the South Caucasus (~1%) and the North Caucasus (15-20%). It seems that the Greater Caucasus did act like a barrier to gene flow. The K=4 analysis of the same dataset that produced K=13 (globe13) also shows the same barrier: all three Armenian samples and Georgians have ~0% of "Amerindian" (which is surely correlated to "Ancient North Eurasian" ancestry and via it with North_European), but North Caucasians and Europeans have 4-10%.  It's clear that this influence did not cross the Greater Caucasus, as Armenians and Georgians lack it. 

39 comments:

Mike Thomas said...

Dienekes
How do populations like Iranians and Indians fit in the "North European" component ?

Alexandros HoMegas said...

" It seems that the Greater Caucasus did act like a barrier to gene flow."

Three isolated language families surived in the Caucusus, those mountains must be really good because in Europe many languages and cultures disapperead.

Davidski said...

Are you claiming that Armenians and Georgians lack ANE?

Lazaridis et al. showed that Abkhasians, Druze and Cypriots have ANE, so how can Armenians and Georgians lack it?

In fact, they don't lack it. They have around 15-16% of it, same as Abkhasians. You can even see this on any decent PCA of West Eurasia, because they pull east of the Levant, where ANE is at less than 10%.

PF said...

Obviously something very relevant was going on in and around the Caucuses.

What's throwing me off, Dienekes, is that there are extremely high G2a concentrations on both sides of the divide -- in Caucasians that have a high North_Euro component, and in those that lack it.

I've been pondering the idea that G2a is somehow connected to very early proto West Asians, who diverged into several groups, some becoming EEFs (by subsuming a WHG-rich pop), some becoming coastal Levantines, Anatolians, some moving to Iran, some morphing into proto Caucasians, etc. Anyways I have to think more about this and look at the specifics of G2a phylogeny, so back to the original point...

How to explain this divide? It doesn't match any other invasion pattern of killing the farmer men.

Dienekes said...

@Davidski
Are you claiming that Armenians and Georgians lack ANE?

I am pretty sure that they lack ANE, because ANE is related to "Amerindian" and the "North_European", and they lack that.

Lazaridis et al. showed that Abkhasians, Druze and Cypriots have ANE, so how can Armenians and Georgians lack it?

Abkhasians register some "Amerindian" in my globe13, so they might have some. The other two groups don't. Druze definitely don't, because they register neither North_European nor Amerindian in K=13 and K=4.

Sometimes 0% in ADMIXTUREdoesn't mean absolute 0%, but just minimum among all included populations. But, if ANE/Amerindian/North_European/whatever had crossed the Greater Caucasus and admixed differentially with Transcaucasian populations, there'd be a cline there, but there isn't at all: Armenians/Georgians/Druze are practically the same in those components as Yemen Jews.

@PF
What's throwing me off, Dienekes, is that there are extremely high G2a concentrations on both sides of the divide -- in Caucasians that have a high North_Euro component, and in those that lack it.

Add Neolithic Europeans who have neither North_Euro nor West_Asian but plenty of G2a. Y-chromosomes aren't good standins for autosomal ancestry, e.g., look at the R1b in Neolithic Spanish, Samara H/G, Yamnaya, Armenians, Basques, Chadic, Turkic speakers; completely different autosomal profiles. Whatever the R1b forefather was 20,000 years ago (or whatever the best estimate is), his descendants clearly went around.

@Mike Thomas
How do populations like Iranians and Indians fit in the "North European" component ?

It's all in the spreadsheet. I think there is some "North_European" in both.

spagetiMeatball said...

Dienekes, why do you think north_european is so the only of your globe 13 components that contains ANE?

What about west asian?

In fact we now know why west_asian and North_european branch so close together on TreeMix.


It's because they are both very heavy in ANE compared to other west eurasian components.

What other way is there to explain North_european and west_asian branching out together, and being so close on PCAs compared to mediterranean and south arabian components?

a said...

https://books.google.ca
The History of Herodotus:
"Herodotus has two nations of Pactyans"...... one portion of Armenia, one India"

/books?id=bKICAAAAQAAJ&pg=PA207&redir_esc=y#v=onepage&q&f=false
Caucasus+Gedrosia
Pathan>Armenian

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedEY4Y3lTUVBaaFp0bC1zZlBDcTZEYlE#gid=0
Caspatyrus>Kashmir

http://freepages.genealogy.rootsweb.ancestry.com/~gallgaedhil/haplo_r1b_ht35.htm
Pakistan [Kashmiri] 16.67%

Dienekes said...

@spagetiMeatball

Dienekes, why do you think north_european is so the only of your globe 13 components that contains ANE?

At K=4, Armenians don't have any "Amerindian" worth speaking of (0,0,0.3%). So, you can't break down their West_Asian to something that contains a lot of ANE (which would register as "Amerindian").

In fact we now know why west_asian and North_european branch so close together on TreeMix.


It's because they are both very heavy in ANE compared to other west eurasian components.


North Europeans have a lot of steppe ancestry which includes West_Asian. That's what's pulling them together.

spagetiMeatball said...

Well there's three possibilities:

1) North european has high ANE, west asian has low ANE. North european has low ANE, west asian has high ANE.

This would be difficult to explain why populations having a high amount of both components top drift statistics with the mal'ta boy.

2) Both north european and west asian have high ANE, compared to other west eurasian components.


This explains the high Fst values of north european/west asian to west eurasian components. They have ANE which is diverged from other european/middle eastern populations, reaching it's highest level along a cline to an individual who lived in siberia 25,000 years ago.

Populations along the meditteranean have the lowest level of this component. Populations along the mediterranean have the lowest levels of north european and west asian both.


Dienekes said...

@spagetiMeatball

I'm not following your argument, what "drift statistics" are you talking about?

spagetiMeatball said...

I'm remembering a lot that Davidski put up on his blog, from both Laz et. al
and the new Haak et. al paper.

Unfortunately I only read the supplementary materials but the drift statistics were clear enough. The populations sharing most alleles were always east europeans, various south central asian groups, and north caucasians.


All groups with either very high west asian or north european.

The west eurasian groups sharing least alleles were all meditteranean groups like sardinians, with low west asian and north european components.

Davidski also calculated f3 statistics himself using the HGDP and the ANE genome. If he can link his file here that would be nice.

spagetiMeatball said...

*Sharing most alleles with MA-1 that should be.

spagetiMeatball said...

So both west asian and north european have to be high in ANE.

a said...

https://docs.google.com/spreadsheets/d/1x8pm8sVcHqceiNFJMO082kxaBF5ePr4__bAK05VQRFw/edit#gid=62882571

Eurogenes has already calculated some of the highest ANE adjacent to the Armenians within the Caucasus; most likely same as Yamanya R1b Z2103 5300+/-B.P

Tabassaran.280019%
Tabassaran.286077%
Tabassaran0.275392%

http://en.wikipedia.org/wiki/Y-DNA_haplogroups_by_populations_of_the_Caucasus

Bagvalins-southwestern Dagestan, 67% R1b most likely R1b Z2103

Tabasarans are an ethnic group who live mostly in Dagestan-39.5%

Dienekes said...

@spagetiMeatball

Unfortunately I only read the supplementary materials but the drift statistics were clear enough. The populations sharing most alleles were always east europeans, various south central asian groups, and north caucasians.


I'm talking about Armenians here, all the groups you mention also have North_European, not only West_Asian. Europeans also have Med that South Asians don't. You can't just compare them directly and attribute their differences to West_Asian or North_European, without considering all their components together.

The one thing that's sure about ANE is that it's related to "Amerindian". Populations that seem to have a lot of ANE like northern Europeans and northern Caucasians have "Amerindian" at K=4, but Armenians, Yemen Jews and many others south of the Caucasus don't.

Dienekes said...

@a

Interesting about Tabasarans. THere were no Tabasaran data when I did globe13 but from what you're saying I'd be surprised if they didn't have a lot of ANE/North_European like others from Dagestan.


spagetiMeatball said...

Okay, so it's clear that North_European and west_asian go together in eurasia quiet frequently, but how can you be sure as to which of them influenced the other? They peak rather disparately geographicallly. So maybe a third group added something to both.

Why can't that third group be ANE?

Btw, armenians are something like 40% west asian? If we associate west asian with ANE, and it looks like it is, then.....

Dienekes said...

Btw, armenians are something like 40% west asian? If we associate west asian with ANE, and it looks like it is, then.....

Your argument is based on a faulty premise ("if we associate west Asian with ANE").

The one thing we can associate with ANE is "Amerindian" because ANE is the only west Eurasian stuff that Amerindians got. Armenians don't have any "Amerindian", so I doubt they have ANE either.

Matt said...

Seems like a good question as to why some of the populations modeled as having an ANE contribution show Amerindian component at this low K4 while others, often don't and instead take on the Asian component....

Possibly some of this ANE is dissolved into the Asian component at this level of resolution?

Now, actually, the Amerindian component at K4 seems to relate more strongly in Europeans to an excess of hunter gatherer ancestry, rather than ANE exactly, based on the patterns we've seen in the published papers.

ANE has some relationship with European hunter gatherers, so that would fit.

For ex' on this, Greek_D shows 94.8 "European", 2.2 "Asian", 0.4 "African", 2.5 "Amerindian" vs Pais_Vasco 94.9 "European", 5.1 "Amerindian" and French_Basque 95.4 "European", 4.6 "Amerindian".

Yet Pais_Vasco seems singularly unlikely to have been more effected by ANE bringing movements from the steppe or to be related to ANE based on isolation by distance, compared to the Greeks, Indo-European speakers relatively close to wherever the heck it is ANE is likely to have expanded from.

What Pais_Vasco probably does have a bit more of though, is WHG ancestry. The presence of Amerind component here in Sardinians, supposedly ANE free, also indicates this.

No doubt there is still some ANE related signal to the Amerind component here, the shape of the distribution just seems more like a European HG signal. Obviously it is hard to tell these apart as European HG ancestry shows a correlation with whatever ANE ancestry in Europe, the case of Southwest Europeans maybe helps tease these apart.

Btw, Dienekes, Southwest Asian seems to be found in the East Africans as well, so if it is the result of drift, must be drift / founder effect common to East Africans and Arabians.

Gary Moore said...

There are some interesting observations about the neighbors of the Armenians:

Compare with this map of the Hittite Empire at its peak with the distribution of YHG Q in Anatolia and the Middle East:

Hittite Empire: http://en.wikipedia.org/wiki/Hittites#mediaviewer/File:Map_Hittite_rule_en.svg

YDNA-Q distribution map: http://www.eupedia.com/europe/Haplogroup_Q_Y-DNA.shtml

Note that the peak distribution of YHG-Q peaks in the area of the core of the former Hittite homeland. Significantly, the frequency of YHG-Q falls off in the Caucasus itself, in the region of Georgia and Armenia. YHG-Q reappears along the eastern shore of the Black Sea farther to the north.





Unknown said...

Dienekes -
Okay so the David Reich paper says that Yamnaya models best as “a mixture of EHG and present-day Near Eastern populations like Armenians (Z = -6.3; SI7).”

So, when these guys say “Armenians” they don’t mean the same thing as your Armenians -- the Armenians who are showing such relatedness to Sardinians and Neolithic Europeans.

Because that should actually draw Yamnaya much closer to Neolithic Europeans.

We’re seeing different data defining “Armenians.”

Arch Hades said...

So where did mainland Greeks get their ~25% North_European component from on your Globe13 then? Medieval Slavs?

According to Patterson et al. 2012='s admixture models, Neolithic farmers near completely replaced the indigenous Hunter-Gatherers in Southern Europe. Only ancient DNA from the far South of Europe and the Aegean can confirm this I suppose, but I would not expect to see much haplogroup R1 around the Aegean or North_ European autosomal components on your ADMIXTURE runs until Bronze age and mainstream Indo-European timescale. Certainly for Southern Europe, the Yamna/Kurgan model for PIE makes solid sense IMO.

The Caucuses and Near East a whole different monster though.


I would be very interested to see where the Yamna come up on your Globe13, probably lots of North_European and West_Asian.

Mike Thomas said...

Dienekes do u think u will construct new dodecad- type Globe data incorporating the new host of aDNA ?

Gary Moore said...

Dienekes wrote:

"The K=4 analysis of the same dataset that produced K=13 (globe13) also shows the same barrier: all three Armenian samples and Georgians have ~0% of "Amerindian" (which is surely correlated to "Ancient North Eurasian" ancestry and via it with North_European), but North Caucasians and Europeans have 4-10%. It's clear that this influence did not cross the Greater Caucasus, as Armenians and Georgians lack it."

Wouldn't the presence of YDNA Q in Anatolia appear to contradict this conclusion? Moreover, it is not clear if all of the YDNA Q in Anatolia is related to the Turkish invasion. For instance, Q1a2b2 (L938) is observed in Lithuania, Britain, and Portugal - not regions associated with Turks - as well as Anatolia. It appears in all likelihood that YDNA Q may have been associated with the Hittites as well as the original speakers of Anatolian languages. The discontinuous distribution of Q in the region along the Black Sea would in turn appear to support a Steppe origin for the speakers of Anatolian languages.

Traces of NE Asian ancestry have been found in other Middle Eastern populations. According to the Europedia article on YHG Q: "Autosomal analyses have confirmed that all Levantine people (Jews, Lebanese, Palestinians, Syrians) possess about 0.5% of Northeast Asian (Mongoloid) admixture. Since these populations lack Mongoloid mtDNA, the presence of Northeast Asian admixture can only be explained by the 2% of Q1b among Levantine men, the only paternal lineage of Mongoloid origin in the region."

It should be noted that YDNA Q1b has been detected in Native American populations in South America. However, the researchers who collected these samples choose to exclude them from their study, citing the presence of Jewish rubber tappers in the Amazon region and the association with SW Eurasian populations, even though it is widely accepted the that this particular YDNA type entered the region from Central Asia fairly recently (that is, around the time of the Bronze Age).

Kurti said...

Another curious finding is the lack of North_European in a latitudinal "column" of populations from the Yemen, through the Levant to the South Caucasus (Georgians and Armenians).


Georgians don't lack it, in Calculators such as Globe13 or v3 they have frequency from 3-12%. While no matter which calculator you use Armenians almost completely lack this component.

Kurti said...

Also Levantines don't lack it, they have 5% of it.

Kurti said...

sorry for the tripple post.

@Dienekes

The break up of the Mal'ta genome showed ~26% of Caucasus_Gedrosia component. That surely doesn't mean Mal'ta was CG admixed but that Caucasus_Gedrosia received ANE admixture. How can North Caucasians with less "North European" have more ANE than North Europeans who have more of this "North European" component?

Obviously Armenians and even more so Georgians have ANE admixture this is visible even from the Lazaridis and Reich papers.

Reichs own statement: "Yamna was East European H&G + 50% Near Eastern farmers rich in ANE.

Gökhan said...

I totally agree with Dienekes

ANE companent in K8 includes Ancient North India (So called "gedrosa") admixture which is not completely Euroasian,but probably derived from Caucasian . That make confusions with understanding the people interaction between cau and Gedrosa before Indo- European invasions. Interaction between Gedrosa and caucasia might be even before neolotic. But probably iranian invasions in bronze age might brought more Gedrosa companent in mesopotamia.

If we seek the neolothic composition in modern days there should be some isolotion areas in Caucasia and Europe. Best candidates for Caucasia - Trans caucasia are Pontic Mountains, Svaneti and Anatolian platue. I think that isolation levels are high in those places. In Eurpoe relatively isolated places are islands. Cyprus and Sardinia may be relatively islolated rather then rest of the europe.

Not only dodecad but also MDLP and Harappa should also be overviewed to understand interaction between those companents.

according to Fst analysis of Harappa Companents

Caucasian have

Mediterranean 0,034
NE Euro 0,036
SW Asian 0,044
Baloch 0,046
S Indian 0,077
NE Asian 0,113
E African 0,114
SE Asian 0,115
Siberian 0,116
Beringian 0,136
W African 0,139
American 0,143
Papuan 0,173
Pygmy 0,185
San 0,198

While American companent one of the highest fst with caucasia Baloch has one of the lowest fst with cau. That shows we can not include gedrosa totally into ANE companent.

Dienekes said...

@Kurti
Obviously Armenians and even more so Georgians have ANE admixture this is visible even from the Lazaridis and Reich papers.

According to Lazaridis et al. (Table 1) the lowest f3-stat for Armenians and Georgians is EEF+South Asian, not EEF+ANE. According to Haber et al. (Table 1) the lowest f3-stat for Armenians is Sardinians+Central/South Asians. According to Haak et al. (Supplementary info, p. 46) it is LBK+Sindhi. I did some f3-stats on Armenians before, and got Sardinian+Velamas_M.

All these are similar, and none are EEF/Sardinian plus MA-1/Native Americans, it's always EEF/Sardinian plus Central/South Asian.

I'm not gonna repeat myself ad nauseam. Armenians don't have "Amerindian" admixture which is the correlate of "ANE", so they don't have ANE.

@Arch Hades
So where did mainland Greeks get their ~25% North_European component from on your Globe13 then? Medieval Slavs?

Greece is accessible from the north via river valleys, so I'm not surprised that they have North_European. Whether all or part of it is from Slavs is TBD.

Basant said...

The fact that the Armenians and other peoples of the Southern Caucus almost completely lack the Northern European component is very pertinent to the questions of the PIE homeland, and Ind-European migration to the Indian Subcontinent and Iran.

The Northern European component is seen at low rates among various Indo-Iranian groups, but higher than the ~1% seen among armenians. If one assumes that the European component indicative of the Indo-European migrations, then the lack of it in the Southern Caucus region would seem to go against a migration from Anatolia or the Potic Step to India and Iran via the Caucus and the Middle East, but rather provide support for a northerly route for the Indo-Iranians around the Caspian consistent with BMAC. This may have some further effect on the Anatolia vs. Pontic Steppe homeland debate.

However, I am somewhat confused by the fact that the Armenians - who are an Indo-European group which has remained close to the PIE homeland, and has seemed to have not mixed much with other groups - lacks the Northern European component almost completely. This may indicate that PIE people also lacked the component, and that the Northern European components among the Indo-Iranians, and central-asian mummies etc, come from some other earlier migration. This observation may also lend credence to the Anatolian hypothesis in which the Proto Indo-Europeans would be expected to have less of a Northern European component.

Or, possibly, some subset of the PIEs went to the Pontic Steppe, became Northern Europeanized, and later migrated to Iran, India etc, while the Hittites, Armenians, etc were left behind and did not have the Northern component.

Either way, thanks for this very thought provoking news.

Basant said...

Also, I find it interesting that I (from Northern India) had a 2% Amerindian component in Genographic 2.0 analysis where as the Armenians, protected by the Caucus Mountains do not. I also wonder what the implications of this are.

n/a said...

Dienekes,

I explained this to you three years ago (at a time when you were claiming your ADMIXTURE results were "consistent with a mainly Neolithic origin of West Eurasian populations, and incompatible with their pre-LGM differentiation."): there's no reason to believe your intra-WEurasian admixture components have any particular correspondence to actual ancient populations.

Since that time, ancient DNA results have proven your "womb of nations" model wrong, as I fully expected they would. We now know that West Eurasians are descended from mixtures of populations that are much more highly differentiated than your ADMIXTURE components. And yet you're still trying to use these components to salvage some sort of out-of-Anatolia argument.

Armenians don't have "Amerindian" admixture which is the correlate of "ANE", so they don't have ANE.

It doesn't matter if there's a correlation between your "Amerindian" component and "ANE" admixture. Who says the correlation is even causal? In any case, the correlation is definitely not 1:1, and absence of your "Amerindian" component does not demonstrate absence of "ANE".

Balaji said...

I agree with Dienekes that there is a discontinuity between the North and the South Caucasus. However, I disagree with him that ANE in Armenians is 0. According to Lazaridis's, Table S14.15, the Abkkhasians have ANE of 18.6%. Georgians and Armenians can be expected to have the same. Regarding Native American-like components at K=4, the recent Haak ADMIXTURE figure may be good to study.

http://biorxiv.org/content/biorxiv/suppl/2015/02/10/013433.DC1/013433-1.pdf

(1) European hunter gatherers such as La Brana, Loschbour and HungaryGamba who have no ANE but only WHG all have some of the pink Native American component at K=4.
(2) Armenians, Georgians and Abkhasians do not have this component. But based on Lazaridis, we can expect that they in fact do have ANE.

Aram Palyan said...

Abkhazia is much more accessible from the North than Georgia and Armenia.

Dienekes said...

@n/a

The point of the "womb of nations" hypothesis is that West Eurasian populations are not very strongly differentiated despite their great geographical distances (I made the point that Gedrosia and Northwest_African had a similar Fst as Gedrosia and South_Asian).

The explanation for this observation was that the different west Eurasian populations had ancestry from a common source that acted as an "attractor", disallowing them from becoming too differentiated from each other, and this attractor were Neolithic migrations from the Middle East. I think that overall this continues to be true. Note that the "womb of nations" hypothesis never claimed that all the ancestry of different populations came from the Middle East, and I listed a whole number of pre-Neolithic populations that people from the womb of nations absorbed in different areas.

You are of course, right that ADMIXTURE components are not as differentiated as the admixing populations, but that is not relevant to the question addressed by the "womb of nations" hypothesis. Because of mixture, Sardinians and Lithuanians have Fst=0.021 (Haak et al. 2015), and the corresponding Mediterranean and North European components have Fst=0.046 ("womb of nations" post), while LBK and WHG/EHG have Fst greater than .08.

So, Fst between ADMIXTURE components underestimates Fst between ancient populations, but is higher than Fst between modern populations. By observing that Fst between ADMIXTURE populations is relatively low, one is not denying the existence of differentiated ancestral populations, but rather that this differentiation must have been "broken" by gene flow. Without this gene flow, Northwest Africans and South Asians would have continental level Fst. In both of them, Fst is inflated by admixture (with Africans and ASI), and deflated by the common recent Neolithic ancestry.

Grey said...

"According to Patterson et al. 2012='s admixture models, Neolithic farmers near completely replaced the indigenous Hunter-Gatherers in Southern Europe."

Maybe but I think stage 1 would have been the Hgs retreating to refuge terrain the farmers couldn't use.

"So where did mainland Greeks get their ~25% North_European component from on your Globe13 then?"

If the original HGs retreated from the farmers into the mountains then if the IE incursion knocked over the farmers for a while that may have allowed the HGs to bounce back.

Slumbery said...

Grey

""So where did mainland Greeks get their ~25% North_European component from on your Globe13 then?"

If the original HGs retreated from the farmers into the mountains then if the IE incursion knocked over the farmers for a while that may have allowed the HGs to bounce back."
---------------
Or a significant part of the ancestry showing-up as "North European" came directly from the IE incursion itself (later a bit more increased by the autosomal impact of the Slavic migration too).

Gary Moore said...

Barant wrote: "The Northern European component is seen at low rates among various Indo-Iranian groups, but higher than the ~1% seen among armenians. If one assumes that the European component indicative of the Indo-European migrations, then the lack of it in the Southern Caucus region would seem to go against a migration from Anatolia or the Potic Step to India and Iran via the Caucus and the Middle East, but rather provide support for a northerly route for the Indo-Iranians around the Caspian consistent with BMAC. This may have some further effect on the Anatolia vs. Pontic Steppe homeland debate. "

Probably the Armenians are almost entirely descended from original Bronze Age Anatolian inhabitants with little admixture with the Hittite-related populations. Evidence from Scandinavia as well as the Hittite homeland in Anatolia suggests that different variants of YHG Q may in fact be a marker for the earliest Indo-Europeans. This is consistent with linguistic evidence which indicates that *PIE shows influences of Northeast Siberian - Native American languages, such as pronoun systems. There is a lot of data to support a model in which the precursor on Indo-European languages formed in the context of a sprachbund involving North American and Dene-Yeniseian languages.


'this'

Ket: kiˑ (near), kiˑ-rʸ, kiˑdǝ, kiˑrʸǝ, kiˑrʸ (masc.), kˈi-rʸe (fem./neut.), kˈi-nʸa (pl.)

Iroquoian: Mohawk kí:ken/ken' í:ken / Cherokee hina (h > k)

*PIE: 'this' *kos, *koh₂, *kod



'that'

Ket: tuˑ (intermediate distance), tuˑ-rʸ, tuˑdǝ, tuˑrʸǝ, tuˑrʸ (masc.), tˈu-rʸe (fem./neut.), tˈu-nʸa (pl.)

Iroquoian: Mohawk hí:ken/tho í:ken / Cherokee na

*PIE: 'that' *só, *séh₂, *tód (s > t)

Note that the modern Cherokee form for 'he, she, it' - nahi - closely follows Armenian na 'he, she, it' as well as Germanic hiz (“this, this one”)


'old'

Ket: sīn / Tlingit shaan (of people) / Navajo sání

*PIE: 'old' *senh₁ó-

Simon_W said...

To my understanding, South Asians, even in the very south, have some shared drift with MA-1. And in the ADMIXTURE analysis in Lazaridis et al. 2014, MA-1 has a little bit of the South Asian component. I'm sure this is related with haplogroup R2 in South Asia.

Regarding Armenians, it can't be denied that the North European admixture there is very low. But the ADMIXTURE analysis in the Haak et al. preprint shows a presence of the hunter-gatherer component quite clealry up to K=15. Now it could be argued that this is unrealistic because they also have a strong South Asian centered component up to K=15. At K=16, the West Asian component appears and replaces the South Asian component. But still there is a faint presence of the hunter-gatherer component up to K=17. At K=18 to 20, this has disappeared, possibly having been swallowed up by the West Asian component.