Population genomics of Bronze Age Eurasia (Allentoft et al. 2015)

[Arame]

This samples are from south shore of Lake Sevan and Syuniq.
This place is far away from Hurrian states. They are not visible in Humanist's map.
Lake Sevan teritory is attested in Urartu period as Uelikuni. The modern Gelarkunik. Many linguists agree that this toponym could have IE etymology. The root is *gel or *wel. The other place name is related to Si-uni-q, Sis-ian ( important town in Siuniq). Armenian traditional historians links this place name with horse. Si -> ji 'horse' in Armenian.
Urartu has a loose control in this places. The south of Syuniq was never controled by Urartu.
The archaeological cultures associated with this aDNA are the Trialeti, Metsamor, Lchashen, Sevan-Uzerlik. Unfortunatly the last three cultures are little known in West, because most literature is still in Russian.

[Kurd]

I've been gone all day, and a lot seems to have happened in terms of discussion.

To address a few points:

1) The BA Armenians aren't (obviously) the direct ancestral population for Iranians. Rather, populations like them constitute around 70%-80% of Iranian genetic ancestry, just like populations similar to Sintashta constitute around 60%-70% of Pashtun genetic ancestry.

2) These stats cut very deep into the phylogeny, so ASI (as understood to be South Asian-specific ENA) has excellent representation via Dai. ASI will have the same relationships to the outgroup populations as any ENA population, which is why Dai does so well (this is all based on how the left populations relate to the right populations. Residual fitting is the name of the game).

3) The Pashtun-Sintashta fits are always the best fits produced by this software, probably because Sintashta are directly ancestral to Pashtuns and company (a fact borne out by the presence of R1a1a lineages in Sintashta which are the exact same lineages found in anywhere from around 50% to 70% of Pashtun males. Also, Sintashta and Andronovo in Allentoft et al. have a hefty share of mtDNA U2 lineages, which constitute the largest share of the modern Pashtun mtDNA gene pool). If they were a broad proxy for general steppe admixture, the fits would be great, but not as amazingly excellent as they are now. Just something to think about.

4) WHG as understood in Lazaridis et al. has a very different meaning now, post-Haak et al. We now know that WHGs were probably admixed between North Eurasian people and some sort of populational group that was basal to North Eurasian people (but they were certainly a stabilized mixture). EHG and MA1 are different spatio-temporal nodes on the "North Eurasian" populational network. The befuddlement a lot of people here are feeling is mainly because they still have the "old model" in mind. The WHG-ANE-ENF model wasn't final.

5) These Pashtun fits are better than what Europeans tend to get, with any combination of modern and ancient populations. If the Pashtun fits are doubted, than the European models are in serious trouble, and the whole Haak et al. paper needs to be reevaluated (and it's main theoretical thrust, in terms of concrete results, has to be thrown into the dustbin). I don't think anyone seriously entertains the undertaking of that route.

6) Finally, qpAdm output overrides anything produced via ADMIXTURE, and to a lesser extent, PCA. As I've written before:

"There is a tendency to look at ADMIXTURE output in these cases, but I think that leads us nowhere. Looking at Chad's ADMIXTURE experiments, that software is very sensitive to slight population/component additions. Drift wrecks havoc with the clusters. At the end of the day, interpreting the output tends to involve a lot of unsubstantiated speculation. Most importantly, there is just no real mechanism for deciding whether the output reflects reality or not.

By contrast, the f4-stat based modelling used in Haak et al. is very robust to things like drift, and doesn't need the "tinkering" involved in ADMIXTURE. Most importantly, the output can be judged on a statistical basis, and is construable as robust, weak, or even as a failure ("infeasible")." In addition, as Shaikorth has noted, ADMIXTURE components overlap.

To summarize, we need to wake up and smell the coffee (), the results are what they are. The only thing we need is Neolithic aDNA from South Central Asia/South Asia, and the picture has been drawn.

Edit: Everest once used ALDER (another piece in the ADMIXTOOLS package) on the HGDP Pashtuns. With Lithuanians, he got Pashtuns to be 66% Lithuanian-admixed. That is quite close to the Sintashta percentages. Here is what he wrote to me:

"Interesting. Alder can calculate admixture % using just 1 reference samples. I tried computer admixture % for Pashtuns using Georgians, Sindhis and then Lithuanians. The admixture using Lithuanian was a whopping 66%."

I agree that the 60% Sintashta + 25% Armenian + 14% Dai is a very good fit for Pashtun, but how can we run with it without aDNA from S Asia.

With discovery of BA or IA aDNA from India, called Ind for the sake of discussion. My prediction is this:

~30% Sintashta + ~40% Ind + ~30% BA Armenian will be a better fit than the above, would make better sense in terms of geography than Dai, and provide a better high k admixture model for Pashtuns.

[Sein]

I agree that the 60% Sintashta + 25% Armenian + 14% Dai is a very good fit for Pashtun, but how can we run with it without aDNA from S Asia.

With discovery of BA or IA aDNA from India, called Ind for the sake of discussion. My prediction is this:

~30% Sintashta + ~40% Ind + ~30% BA Armenian will be a better fit than the above, would make better sense in terms of geography than Dai, and provide a better high k admixture model for Pashtuns.

The only thing I can say is that we must await aDNA from southern Central Asia and South Asia proper. But as of now, this model is the best we can get with the aDNA data we have.

[Arbogan]

Just a side note, but this isn't an equal comparison. The BA Armenians are an ancient population in the vicinity of modern NW Iranians. By contrast, the Burusho are a modern population in the vicinity of modern Pashtuns. Pashtuns will naturally come out as predominantly Burusho (with a big caveat to be noted), since both Pashtuns and Burusho are living end points on a shared deep populational history, with shared source populations for both ethnic groups.

Also, having it based on actual ancient populations that are "prior selected" makes much more historical and biological sense then what ADMIXTURE does. ADMIXTURE simply produces panmictic populations (at varying K levels to boot), and panmixia has never really been at play in the actual history of our species. By contrast, we are actually descended (albeit in very complex and multifaceted ways) from the aDNA samples that are retrieved, and continue to be retrieved, by these research labs.

Now, here is the big caveat to be noted. Basically, qpAdm will probably have the model as "infeasible". That's the beauty of qpAdm, it isn't a Dodecad oracle. If a model can't work, it won't. In this case, I'm sure it won't, since the ENA differential doesn't make sense. Pashtuns have intermediate ENA affinity in relation to Burusho and Sintashta, but Pashtuns and Burusho have nearly identical affinities to North Eurasian and Southwest Asian populations, when one adjusts for the heightened ENA affinity in the Burusho. Sintashta just wouldn't figure into it, but Dai wouldn't either. And the greater ENA pull in the Burusho will make them a poor proxy for Pashtuns in terms of f4-relationships. So, the model will just fail ("infeasible").

Here is an example of a "infeasible" model for Iranians (it doesn't work):

left pops:
Iranian
RISE_baSin
RISE_baArm
Dai

right pops:
Ju_hoan_North
Mbuti
Karitiana
Surui
Ulchi

0 Iranian 8
1 RISE_baSin 5
2 RISE_baArm 7
3 Dai 10
4 Ju_hoan_North 5
5 Mbuti 10
6 Karitiana 12
7 Surui 8
8 Ulchi 25
jackknife block size: 0.050
snps: 525444 indivs: 90
number of blocks for block jackknife: 720
dof (jackknife): 605.965
numsnps used: 335605
codimension 1
f4info:
f4rank: 2 dof: 2 chisq: 2.228 tail: 0.328213251 dofdiff: 4 chisqdiff: -2.228 taildiff: 1
B:
scale 1.000 1.000
Mbuti -0.005 -0.040
Karitiana 0.962 1.221
Surui 0.945 0.877
Ulchi 1.477 -1.318
A:
scale 160.465 2548.580
RISE_baSin 0.083 1.700
RISE_baArm -0.027 0.324
Dai 1.730 -0.066


full rank 1
f4info:
f4rank: 3 dof: 0 chisq: 0.000 tail: 1 dofdiff: 2 chisqdiff: 2.228 taildiff: 0.328213251
B:
scale 1.000 1.000 1.000
Mbuti -0.006 -0.086 1.852
Karitiana 0.962 1.191 0.533
Surui 0.947 0.880 -0.536
Ulchi 1.476 -1.341 0.004
A:
scale 160.562 2598.118 16762.745
RISE_baSin 0.082 1.708 0.278
RISE_baArm -0.042 0.281 -1.709
Dai 1.730 -0.074 -0.055


best coefficients: -0.227 1.198 0.030
ssres:
-0.000230794 -0.000199664 -0.000056880 -0.000137406
-1.359687850 -1.176288948 -0.335096890 -0.809508082

Jackknife mean: -0.183558573 1.153392901 0.030165672
std. errors: 0.228 0.229 0.016

error covariance (* 1000000)
52086 -52247 161
-52247 52655 -408
161 -408 247


fixed pat wt dof chisq tail prob
000 0 2 2.228 0 -0.227 1.198 0.030 infeasible
001 1 3 5.238 0 -0.394 1.394 0.000 infeasible
010 1 3 40.571 8.06223e-09 0.967 -0.000 0.033
100 1 3 3.796 0.284399 -0.000 0.970 0.030
011 2 4 46.234 2.20113e-09 1.000 -0.000 0.000
101 2 4 8.941 0.0625991 0.000 1.000 0.000
110 2 4 3105.744 0 0.000 -0.000 1.000
best pat: 000 0 - -
best pat: 100 0.284399 chi(nested): 1.567 p-value for nested model: 0.210584
best pat: 101 0.0625991 chi(nested): 5.145 p-value for nested model: 0.0233121

## end of run

Basically, one can not model Iranians as Sintashta + BA Armenian + Dai (it's a failure), even though that is an excellent model for Pashtuns. Since Pashtuns and Iranians tend to overlap considerably in ADMIXTURE, this striking difference tells us something about the broad genetic roots of both populations.

I think if you added south-asian population without WHG, but are closer to pre-indo-iranians, you'll get a much better idea for the ancestral group to pashtoos. The thing that seems iffy about the whole buisness, isn't that pashtoos don't have ancestry with link to indo-iranians from the steppe. That is granted. It's the amount of such ancestry. 60% sintashta is tad bit unrealistic. It isn't showing in IBD runs. If pashtoons indeed are 66% lithuanian-like, it doesn't show up in any tools. Save for qpAdm. Which means one of two things:

1. qpAdm is showing a false positives
2. pashtoos might be a related to a west-eurasian group that is closely related to Balto-Slavic-indo-iranian groups, but is not quite like it. But more southern shifted.

[Sein]

I think if you added south-asian population without WHG, but are closer to pre-indo-iranians, you'll get a much better idea for the ancestral group to pashtoos. The thing that seems iffy about the whole buisness, isn't that pashtoos don't have ancestry with link to indo-iranians from the steppe. That is granted. It's the amount of such ancestry. 60% sintashta is tad bit unrealistic. It isn't showing in IBD runs. If pashtoons indeed are 66% lithuanian-like, it doesn't show up in any tools. Save for qpAdm. Which means one of two things:

1. qpAdm is showing a false positives
2. pashtoos might be a related to a west-eurasian group that is closely related to Balto-Slavic-indo-iranian groups, but is not quite like it. But more southern shifted.

I have to go to work, so I'll be able to rejoin the discussion tonight. But for now:

Point 1 is not possible. Looking at all the angles, that wouldn't make any sense. One has to understand how these things actually work, the technical details are essential.

Point 2 doesn't make sense, since qpAdm and formal stats (d-statistics and such) are how one would properly gauge such relations. In addition, if anything like that was at play, the models wouldn't be as amazingly excellent as they are, something would be missing. For example, Corded Ware is somewhat different from Sintashta in formal stats, despite the ADMIXTURE similarity. Something is missing in qpAdm models with Sintashta as being predominantly Corded Ware-derived, the models aren't great. That isn't the case with Pashtuns and Sintashta.

Also, the IBD links are pretty strong, despite the fact that the samples aren't of ideal quality for IBD analysis, and despite the fact that it is very hard to detect IBD over the time scale which currently concerns us. I don't know where this notion stems from.

In addition, Sintashta have the right R1a1a subclades. We can be 100% certain that 50%-70% of Pashtun males are direct descendants of Sintashta and/or it's descendant/related steppe cultures. Even the mtDNA shows links. Taking this into consideration, it is of no real surprise that Pashtuns are predominantly steppe-derived in terms of genome-wide ancestry.

Edit: Also, Pashtuns came out as 66% Lithuanian-admixed in ALDER, not qpAdm. This was more than a year ago. At the time, I dismissed the results. But now, it seems that Everest was on to something.

[parasar]

http://eurogenes.blogspot.com/2015/0...et-al.html?m=0
Sintashta looks very 'European'
In some ways more than Corded Ware.

[Arbogan]

I have to go to work, so I'll be able to rejoin the discussion tonight. But for now:

Point 1 is not possible. Looking at all the angles, that wouldn't make any sense. One has to understand how these things actually work, the technical details are essential.

Point 2 doesn't make sense, since qpAdm and formal stats (d-statistics and such) are how one would properly gauge such relations. In addition, if anything like that was at play, the models wouldn't be as amazingly excellent as they are, something would be missing. For example, Corded Ware is somewhat different from Sintashta in formal stats, despite the ADMIXTURE similarity. Something is missing in qpAdm models with Sintashta as being predominantly Corded Ware-derived, the models aren't great. That isn't the case with Pashtuns and Sintashta.

Also, the IBD links are pretty strong, despite the fact that the samples aren't of ideal quality for IBD analysis, and despite the fact that it is very hard to detect IBD over the time scale which currently concerns us. I don't know where this notion stems from.

In addition, Sintashta have the right R1a1a subclades. We can be 100% certain that 50%-70% of Pashtun males are direct descendants of Sintashta and/or it's descendant/related steppe cultures. Even the mtDNA shows links. Taking this into consideration, it is of no real surprise that Pashtuns are predominantly steppe-derived in terms of genome-wide ancestry.

Edit: Also, Pashtuns came out as 66% Lithuanian-admixed in ALDER, not qpAdm. This was more than a year ago. At the time, I dismissed the results. But now, it seems that Everest was on to something.

Here is an interesting experiment. We could attempt an inverse run. Have sintashta/andronovo ran with corded-ware, a south-central-Asian pop and DAI. You can even use pashtoos as the 2nd derivative. It would show the unidirectional markers shared by sintashta and pashtoons. Then you can see whether it's a good fit. Since we're assuming that pashtoos are derivatives of sintashta or something related. It would show sintashtas affinity to south-central-Asians.

Regarding IBD. Not really. We've seen that modern day Europeans are closer related to WHGs, and overally share more mutational similarities with them, despite being generational drifted via hundreds of generations. It implies a genetic relationship. The same would be true for Pashto's. And the time frame seperates them even less.

[Sein]

Here is an interesting experiment. We could attempt an inverse run. Have sintashta/andronovo ran with corded-ware, a south-central-Asian pop and DAI. You can even use pashtoos as the 2nd derivative. It would show the unidirectional markers shared by sintashta and pashtoons. Then you can see whether it's a good fit. Since we're assuming that pashtoos are derivatives of sintashta or something related. It would show sintashtas affinity to south-central-Asians.

Regarding IBD. Not really. We've seen that modern day Europeans are closer related to WHGs, and overally share more mutational similarities with them, despite being generational drifted via hundreds of generations. It implies a genetic relationship. The same would be true for Pashto's. And the time frame seperates them even less.

(I'm on my 15 min break)

David has tried this, all the models were failures. Sintashta just can't be modeled as Pashtun-admixed, or as admixed with any other South Central Asian population, even though they provide excellent models for modern South Central Asians.

With IBD, I don't know where this notion comes from. I haven't really seen any IBD run with these ancient genomes. But the GedMatch Yamnaya samples provided very substantial hits for the South Central Asians + South Asians who posted their results, if I recall correctly. That is indicative of something.

[Dr_McNinja]

Sintashta was R1a-Z93-Z2124, right? If so, it's extremely likely they are the direct ancestors of most R1a-Z93 Afghans (and a sizable chunk of Indian R1a, mostly from the South/Southwest) since they're mostly Z2124. We'd need next generation level sequencing of ancient Sintashta Y-DNA to confirm it (along with FGC/Big Y of Afghan R1a).

It is weird to think that Admixture might be totally screwing this up though. There's like none of the WHG admixture of Sintashta left, it's not even at Tajik or Haryana levels in what we've seen so far. One explanation is that 3000 years is a long time and while being descended from Sintashta, the Pashtun people were heavily autosomally influenced by other West Asian populations ("Gedrosian"? Either as a late arrival or what the local contribution was. It's ground zero of a WHG "sink", and WHG goes up around it in all directions except west and south). qpAdm could just be picking up the connection that Admixture isn't best equipped for.

The other is that Admixture is totally unreliable here and the WHG autosomal signature is there but being picked up as ENF, ANE, and/or something else. This seems exceedingly unlikely (because, for instance, WHG is found throughout South Asia in a more discernible spread) but it has to be a possibility since we can be fairly certain the end solution should be Sintashta -> Pashtun.

Getting ancient DNA from Iran and Afghanistan would be a huge help.

WHG is still there in higher quantities among some groups like Pamiri Tajiks, there since probably before the population movements of the last 2000 years. Has there been any study on their haplogroups? Are they R1a-Z2124 too?

Makes me wonder when we'll find out who brought L657 into South Asia. I wouldn't be surprised if they're autosomally similar to Sintashta.

[DMXX]

Going through each of these points individually:

1) The BA Armenians aren't (obviously) the direct ancestral population for Iranians. Rather, populations like them constitute around 70%-80% of Iranian genetic ancestry, just like populations similar to Sintashta constitute around 60%-70% of Pashtun genetic ancestry.

See later for a detailed response regarding why the Bronze Age Armenians constitute around 70-80% of Iranian genetic ancestry.

2) These stats cut very deep into the phylogeny, so ASI (as understood to be South Asian-specific ENA) has excellent representation via Dai. ASI will have the same relationships to the outgroup populations as any ENA population, which is why Dai does so well (this is all based on how the left populations relate to the right populations. Residual fitting is the name of the game).

Once more, not taking into consideration the possibility(/probability?) that ASI could be a hybrid component and not simply "South Asian-specific ENA".

Do you have any actual evidence that would, beyond any reasonable doubt, push observers to conclude ASI is merely "South Asian-specific ENA" and is not also a hybrid construct the same way ANI is?

At least one other regular peruser of autosomal genetic profiles in the genetic genealogy community acknowledges this could well be the case.

Do you have any empirically based insights that dismiss this idea?

3) The Pashtun-Sintashta fits are always the best fits produced by this software, probably because Sintashta are directly ancestral to Pashtuns and company (a fact borne out by the presence of R1a1a lineages in Sintashta which are the exact same lineages found in anywhere from around 50% to 70% of Pashtun males. Also, Sintashta and Andronovo in Allentoft et al. have a hefty share of mtDNA U2 lineages, which constitute the largest share of the modern Pashtun mtDNA gene pool). If they were a broad proxy for general steppe admixture, the fits would be great, but not as amazingly excellent as they are now. Just something to think about.

Uniparental markers by themselves are not informative given their liability towards founder effects. This is very relevant in the case of Pashtuns and not a recognised issue with Iranians, as the former still abide by a tribal structure.

Thus, asserting anything resembling a one-to-one correlation (with this frequent assertive language, highlighted) between modern Pashtuns and an (undeniably) ancestral population via uniparental markers and statistical fits runs against conventional wisdom in population genetics.

Actually, it's the first I've seen such an audacious comparison in years.

4) WHG as understood in Lazaridis et al. has a very different meaning now, post-Haak et al. We now know that WHGs were probably admixed between North Eurasian people and some sort of populational group that was basal to North Eurasian people (but they were certainly a stabilized mixture). EHG and MA1 are different spatio-temporal nodes on the "North Eurasian" populational network. The befuddlement a lot of people here are feeling is mainly because they still have the "old model" in mind. The WHG-ANE-ENF model wasn't final.

Given nobody so far has invoked "spatio-temporal nodes" or contrasting models, this is irrelevant. I see no overt "befuddlement", or attempt to marry these findings with prior models in this specific discussion. The three-way model applicable to Europeans is clearly not applicable to Asian populations (although we can safely categorise the steppe groups as very European-like). This is a false equivalence and a distraction from the point below.

Instead, numerous observers (both on this forum and elsewhere) have noticed the rather marked differences in ADMIXTURE component scores between the Eurasian steppe groups with contemporary populations in South-Central Asia. Shaikorth's explanation regarding hidden component scores will surely account for some of the observed disparity, but whether it accounts for all is a quandary which literal interpretors of these qpAdm statistics will have to satisfactorily explain.

The onus is not on those with a non-literal interpretation (i.e. myself; my own view in case people missed this; significant contribution to Afghans, less so for Iranians and Kurds).

5) These Pashtun fits are better than what Europeans tend to get, with any combination of modern and ancient populations. If the Pashtun fits are doubted, than the European models are in serious trouble, and the whole Haak et al. paper needs to be reevaluated (and it's main theoretical thrust, in terms of concrete results, has to be thrown into the dustbin). I don't think anyone seriously entertains the undertaking of that route.

Repetition of prior point(s) (argumentum ad nauseam).

Nobody is contesting the chisq values point towards an excellent statistical fit.

An excellent statistical fit, however, does not (and never will or should) result in wholesale reassessment of ancestral origins in the manner that has been undertaken recently.

6) Finally, qpAdm output overrides anything produced via ADMIXTURE, and to a lesser extent, PCA. As I've written before:
...

qpAdm does look like a superior tool to these. However, qpAdm alone is not a perfect program. I have previously stated some of the overt issues with this tool (e.g. output completely reliant on input populations).

In case some appeal to authority is required to reduce the dependency on re-imagining the ethnogenesis of various Eurasian populations through this tool, some comments from the qpAdm authors regarding its' applicability (their readme file):

qpAdm takes a parameter file in exactly the same format as qpWave. The first population of (popleft) is the target population and the main pourpose of the program is to provide admixture weights, writing the target as a linear combination of the other populations of popleft.

CAVEATS
1) It is important to realize that the answers are invalid if there has been post admixture gene-flow between left and right populations.
2) Weights may be negative. If significantly so, the admixture model is probably wrong.
3) We recommend keeping popright small. If large the covariance matrix of f4-statististics is likely to be poorly estimated. I then don't trust
the computed p-values although the admixture weights seem to usually be reasonable.

Directly fitting the context of this discussion into these two highlighted points:
1) The genetic background of posterior populations in the timeline residing in South-Central Asia are absolutely essential if we are to arrive at any proportions that approach reality. Hypothetical speculations aside, it's abundantly clear that Pashtuns cannot be a three-way mix between Sintashta, Georgians and Dai. The strength of a statistical fit simply does not override either the archaeological, linguistic or (pre)historical narrative.

2) Conversely, there is no satisfactory reason (only special pleading) to assume that Pashtuns have remained genetically static since the hypothetical admixture juncture indicated by qpAdm. Once more, the strength of a statistical fit simply does not override either the archaeological, linguistic or (pre)historical narrative.

To summarize, we need to wake up and smell the coffee (), the results are what they are. The only thing we need is Neolithic aDNA from South Central Asia/South Asia, and the picture has been drawn.

Please read above.

Neolithic remains alone will not yield a definitive answer to the accuracy of these results. Data from both Mesolithic and Iron Age South-Central Asia is also required.

Edit: Everest once used ALDER (another piece in the ADMIXTOOLS package) on the HGDP Pashtuns. With Lithuanians, he got Pashtuns to be 66% Lithuanian-admixed. That is quite close to the Sintashta percentages. Here is what he wrote to me:

Several comments above, you raised the red herring argument of (unspecified) people being "befuddled" with "old models".

Here, you appear to favourably cite a prior correspondence with everest59, despite his very objection now to the literal interpretation of this "Pashtuns are predominantly Sintashta" notion. Reconsider using this as anything resembling evidence.

Given everest59's recent comments disagree with your apparent position, I will respectfully ignore any acknowledgement of authority that was supposed to be generated by your invocation of his ALDER experimentation.

Just a side note, but this isn't an equal comparison. The BA Armenians are an ancient population in the vicinity of modern NW Iranians. By contrast, the Burusho are a modern population in the vicinity of modern Pashtuns. Pashtuns will naturally come out as predominantly Burusho (with a big caveat to be noted), since both Pashtuns and Burusho are living end points on a shared deep populational history, with shared source populations for both ethnic groups.

I am grateful to finally see at least one comment here directly addressing the posts of mine you quoted, rather than ad verbatim repetition or irrelevant distractions from the points previously raised.

The Bronze Age Armenians actually appeared to resemble modern Iranians quite well in several of the ADMIXTURE runs and modern Iranians tend to score the lowest GD with them. This does not contradict my earlier point that they should be not be considered as ancestral to Iranians (there is no substantial archaeological imperative to assume this outside of the Kura-Araxes culture, which was only apparent in the northwestern portion of the country). Thus, their apparent similarity to Iranians is an artefact of their status as West Asians with an appreciable pull towards a north/northeastern trajectory. This is precisely how Iranians and Kurds may be characterised relative to modern Armenians.

Indeed, given the direction of this extra north-shifting ancestry is most likely from the steppelands in both Bronze Age Armenians and Iranians/Kurds, my prior point regarding "masking" of the steppe derived component in Iranians through qpAdm remains fully relevant.

They are, for all intents and purposes, a Bronze Age equivalent of what eventually transpired in the Iranian plateau (as supported by archaeology, linguistics, uniparental markers etc.). No surprise they represented the bulk of Iran's ancestral make-up in these runs, despite not being ancestral to Iranians.

Also, having it based on actual ancient populations that are "prior selected" makes much more historical and biological sense then what ADMIXTURE does.

Agreed. Which is precisely why modelling Pashtuns as a combination of Sintashta (ancient), Georgian (modern) and Dai (modern) is a statistical "wonder" without a definitive context.

Population genetics is not simply a mathematical problem; a historical, linguistic and archaeological context on the backdrop of prior data must be established and/or reconciled.

I note a complete absence of any mention of archaeological data in your rebuttals. This is unusual, given Gimbutas' models for the Eurasian steppe hypothesis (which the Haak and Allentoft papers have nicely confirmed in my opinion) was based on these, as is the very foundation of research on Indo-Iranian prehistory.

Archaeologists have near-consistently recognised that a clear trail from Sintashta or Andronovo through to the Afghan heartland does not exist.

If we were to model Pashtuns as mostly (>60%) Sintashta and assume the urheimat of the historical Pashtuns is the area encompassing southern Afghanistan, one would naturally expect evidence of their pastoralist expansions leading that extent southwards. Merely marvelling at the statistical strength of a model does not replace the clear objections of said model to adjunct disciplines for population genetics. In that respect, your perspective is fully analogous with Bouckaert et al.'s attempts at modelling the Indo-European languages with statistical software.

Two questions based on this proposition:
a) Assuming the above is your currently preferred stance, could you provide archaeological evidence of a mostly Andronovo-derived culture bypassing the BMAC en route towards southern Afghanistan?
b) If the above is not your currently preferred stance, could you provide archaeological evidence accounting for why near-complete culturo-archaeological synthesis took place between the Andronovo tribes and BMAC, yet the Pashtun people are allegedly mostly Sintashta-derived?

Please note this is not a strawman argument given a) Andronovo and Sintashta are commonly recognised as proto-Iranian and proto-Indo-Iranian respectively, and b) Allentoft has provided us with their genetic data.

ADMIXTURE simply produces panmictic populations (at varying K levels to boot), and panmixia has never really been at play in the actual history of our species. By contrast, we are actually descended (albeit in very complex and multifaceted ways) from the aDNA samples that are retrieved, and continue to be retrieved, by these research labs.

With all due respect, technical information regarding ADMIXTURE is, once more, irrelevant here.

The issue, once more, is the assumption that the qpAdm statistics should be interpreted literally.

Now, here is the big caveat to be noted. Basically, qpAdm will probably have the model as "infeasible". That's the beauty of qpAdm, it isn't a Dodecad oracle. If a model can't work, it won't.

I am familiar with qpAdm as I regularly read Eurogenes. The technical explanation will be helpful for others, no doubt.

In this case, I'm sure it won't, since the ENA differential doesn't make sense. Pashtuns have intermediate ENA affinity in relation to Burusho and Sintashta, but Pashtuns and Burusho have nearly identical affinities to North Eurasian and Southwest Asian populations, when one adjusts for the heightened ENA affinity in the Burusho. Sintashta just wouldn't figure into it, but Dai wouldn't either. And the greater ENA pull in the Burusho will make them a poor proxy for Pashtuns in terms of f4-relationships. So, the model will just fail ("infeasible").
...

Your reasoning is appreciated.

However, it is an inconsistent debating style to give a brief description of how things should be for your currently favoured scenario, whereas Iranians are re-modelled using qpAdm.

In order to support your perspective, and in combination with the above requests, would you kindly obtain a run for the HGDP Pashtuns (all, not ten) in qpAdm using Sintashta, the Burusho and Georgians?

Basically, one can not model Iranians as Sintashta + BA Armenian + Dai (it's a failure), even though that is an excellent model for Pashtuns.

For at least the third time, unnecessary repetition of an earlier point (argumentum ad nauseam). I already addressed why the fits appear paltry for Iranians.

Finally, regarding IBD segment scoring; a user called Srkz had produced maps revealing the IBD segment sharing between various Andronovo and Sintashta samples with West and South-Central Asians was largely equivalent with the exception of Pamiri Tajiks, who exhibited a greater number. These have since been taken down. I can verify Arbogan's claim.

I would appreciate more direct rebuttals in future to minimise the degree of (no doubt unintentional) obfuscation shown above.