search this blog

Tuesday, June 30, 2020

The precursor of the Trojans

Who remembers kum4 from Omrak et al. 2016? I'm pretty sure now that this individual packs a lot of ancestry from the Pontic-Caspian (PC) steppe.

If so, that's a big deal, because her Chalcolithic (or Late Neolithic?) burial was located at Kumtepe. That is, in the same part of Anatolia as the later settlement of Troy, which may have been founded by early Anatolian speakers from Eastern Europe (see here).

The qpAdm mixture models below, featuring kum4 and the likely older kum6, also from Kumtepe, are based on qpfstats output. qpfstats is a new program from the David Reich Lab specifically designed to help analyze low coverage ancients (see here). And kum4 is certainly that.

RUS_Progress_En 0.383±0.114
TUR_Barcin_N 0.617±0.114
chisq 7.868
tail prob 0.247957
Full output

IRN_Seh_Gabi_C 0.325±0.150
TUR_Barcin_N 0.675±0.150
chisq 14.736
tail prob 0.0224096
Full output

RUS_Progress_En 0.121±0.042
TUR_Barcin_N 0.879±0.042
chisq 21.790
tail prob 0.00132149
Full output

IRN_Seh_Gabi_C 0.283±0.059
TUR_Barcin_N 0.717±0.059
chisq 6.289
tail prob 0.391566
Full output

Indeed, kum4 and kum6 offer just ~10,000 and ~100,000 "valid SNPs", respectively (see here). However, if nothing else, the results are clearly not random.

For one, because they fit the expected pattern, with the likely older individual lacking ancestry from the PC steppe (her model with RUS_Progress_En shows a weak statistical fit). Moreover, the qpAdm mixture ratios align almost perfectly with the results in my Principal Component Analysis (PCA) of ancient West Eurasian genetic variation. Coincidence?

See also...

Perhaps a hint of things to come

Saturday, June 27, 2020

Major updates to ADMIXTOOLS

An important message from Nick Patterson:

Dear Eurogenes bloggers,

Many of you use ADMIXTOOLS and you might like to know that there is a new release on github [LINK] with some important enhancements.

From the README

*** NEW ***


Version 7.0 has numerous upgrades.

a) Two new executables --qpfstats qpfmv allow precomputation of f-statistic basis. This can greatly reduce computation costs.
b) qpAdm, qpWave, qpGraph support qpfstats output as input.
*** This is a much improved way of running with allsnps: YES. ***
c) A new experimental feature of qpGraph (halfscore: YES) allows comparison of 2 phylogenies + a (weak) goodness of fit score. Be careful if running with a large number of populations and consider reducing block size say blgsize: .005


Note that several of the new ideas implemented in version 7.0 were developed collaboratively with Robert Maier, who has implemented them along with the great majority of other ADMIXTOOLS functionality in R: See
Executables run fast, and it has features not available in this C version, such as interactive exploration of graph phylogenies.
A manuscript describing the algorithmic ideas and providing documentation of the methods is in preparation.

qpfstats is the most important new executable. This estimates f-statistics and covariance on a basis.

a) This can be passed into other programs of the package without having to reaccess the genotype files, greatly speeding the computations.
b) In allsnps: YES mode a new computation is carried out (explained in qpfs.pdf) that is much more logical when there is a lot of missing data. Sometimes standard errors are greatly reduced.
qpfstats can be used with up to 30 populations. Much beyond that the output files become large.

As usual there may be bugs...

Nick Patterson 6/27/2020

Update 29/06/2020: As pointed out above, qpfstats is the most important new executable. Indeed, Nick Patterson now recommendeds that qpAdm analyses run with the allsnps: YES flag should be based on qpfstats output.

Several of my recent blog posts featured qpAdm models run with the allsnps: YES flag, but they were based on genotype data because obviously I didn't know anything about qpfstats at the time.

So I went back and ran some of these models again, just to make sure that they were still relevant. Below are three examples which you can compare to the original analyses here, here and here, respectively.

RUS_Maykop_Novosvobodnaya 0.281±0.042
TUR_Arslantepe_LC 0.719±0.042
chisq 10.923
tail prob 0.449752
Full output

RUS_Vonyuchka_En 0.137±0.031
TUR_Buyukkaya_EC 0.863±0.031
chisq 15.074
tail prob 0.0889099
Full output

RUS_Progress_En 0.083±0.020
UKR_N 0.917±0.020
chisq 6.825
tail prob 0.65538
Full output

As far as I can tell, they're very similar to the original runs, which is a relief, because it means that the conclusions in my blog posts still make sense.

Wednesday, June 24, 2020

Armenian Highland population prehistory

A new preprint at bioRxiv claims that some sort of large-scale population movement resulted in the spread of Sardinian-like ancestry into both the Armenian Highland and East Africa during or just after the Middle-Late Bronze Age. See Hovhannisyan et al. here.

In all seriousness, my suggestion is that the authors should familiarize themselves with the scientific concept of the sanity check and then try again.

For what it's worth, here's a brief outline of the population history of the Armenian Highland based on what I've learned about the topic from ancient DNA in recent years:

- overall, the Neolithic populations of the Armenian Highland were surely very similar to the Caucasus_lowlands_LN samples from what is now Azerbaijan from the recent Skourtanioti et al. paper (see here)

- Chalcolithic era migrations from the Pontic-Caspian steppe and/or the North Caucasus introduced steppe ancestry to the Armenian Highland, bringing at least some of its populations closer genetically to those of Eastern Europe (a somewhat outdated but still useful blog post about this subject is found here)

- population expansions during the Early Bronze Age associated with the Kura-Araxes cultural phenomenon, which may have originated in what is now Armenia, resulted in a resurgence of indigenous Caucasus hunter-gatherer (CHG) ancestry across the Caucasus, as well as its spread to many other parts of West Asia (see here)

- another significant pulse of Eastern European admixture affected the Armenian Highland during the Middle-Late Bronze Age and Early Iron Age (see here)

- it's not yet completely clear what happened in the Armenian Highland during the Iron Age in terms of significant genetic shifts, due to the lack of ancient human samples from the region dating to this period, but it's still possible that the speakers of proto-Armenian arrived there from the Balkans at this time

- the present-day Armenian gene pool is the result of the processes described above, as well as later events, such as those associated with the Urartian and Ottoman Empires.

Indeed, it's probably not a coincidence that present-day Armenians cluster more or less between the prehistoric populations from the Armenian Highland and surrounds in the Principal Component Analysis (PCA) below.

To see a more detailed and interactive version of the plot, copy paste the data from the text file here into the relevant field at the Vahaduo Globabl25 PCA Views here.



See also...

Armenian confirmation bias

Perhaps a hint of things to come

Understanding the Eneolithic steppe

Tuesday, June 16, 2020

Like three peas in a pod

One of the most interesting questions still waiting to be answered by ancient DNA is where exactly did the ancestors of the present-day European and South Asian bearers of Y-haplogroup R1a part their ways? Indeed, the answer to this question is likely to be informative about the place and time of the split between the Balto-Slavic and Indo-Iranian language families.

I was doing some reading today and discovered that the peoples associated with the Bronze Age Fatyanovo-Balanovo and Unetice archeological cultures shared strikingly similar metalwork, despite being separated by well over two thousand kilometers of forest and steppe. Apparently, this similarity is especially pronounced in the metalwork of the Unetice culture from what is now Slovakia (see Ancient Metallurgy in the USSR: The Early Metal Age, page 136).

S11953 is currently the only sample from Slovakia associated with the Unetice culture (Sirak et al. 2020). There are no Fatyanovo-Balanovo samples available yet. However, as far as I can tell, I0432 from Samara, Russia, should be a decent stand in (Mathieson et al. 2015).

Of course, both S11953 and I0432 belong to Y-haplogroup R1a. Moreover, S11953 belongs to a typically Balto-Slavic subclade of R1a, while I0432 belongs to a closely related subclade that is dominant nowadays among the Indo-Iranian speakers of Asia.

S11953 is younger than I0432, but this doesn't necessarily mean that his ancestors arrived in East Central Europe from deep in Russia during the Bronze Age. Indeed, the opposite is more likely to be true. That is, I0432 is probably the recent decedent of migrants from somewhere near the North Carpathians, because he shows elevated European Neolithic farmer ancestry compared to earlier ancients from the Samara region (see here).

Below is a Principal Component Analysis (PCA) showing how S11953 and I0432 compare to each other in the context of ancient West Eurasian genetic variation. Obviously, they're sitting in the same part of the plot, which suggests that they harbor very similar ratios of ancient genetic components and probably share relatively recent ancestry. The relevant PCA datasheet is available here.

I've also highlighted myself, Davidski, on the plot. That's because I share the same Balto-Slavic-specific subclade of R1a with S11953 and, in terms of overall ancestry, I'm similar to both S11953 and I0432. Moreover, I'm the speaker of Polish, which is a Balto-Slavic language. What are the chances that we're dealing here with a remarkable string of coincidences? Indeed, was the North Carpathian region perhaps the homeland of the language ancestral to both Balto-Slavic and Indo-Iranian?

However, please note that there's nothing unusual or remarkable about my ancestry. The vast majority of people of Central, Eastern and Northern European origin - that is, mostly the speakers of Balto-Slavic, Germanic and Celtic languages - would also land in this part of the plot.

See also...

On the doorstep of India

Y-haplogroup R1a and mental health

The mystery of the Sintashta people

Saturday, June 13, 2020

The Abashevo axe did it (Mednikova et al. 2020)

Open access at the Journal of Imaging over at this LINK. From the paper, emphasis is mine:

A massive bronze battle axe from the Abashevo archaeological culture was studied using neutron tomography and manufacturing modeling from production molds. Detailed structural data were acquired to simulate and model possible injuries and wounds caused by this battle axe. We report the results of neutron tomography experiments on the bronze battle axe, as well as manufactured plastic and virtual models of the traumas obtained at different strike angles from this axe. The reconstructed 3D models of the battle axe, plastic imprint model, and real wound and trauma traces on the bones of the ancient peoples of the Abashevo archaeological culture were obtained. Skulls with traces of injuries originate from archaeological excavations of the Pepkino burial mound of the Abashevo culture in the Volga region. The reconstruction and identification of the injuries and type of weapon on the restored skulls were performed. The complementary use of 3D visualization methods allowed us to make some assumptions on the cause of death of the people of the Abashevo culture and possible intra-tribal conflict in this cultural society. The obtained structural and anthropological data can be used to develop new concepts and methods for the archaeology of conflict.


Human skeletal remains from excavations of the Pepkino burial mound bear many traumatic wounds on the skulls and postcranial bones (Figure 4). The primary hypothesis is that young men of the Abashevo culture fell at the hands of enemies, which were the representatives of another tribe or culture [14,16]. After their discovery in the XX century, the skulls of killed people of the Abashevo culture were restored using anthropological paste, including beeswax.


A simple explanation for obtaining such injuries is the conclusion that the victim stood face to face with their assaulter and tried to back away from the battle axe, but fell and received other lethal wounds. The superficial trauma by the battle axe as well as serious damage to a bone structure and deep cracks in the skull are visible in the upper part of the model.


The comparison of the real bronze axe with the model obtained from molds indicates their complete identity and the belonging of these axes from different archaeological sites of the Abashevo culture to the same cultural group. This conclusion may indicate intra-cultural conflict among the Abashevo people. As a final note, the presented results of quite diverse imaging methods indicate a new direction in the archaeology of conflicts and the applicability of 3D modeling methods to identify both weapons technologies and the specifics of the use of these weapons to injure humans.


Mednikova et al., The Reconstruction of a Bronze Battle Axe and Comparison of Inflicted Damage Injuries Using Neutron Tomography, Manufacturing Modeling, and X-ray Microtomography Data, J. Imaging 2020, 6(6), 45;

See also...

Tuesday, June 9, 2020

Maykop ancestry in Copper Age Arslantepe

At least four individuals from the Late Chalcolithic (LC) burial site of Arslantepe show ancestry typical of the population associated with the contemporaneous Maykop culture in the North Caucasus. They are ART018, ART020, ART027 and ART039 from the recent Skourtanioti et al. paper. I've labeled them TUR_Arslantepe_LC_Maykop in my qpAdm mixture model below:

RUS_Maykop_Novosvobodnaya 0.318±0.041
TUR_Arslantepe_LC 0.682±0.041
chisq 9.969
tail prob 0.533159
Full output

Considering the tight statistical fit, I think it's even possible that some of these people harbor direct ancestry from Maykop Novosvobodnaya. Here's a Principal Component Analysis (PCA) showing why my qpAdm model works so well. It was produced with the data in the text file here and the Vahaduo PCA tools here.

Moreover, one of the Arslantepe males, ART038, belongs to Y-haplogroup R1b-V1636 (R1b1a2). This is clearly a marker of paternal steppe ancestry, because it's been reported in two Eneolithic samples from the southernmost part of the Pontic-Caspian steppe near the North Caucasus foothills (see here). These individuals are dated to ~4,200 calBCE, so they lived about a thousand years earlier than ART038.

ART038 probably lacks steppe and Maykop-related ancestries on his autosomes. Nevertheless, my point about his Y-haplogroup stands, because autosomal admixture can be bred out and disappear completely within a couple hundred years, or about 6 to 8 generations.

Interestingly, Skourtanioti et al. argued against the possibility of significant steppe and Maykop-related ancestries in the Arslantepe LC samples. They also didn't see R1b-V1636 as an obvious signal of paternal steppe ancestry. I find this very puzzling indeed, because to me it seems way off the mark. From the paper:

However, R1b-V1636 and R1b-Z2103 lineages split long before (~17 kya) and therefore there is no direct evidence for an early incursion from the Pontic steppe during the main era of Arslantepe. Lineage L2-L595 found in ALA084 (Alalakh) has previously been reported in one individual from Chalcolithic Northern Iran (Narasimhan et al., 2019) and in three males from the Late Maykop phase in the North Caucasus (Wang et al., 2019). These three share ancestry from the common Anatolian/Iranian ancestry cline described here, which indicates a widespread distribution that also reached the southern margins of the steppe zone north of the Caucasus mountain range.

See also...

Perhaps a hint of things to come

An early Mitanni?

How relevant is Arslantepe to the PIE homeland debate?

Tuesday, June 2, 2020

Perhaps a hint of things to come

It's still a mystery how the Hittites and other Anatolian speakers ended up in the Near East. However, the leading theory is that their ancestors migrated from the steppes of Eastern Europe to western Anatolia via the Balkans sometime during the Copper Age.

Consider the qpAdm mixture models below, made possible thanks to some of the ancient samples published recently along with Skourtanioti et al. 2020. The key ancients are described in a text file available here.

AZE_Caucasus_lowlands_LN 0.471±0.094
RUS_Vonyuchka_En 0.148±0.040
TUR_Barcin_N 0.381±0.069
chisq 12.874
tail prob 0.116261
Full output

RUS_Vonyuchka_En 0.107±0.029
TUR_Buyukkaya_EC 0.893±0.029
chisq 12.107
tail prob 0.207331
Full output

I'd say it's quite clear now that TUR_Barcin_C harbors minor ancestry from the Pontic-Caspian (PC) steppe. The reason this isn't widely accepted yet is because demonstrating it convincingly hasn't been possible without a proximate Anatolian ancestry source for TUR_Barcin_C, precisely like TUR_Buyukkaya_EC.

Admittedly, though, the statistical fits in my models aren't all that great. I suspect the problem lies with RUS_Vonyuchka_En, which is likely to be a rather poor stand in for the people who brought steppe ancestry, and possibly early Anatolian speech, to western Anatolia.

So let's see what happens when I try a more proximate reference for the steppe ancestry in TUR_Barcin_C. How about Yamnaya_BGR, an individual of mixed Balkan and steppe origin from what is now Bulgaria?

AZE_Caucasus_lowlands_LN 0.518±0.075
TUR_Barcin_N 0.203±0.056
Yamnaya_BGR 0.279±0.067
chisq 10.602
tail prob 0.225269
Full output

TUR_Buyukkaya_EC 0.749±0.058
Yamnaya_BGR 0.251±0.058
chisq 9.687
tail prob 0.376414
Full output

That's a little better. Unfortunately, the problem now is that the models are anachronistic, because TUR_Barcin_C is about a thousand years older than Yamnaya_BGR. Clearly, we need more Copper Age samples from the western edge of the PC steppe, the eastern Balkans, and especially northwestern Anatolia.

The Principal Component Analysis (PCA) below effectively illustrates why my qpAdm models work. It was produced with Global25 data using the Vahaduo PCA tools freely available here. Note that TUR_Barcin_C is shifted away from the essentially perfect cline formed by AZE_Caucasus_lowlands_LN, TUR_Barcin_N and TUR_Buyukkaya_EC towards samples from ancient Eastern Europe, including Yamnaya_BGR.

See also...

Steppe invaders in the Bronze Age Balkans