search this blog

Sunday, September 10, 2017

Your ancient ancestry #1

This is the first of a series of guides to modeling your ancient ancestry with the Global 10/nMonte2 method.

I do already have a user guide for running Global 10 and Basal-rich K7 data with nMonte and 4Mix (see here). However, in this series I’m going to recommend specific models that produce results similar to those from my experiments with other methods, such as qpAdm, as well as from scientific literature. Hopefully, this will help users achieve more sensible and accurate outcomes, and avoid problems such as overfitting.

Let’s start with models for modern-day Europeans that focus on Yamnaya-related ancestry, which very likely represents a genetic signal of early Indo-European dispersals during the Early to Middle Bronze Age from the Pontic-Caspian steppe.

It’s now clear via a wide range of methods that about half of the genomes of modern-day Eastern and Northern Europeans, and up to about a quarter of the genomes of modern-day Southern Europeans, are derived from such Yamnaya-related sources. Any tests dealing with ancient European substructures that don’t, one way or another, reflect this robust inference must be considered inadequate.

So if my models are to be useful, then this is what they must show. And indeed they do. Here are a few examples focusing on modern-day and ancient England, in chronological order:

Yamnaya_Samara 49.75
Barcin_N 32.3
Hungary_HG 17.95

distance%=0.5318 / distance=0.005318

Yamnaya_Samara 45.65
Barcin_N 33.35
Hungary_HG 21

distance%=0.4668 / distance=0.004668

Yamnaya_Samara 44.95
Barcin_N 31.6
Hungary_HG 23.45

distance%=0.5409 / distance=0.005409

Yamnaya_Samara 44.55
Barcin_N 36.95
Hungary_HG 18.5

distance%=0.3699 / distance=0.003699

Yamnaya_Samara 45.2
Barcin_N 36.85
Hungary_HG 17.95

distance%=0.4875 / distance=0.004875

The full output is available in a zip folder HERE. I’m not claiming that these ancestry proportions are perfect, especially for Southern Europeans, who generally have very complex ancestry, but they do make a lot of sense.

One obvious problem with the Global 10 is that some of its dimensions, or PCs, exaggerate affinity between modern-day and Mesolithic Europeans. This is especially true for PC6. Hence, to try and mitigate this problem I decided to remove PC6 from the Global 10 datasheet used in my analysis.

To try these models on your own genome, remove PC6 from your Global 10 coordinates file, and use the data text files provided in the zip folder linked to above. It’s best to rely on the datasheets specifically designed for your ethnic group or region of Europe. But feel free to tweak my models. There’s no harm in experimenting if you’re cautious and sensible about it. Indeed, using Iberia_HG or Loschbour along with Hungary_HG appears to produce more accurate outcomes for many Western Europeans.

The important, but often neglected, point to keep in mind is that I designed the Global 10 to help replicate results from more reliable but technically less accessible methods, and not to challenge any generally accepted models.

In the near future, a wider choice of ancient samples should enable me to fine tune and improve the models. For instance, a slightly more eastern-shifted forager reference population than Hungary_HG, such as the yet to be published Lithuanian Narva samples (see here), will probably shift the results slightly for Northeast Europeans, perhaps by bringing down their Yamnaya-related ancestry proportions by a few per cent.

Moreover, adding a wide range of yet to be published Middle to Late Neolithic European samples, such as those from the Globular Amphora Culture (GAC), should prove an interesting exercise.

See also...

Global 10: A fresh look at global genetic diversity


pequerobles said...


if i mail you my 23andme, and ancestry dna raw data, would you be able to do this calculator for me?


Sofia Aurora said...

Dear David
It's the first time that i post in your blog and please forgive any misconceptions or informalities of my post.
The reason that i decided to post is because i have read in April 2017 a post of yours about David Reich's speech in the American Philosophical Society about aDNA and the proof we deduct from it in favour of the Steppe solution of the Proto-Aryans (i don't call it steppe hypothesis, because we have the solution to the Proto-Aryan origins and that is the Kurgan explanation)!
In that post you said that pretty soon a paper will follow the speech in Nature journal (or in Science if i don't remember well).
Did that paper come out?
If yes can you PLEASE publish the internet address of it?
If not do you know when about is going to be published?
Respectfully Yours

Steven said...


Could you run this calculator for me if I mail you my 23andMe raw data?

Davidski said...

For anyone who wants the Global 10 coordinates to run my models, see here.

Chad Rohlfsen said...

It would probably be better to use Bichon for the HG part. KO1 is 15% Barcin and 15% EHG.

Davidski said...

Yes, I remember that qpGraph model showing Hungary_HG as 15% Barcin, but I haven't been able to see that in any of my tests.

I used Hungary_HG to basically represent everything non-steppe and non-Barcin, but yeah, Bichon could work too.

Indeed, for many people in Western Europe, using Iberia_HG or Loschbour alongside Hungary_HG produces interesting outcomes, showing a mix of Western European HG and Hungary_HG.

Samuel Andrews said...

Everyone should download the zip folder. It has nMonte results for all West Eurasian populations.

My results.
"Barcin_Neolithic" 44.45
"Yamnaya_Samara" 35.45
"Hungary_HG" 18.4
"Yoruba" 1.05
"Karitiana" 0.65
"distance%=0.5059 / distance=0.005059"

Davidski said...

@Sofia Aurora

The paper that I was referring to actually turned out to be three papers/preprints. I blogged about them here (you'll find links to the papers in these blog posts).

The genomic history of Southeastern Europe (Mathieson et al. 2017 preprint)

The Bell Beaker Behemoth (Olalde et al. 2017 preprint)

Steppe admixture in Mycenaeans, lots of Caucasus admixture already in Minoans (Lazaridis et al. 2017)

You'll also find a summary of this year's main findings from ancient DNA at the link here.

The big paper on South Asia is yet to be published, hopefully soon, and don't expect any surprises when it does come.

Sofia Aurora said...

Dear David
Thank you for answering my post.
I agree with you about South Asia but i am eager to find out if there are going to be any new haplogroups both from the steppe populations and from the aboriginals of South Asia (perhaps any haplogroup clarifing the origin of the Veddoids?)

Matt said...

This looks like it will be very useful to some folk.

Heck of a lot of Hungary_HG ancestry. Like, assuming Yamnaya 60% EHG (for the sake of assumptions; it could be anywhere as low as 45% and possibly 50%), then Anglo Saxon 27% EHG, and they have almost as much Hungary_HG (23.45%). The Latvians could actually have slightly more Hungary_HG like ancestry.

Nothing like IrelandMN+Yamnaya can cope with these fractions (e.g. Bell Beaker here would need something like a 37:63 WHG:Barcin_N ancestor to mix with Yamnaya).

As another thing, can I ask about the Anatolia_BA fractions in the SE European models? Do you think that's actually a real phenomenon, or is it more like that's the best way of representing it at the moment?

On the whole topic of different HG ancestry in Europeans, I was thinking about this paper with extra Iberian and Romanian HG from earlier in the year - I sort of forgot about it with all the other stuff going on, but thinking back, was wondering if the HG genomes (the Iberian Canes and Chan in particular) could be any use in trying to isolate different Middle Neolithic and HG streams, compared to just using the variances in El Miron, ANE, etc. affinities.

(Semi on topic, somewhat more convinced today after the other thread in Olalde 2017's finding of a Central European MN only contribution to Bell Beaker Germany. Though would still like to see it replicated with the new qpAdm methodology, using many ancient samples, as they were still running with world focii plus a few relevant ancients in Olalde, which at the moment does not seem like clearly the best methodology).

Davidski said...

@Sofia Aurora

As far as I know, we shouldn't expect anything to contradict the present findings. In other words, expect more R1 in ancient steppe populations and in steppe-derived groups in South Asia. The earlier South Asian samples from Harappan sites will show a different set of haplogroups.

I also know that efforts were made to get Mesolithic South Asian samples, but I don't know how that turned out, so we might see Y-DNA from Mesolithic South Asia or we might not.

Davidski said...


Heck of a lot of Hungary_HG ancestry. Like, assuming Yamnaya 60% EHG (for the sake of assumptions; it could be anywhere as low as 45% and possibly 50%), then Anglo Saxon 27% EHG, and they have almost as much Hungary_HG (23.45%). The Latvians could actually have slightly more Hungary_HG like ancestry.

Nothing like IrelandMN+Yamnaya can cope with these fractions (e.g. Bell Beaker here would need something like a 37:63 WHG:Barcin_N ancestor to mix with Yamnaya).

The HG ratio in these models can be forced down by using Villabruna or Bichon. But in any case, I think it's quite certain now that Yamnaya picked up a lot of HG-rich Neolithic MN admixture from groups like GAC and TRB. Also, I'd say there's a good chance that western Yamnaya was more HG-rich than Yamnaya Samara.

As another thing, can I ask about the Anatolia_BA fractions in the SE European models? Do you think that's actually a real phenomenon, or is it more like that's the best way of representing it at the moment?

I think it's real, and indeed Anatolia_BA looks like a very proximate reference for it. Adding Anatolia_BA to the Balkan models produces very sound results in my view. South Italy and Sicily are more complex, and they probably need something from the Levant for realistic outcomes, but I don't know how to fit it all in.

Anthro Survey said...

In case anyone's interested, this is my best model for North Italy to date and they make good sense. Essentially, a mix of local EEFs, incomer Urnfield-like groups, and some CHG-rich admixture from the East Med. In my humble opinion, this East Med wave probably arrived relatively late and was even more Armenia-shifted than Anatolia Chl. This means one thing: more endogenous EEF ancestry in North Italy and a minimum of ~12% post-Neolithic ancestry from the Near East. That's because ANatolia_Chl=EEF+Armenia, more or less.

South Italy models as Bergamo+Levant_BA+more Anatolia_Chl. The post-Neolithic contributions there are copious.

[1] "distance%=0.0649 / distance=0.000649"


Halberstadt_LBA:I0099 46.40
Anatolia_ChL:I1584 25.80
LBK_EN:I0795 9.85
Portugal_LN:DolmenAnsiao96B 5.45
Portugal_LN:MonteCanelas337A 5.35
LBK_EN:I0054 5.10
LBK_EN:I0046 2.05

Anthro Survey said...

It's interesting that Tuscans don't "take" much of the Levant_BA, while that type of shift is important in the south relative to N. Italians.

Could this correspond to those elusive Syrians that migrated to S. Italy during Roman times and the genetic cline Roman rule introduced thereof as Davidski has suggested in the past?

[1] "distance%=0.1536 / distance=0.001536"


Italian_Bergamo 49.45
Anatolia_ChL:I1584 29.30
Levant_BA:I1705 9.90
Levant_BA:I1730 7.35
Levant_BA:I1706 4.00

[1] "distance%=0.2283 / distance=0.002283"


Italian_Bergamo 76.2
Anatolia_ChL:I1584 21.2
Levant_BA:I1705 2.6
Levant_BA:I1706 0.0
Levant_BA:I1730 0.0

Rob said...

@ Anthro

USe either the best quality sample, or an average, so you don't get multiple redundant individuals (eg 3x "Levant BAs")

I don;t get any Levant BA for Tuscans.

"Baalberge_MN" 39.85
"Armenia_EBA" 27.9
"Hungary_CA:I1497" 17.95
"Yamnaya_Samara" 13.5

(using 3000 BC source populations)

For a different south European:

"Baalberge_MN" 43.9
"Armenia_EBA" 24.2
"Yamnaya_Samara" 17.35
"Hungary_CA:I1497" 10.5
"Kumtepe_LN:kum4" 3.1

Seinundzeit said...


Those are excellent suggestions.

On a somewhat related note: I've been trying to explore the differences that exist between the ancient "steppe-derived" populations which seemingly lacked significant EEF/MN admixture.

For what it's worth, I've found that Srubnaya_outlier and early Latvian "Corded Ware" lie on opposite ends of an ANE-to-WHG continuum with regard to their non-CHG ancestry, while Yamnaya_Samara is somewhat intermediate in relation to those two populations.

The reference populations were Iran_N, CHG, Levant_N, ANE, and WHG. I applied scaling to the data (and just to clarify, by "scaling" I do not mean to imply the weighting of the PCs by the eigenvalues. That doesn't work).


59.8% ANE + 14.2% WHG

26.0% CHG



31.4% WHG + 24.2% ANE

44.5% CHG



44.8% WHG + 5.7% ANE

49.5% CHG


For comparison, this is how EHG fare:

59.9% ANE

39.7% WHG

0.4% CHG


It seems like there was some substantive genetic substructure among the hunter-gatherers of Northeastern/Eastern Europe + the Eurasian steppes.

Also, interesting that CHG is the unifying factor for these populations. Not sure if that has any bearing on the PIE question, but one is tempted to draw the connection.

Anyway, I've tended to focus on West and South Central Asians, when modelling contemporary West Eurasian populations.

But after observing the patterns above, I decided to give European populations a spin, just to see how they stack-up when one throws in the Srubnaya_outlier, EHG, Yamnaya_Samara, and early Corded Ware.

(To be continued)

Rob said...

"Also, interesting that CHG is the unifying factor for these populations"

Sorry to laugh, but did you just figure that out now ?

Seinundzeit said...

Continuing from where we left off...


87.55% Boncuklu_N

8.95% WHG

3.50% Levant_N



36.0% Levant_N

26.1% Latvia_LN1

19.0% WHG + 17.6% Boncuklu_N

1.2% Gambian



44.40% Boncuklu_N + 16.50% WHG

29.85% Latvia_LN1 + 9.25% Yamnaya_Samara



34.0% Boncuklu_N + 23.5% WHG

28.2% Latvia_LN1 + 14.2% Yamnaya_Samara



45.70% Latvia_LN1

28.80% WHG + 24.25% Boncuklu_N + 1.25% EHG



26.3% Latvia_LN1 + 18.3% EHG + 3.4% Srubnaya_outlier

29.9% Boncuklu_N + 17.2% WHG

4.8% Ulchi



40.60% EHG + 8.00% Latvia_LN1 + 2.15% Srubnaya_outlier

29.85% Boncuklu_N

19.40% Ulchi



50.5% EHG

20.8% Boncuklu_N + 8.4% WHG

20.3% Ulchi


So one sees some very interesting patterns, and those patterns seem to be quite sensible.

For comparison, West Eurasian populations at the southeastern periphery, using the exact same reference populations.


42.3% Iran_N + 10.9% Iran_Chl

36.7% Srubnaya_outlier

10.1% ASI



33.1% Iran_N + 20.9% Iran_Chl

32.5% Srubnaya_outlier

13.4% ASI


Karlani Pashtun, central highlands

41.60% Iran_Chl + 11.55% Iran_N

36.60% Srubnaya_outlier

9.45% ASI

0.80% Mongola


Sarbani Pashtun, southwestern plateau

39.15% Iran_Chl + 20.70% Iran_N

30.40% Srubnaya_outlier

6.25% ASI

3.50% Mongola



34.30% Srubnaya_outlier + 16.70% Yamnaya_Samara + 0.10% WHG

40.40% Iran_Chl

5.85% ASI

2.65% Mongola



51.00% Iran_Chl + 1.20% Levant_EBA

25.95% Srubnaya_outlier + 13.45% Yamnaya_Samara + 2.60% WHG + 1.55% Boncuklu_N

4.25% Mongola


Interesting stuff; my general impression is that the Indo-European expansion into southern Central Asia (Tajikistan/Uzbekistan/Turkmenistan and Afghanistan/northern Pakistan) might have involved a population that was very much ANE-rich (shifted very strongly towards ANE rather than WHG on the European HG continuum, like the Srubnaya_outlier).

By contrast, mainland Europe seems to have been subjected to an IE expansion involving a population very heavily skewed towards WHG, like the early Latvian Corded Ware sample.

Also, seems like populations in the Volga-Ural region are exceedingly interesting (although, things tend to get rather interesting even for some Slavic Russian populations).

Still, more samples needed; it'll all come together eventually.

On a personal level, I’m incredibly stoked about those Swat valley results.

Davidski said...

I'm pretty sure that CHG is not the unifying factor on the Bronze Age steppe, because the southern ancestry in Khvalynsk, Yamnaya and Sredny Stog, all groups identified as early or even Proto-Indo-European, harbor starkly different types of southern ancestry, which look like they arrived on the steppe from different sources and times. So the unifying factor is R1-rich EHG and mobility on the steppe.

Seinundzeit said...


"Sorry to laugh, but did you just figure that out now ?"

I wouldn't rush to conclusions. I said that one is "tempted" to draw a connection, but one mustn't act on every temptation.

Like I said, we need far better sampling; a resolution to the PIE question isn't immediately at hand.

That being said, LPIE is pretty much settled, no? Everything points towards the steppe.

Rob said...

@ Sein

I was referring to your genetic deductions, with which I agree
I wasn't making implications for PIE linguistically. For the latter, a protracted expansion of archaic PIE languages means that I shy away from precisely such kind of deductions that PIE came with CHG this or R1 - that

Anthro Survey said...


I find that fits are crisper but no less sensible when I use "redundant" samples for some ancestral streams like I did with Neolithics(who are to be interpreted as local agriculturalists prior to the influx of transalpine steppe-rich populations in my Bergamo model). We're working with a 10D PCA and sometimes the certainty about the finer Neolithic variation of an area just isn't there. Given how similar Neolithic Western Europeans were in their deep ancestry, my Neolithic "shotgun" approach is harmless in the long run.

Also, I tend to focus on more recent population dynamics. This is especially so when it comes to steppe ancestry. In particular, using Urnfield and Beaker genomes for steppe ancestry mediators works out great. The fits are excellent, which is unsurprising considering temporal and spatial proximity, and the proportion of Yamna-related ancestry can be calculated indirectly this way.

The Yamnaya you got for Tuscans is curiously low compared to the ~20% figure from previous formal stats, admixture readouts, and nMonte models that use Urnfield/Beaker-like inputs.

Rob said...

@ AnthroS

I would think modern central - north Italians have 5-8% more steppe admixture than the Bronze Age forebears, by way of the Celtic and historic Germanic invasions.

I also use more recent sources, as you, did, but by doing runs with different -aged source populations, it just shows you how nMonte results can "jump around" massively depending on what you throw in. and formal models have potential limitations, such as available outgroups, and how the modeller themselves attempts the run.

But here is one with 2000 BC sources included.

"Bell_Beaker_Germany" 56.15
"Mycenaean" 30.35
"Jordan_EBA:I1730" 6.5
"Iran_Chalcolithic:I1665" 3.85
"Hungary_CA:I1497" 1.65

"Anatolia_BA" 53.6
"Bell_Beaker_Germany" 36.1
"Jordan_EBA:I1730" 6.75
"Vatya:RISE479" 2.35

Rob said...

So i think nMonte can suggest the 'big picture', modelling:

"Barcin_Neolithic:I1099" 64.3
"Kotias:KK1" 17.2
"Karelia_HG" 16.95
"Villabruna:I9030" 1.55

So here can can estimate that Italian Tuscans in the 21 century, have 30 % 'steppe' (15 % CHG + 15% EHG). If 10% is more recent prehistoric (say), 25% is Copper - Bronze Age stuff.
The question is what came via the Alps and what, say, via the Balkans ?
Did it all come at once, eg with Beaker or Polada, or was it successive waves ?
I think that get's a bit complex for nMonte.

Anthro Survey said...


It's a valid point regarding Celtic migrations--indeed, I think Halstatt and LaTene waves were significant. It is the Germanics I'm not so sure about, unless we're talking about Veneto and Friuli. My guess is that the Celtic wave wasn't too far off from Urnfield-like groups---albeit maybe w/extra EEF. In all likelihood, the Bergamasque and others from Lombardy/Emilia probably have next to zero Germanic influence despite having been a political center of the Lombard kingdom. This is somewhat analogous to the situation in Anatolia where Turkic ancestry is insignificant around Konya(Seljuk capital) but skyrockets, counter-intuitively, in western locales like Aydin(there is an interesting reason for that but don't wanna delve into that now).

Now, regarding the recent model----I tried something similar a while back. Beakers+Minoans to model Bergamo. It was a superb fit, but, again, we should be careful not to take Minoans at face value. I treat Minoans(and Mycenaenans in your case) as a geometrical sum of EEF+("Armenia related"+Anatolia_N). It's a sum of two unknowns, ultimately, and impossible to determine the magnitude of post-Neolithic ancestry w/certainty. For Bergamo, I made the assumption that it came relatively late----at a time when Anatolia was quite "KA"-ized. Mathematically, it would imply ~13% foreign "east med" admixture and the other 13% would simply increase the local agriculturalist score.

Anthro Survey said...

In the big picture model, it could be that Karelia is somehow eating up the WHG. In fact, I'm almost sure that's the case because you're not left with any extra CHG.

Part of the problem is that even though Karelia makes a good fit on a 2D PCA, it's not the case in 10D, with more components of variance. Most likely, Karelia specifically was not the group that contributed to the formation of the Yamnaya(or relevant early steppics), but some other EHGs.

This aspect is most apparent when you're modeling North Africans. It's impossible to get anything under 3% there. Very poor with Neolithic Levant, but slightly better with Natufians for the basal-rich starters.

Shaikorth said...


Yamnaya improves the fit for Volga populations like Chuvash a lot compared to Latvia_LN, considering geography this was expected. With PC6 removed:


Karelia_HG 59.5
Ulchi 21.7
Barcin_N 18.9
Hungary_HG 0.0
SHG 0.0
Yamnaya_Samara 0.0


Karelia_HG 36.8
Barcin_N 23.8
Ulchi 21.4
Yamnaya_Samara 18.1
Hungary_HG 0.0
SHG 0.0

Anthro Survey said...


My India doodlings... Tamil Brahmins. The second model makes a lot of sense historically and can be conveniently taken at face value: Brahmin migrants from the north mixed to a limited degree with ASI/Iran agriculturalists to create the southern Brahmin caste. Fit improvement seen.

[1] "distance%=1.0116 / distance=0.010116"


Paniya 57.45
Iran_N:I1945 27.50
Yamnaya_Samara:I0441 15.00
Yamnaya_Samara:I0357 0.05

[1] "distance%=0.4236 / distance=0.004236"


Brahmin_UP 66.8
Paniya 20.7
Iran_N:I1945 12.5

David Rabaez said...

[1] "distance%=0.302 / distance=0.00302"

"Barcin_N" 60.95
"Hungary_HG" 22
"Yamnaya_Samara" 17.05

[1] "distance%=0.6489 / distance=0.006489"


Barcin_N 63.65
Hungary_HG 18.00
Yamnaya_Samara 16.60
Yoruba 1.75

[1] "distance%=0.6778 / distance=0.006778"


Barcin_N 67.15
Hungary_HG 25.90
Yamnaya_Samara 5.65
Yoruba 1.30

[1] "distance%=0.3586 / distance=0.003586"


Barcin_N 65.6
Hungary_HG 25.3
Yamnaya_Samara 7.4
Yoruba 1.7

Davidski said...

@David Rabaez

Try Iberia_HG instead of Hungary_HG. It seems that slightly different types of hunter-gatherer ancestry can shift the Yamnaya percentage in this test.

huijbregts said...

For most Europeans the Global 10 is a tale of two dimensions: PC4 and PC6. These carry the most variance and on visual inspection show the most structure.
So it seems plausible to relate these dimensions to the big migrations EEF and Steppe.
Indeed a plot of the Europeans on PC4 and PC6 shows very distinct clines, and captures most of the European structure:
(I dropped HGs and Caucasus).

Curiously these clines are not aligned along the principal dimensions but markedly rotated. Even after redoing the PCA, the clines are still rotated. Anybody got an idea why?

Matt said...

Anthro Survey: Most likely, Karelia specifically was not the group that contributed to the formation of the Yamnaya(or relevant early steppics), but some other EHGs.

From what I remember, the Samara and Karelia HG samples actually share relatively low drift under f3 stats with each other compared to WHG (esp. WHG from the Loschbour-Bichon-Rochedane subgroup. So population structure between different EHG groups (and Samara and Karelia are a huge distance apart), is plausible.

(As much as these northern populations of EHG and WHG were obviously relatively homogenous over distance compared to what we see in the Near East).

Vincent said...

@ Anthro Survey

"Could this correspond to those elusive Syrians that migrated to S. Italy during Roman times and the genetic cline Roman rule introduced thereof as Davidski has suggested in the past?"

No there's no Syrian admixture from Roman times in South Italy. The genetic cline comes from more CHG and less EHG in the south (see Sazzini, & Sarno). Sicilians cluster with Mycenaeans and Minoans in Lazaridis.

Davidski said...


You'll end up being very disappointed if you really believe that modern Sicilians cluster with Minoans and Mycenaeans.

They're sort of similar. But Sicilians have some North African and Levant admixture that is missing in Minoans and Mycenaeans.

Tesmos said...

Is this model outdated?

"Yamnaya_Samara:I0231" 43.2
"LBK_EN:I0025" 39.1
"Loschbour:Loschbour" 17.7

Davidski said...

Still fairly current. It'll have to change when we get samples from western Yamnaya.

Yamnayan Vegan said...

@Davidski which datasheets would be the best for the peoples of Caucasus and Anatolia?

Davidski said...

@Yamnayan Vegan

I haven't looked at West Asia yet. But the model for Cypriots should be useful for many West Asian groups, especially if you tweak it a little.

Vincent said...


North African/Levant admixture in Sicilians is no more than 5%, and it's even less in other South Italians (close to zero in most).

Sicilians are VERY similar to Mycenaeans and a bit more 'northern' than Minoans. They're shifted toward steppe/CHG, not Africa or the Levant.