search this blog

Monday, February 13, 2017

Mitogenome diversity in Sardinians

Open access at Mol Biol Evol:

Sardinians are “outliers” in the European genetic landscape and, according to paleogenomic nuclear data, the closest to early European Neolithic farmers. To learn more about their genetic ancestry, we analyzed 3,491 modern and 21 ancient mitogenomes from Sardinia. We observed that 78.4% of modern mitogenomes cluster into 89 haplogroups that most likely arose in situ. For each Sardinian-Specific Haplogroup (SSH), we also identified the upstream node in the phylogeny, from which non-Sardinian mitogenomes radiate. This provided minimum and maximum time estimates for the presence of each SSH on the island. In agreement with demographic evidence, almost all SSHs coalesce in the post-Nuragic, Nuragic and Neolithic-Copper Age periods. For some rare SSHs, however, we could not dismiss the possibility that they might have been on the island prior to the Neolithic, a scenario that would be in agreement with archeological evidence of a Mesolithic occupation of Sardinia.

Olivieri et al., Mitogenome Diversity in Sardinians: a Genetic Window onto an Island's Past, Mol Biol Evol, Published: 08 February 2017, DOI:

See also...

Something unexpected from Mesolithic Sardinia


Samuel Andrews said...

As interesting Sardinians are it'd be better if energy and money was spent getting 3,400 mito genomes from non-bottle necked, tiny populations.

Samuel Andrews said...

Stupid reasoning which has lead to wrong conclusions about European genetics for years...
"However, it is also important to realise that
even if H3 (and H1) arrived in Sardinia only with the Neolithic, they most likely came from
either Spain or elsewhere in the western Mediterranean, and not from the Near East. This
would imply that they are likely the result of autochthonous west Mediterranean Mesolithic
acculturation, in the wider European context. "

80-90% of European mtDNA is from Late Mesolithic/Neolithic Western Asia, get used to it.

Davidski said...

Yeah, not much chance that H1 and H3 are European.

Nirjhar007 said...

Dave, Sam

What are the most frequent European Mtdna groups , apart from U5 ?.

Unknown said...

Hello again Davidski. It's little off-topic, but now that some prehistoric east asian genomes are finally sequenced, namely DevilGate and Sanganji/Shikkariabe Jomon. Perhaps could you try to create a new admixture calculator with them? Would be nice if we finally had a calculator with at least a few ancestral east asian components on Gedmatch.

xibler said...


"Yeah, not much chance that H1 and H3 are European."

So where did they come from, then? Just wondering what your theory was.

Unknown said...

I think H came from a Southeastern European refugium it was dominant in Romania at least 7500 years ago and seems (not enough data) to be dominant before that. Of course H has been found and confirmed (RFLPS) in North Africa 10-20kya (not sure of exact date not enough time to research) along with U5 and U2 so this was from a European population and probably did not originate in the near east (although some of there mtdna did) and even though it hasnt been confirmed it has been suggested in Mesolithic Iberia and HV* has basically been confirmered in South Italy around 30kya. It seems to have origininated in Southeast Europe area maybe Italy but more likely further east and by 20kya had spread to North Africa and maybe Iberia and possible coastal areas of France ect. In modern Europe most H seems to have a recent origin from Southeast Europe (maybe Hamangia pre Hamangia) which is probably also assoicated with R1b possibly while R1a may not be directly associated with H until very recent times (i.e 3-5kya). It seems all H in Europe has a recent SE European origin except H1 and maybe others like H6 and H7 which may have a south western origin and may be associated with I2 and H2 and H3? which may come from the neolithic near east perhaps a back migration? The oldest anatolians from 11kya both had N1a mtdna and G2a y-dna Ancient dna suggests it was rare in the near east where it starts showing up around 8,700 years ago as a minority and no ancient dna evidence shows it was there before 9,000kya of course maybe certain groups in the near east had totally different mtdna than other groups and it does come from one of these groups the strongest evidence that currently exist for a Near Eastern origin is Kumtepe6 in my opinion.

xibler said...

@John Smith,

Thumbs up!

This H1a3 thanks you for that.

Gaspar said...


H1 was already in Karlsdorf Germany in Early Neolithic times, did it go to Spain from Germany?

Gioiello said...

@ Samuel Andrews

I have to read the paper, but, after that also on Anthrogenica more and nore people is giving reason to me and my theory of an Italian Refugium, the defeat for you and for all Levantines is complete, similar to many defeats that Romans gave to what you think to be and you aren't...

Samuel Andrews said...

Everyone should save this data somewhere.

Haplogroup Frequencies in modern and ancient Europe.

H subclade frequencies in modern and ancient Europe.

Hunter Gatherers

Hungary Neolithic.

Germany Neolithic.

Iberia Neolithic.

Romania Neolithic.

Steppe Chalcolithic.

North Europe Bronze age/Late Neolithic.

Crete Bronze age.

Samuel Andrews said...

Here's why I think H1 and H3 are of Near Eastern origin.

H1 and H3 pop up throughout Neolithic Europe. Today H1 is very popular in Europe and NorthWest AFrica. Today H3 is very popular in Western Europe and Northwest Africa. NorthWest Africans have a tiny fraction of Paleo European ancestry, so how could H1 and H3 be of Paleo European origin?

No mtDNAs from Paleo Western Europe belong to H1 or H3 but plenty of mtDNAs from Neolithic Europe do. Neolithic Europeans traced most of their ancestry(70-100%) to Neolithic Anatolia/Near East.

Samuel Andrews said...



xibler said...

@Samuel Andrews

What made H1 and H3 so popular -- coming from the near east and all, who were their y cohorts? Did they fare as well? Just curious.

Unknown said...


We dont know all the paleoeuropean populations yet I think. Too many holes.

I favour the North African origin for H1 and H3. The facts fit (paleolithic presence, pattern of expansion from Iberia, absence from the Near East etc) and the first Bell Beakers are associated with North Africa (archaeologically/culturally).

Matt said...

Haven't read through, but wonder how many "Sardinian Specific Haplogroups" will remain so as more populations are sequenced at similar depth? Or is high population coverage sampling extant enough that this isn't an issue?

Reading physorg's coverage also wonder more about the possibility of connections between Southern European hunter gatherers and the Near East ("The most plausible candidates would include haplogroups K1a2d and U5b1i1, which together comprise almost 3 percent of modern Sardinians, and possibly others. Such a scenario would not only support archaeological evidence of a Mesolithic occupation of Sardinia, but could also suggest a dual ancestral origin of its first inhabitants. K1a2d is of Late Paleolithic Near Eastern ancestry, whereas U5b1i1 harbours deep ancestral roots in Paleolithic Western Europe").

IIRC early work on La Brana suggested Iberian HG had no more Basal Eurasian than Loschbour, throwing doubt on any connection between S Europe and Near East prior to Neolithic.

But how we calculate Basal Eurasian seems now subject to more question and Lazaridis's best fit estimates from 2016 even models SHG with some Basal Eurasian (10%) and Swiss HG (2.5%?) -

Won't change the basic finding of IBS / f3 relatedness to S European HG to be found in NE Europe, but I wonder if there won't be some connections found between S Europe pre-Neolithic and Near East.

(Small Experiment: Take the numbers from the above Lazaridis graph, fit Anatolia_N as 25% BE at 0% WHG, Early_EN at 24% BE at 8% WHG, EuropeMNChl at 20% BE at 23% WHG, regression to 100% WHG would be expeted to be 3% BE. (Or if you take EN as 5% WHG and MN as 20% WHG is implied to be 0.25% BE).

While taking CHG at 35% BE as 0% EHG, Yamnaya as 20% BE as 66% EHG, Samara_Eneolithic as 17% BE and 16% CHG, regression to 100% EHG should be 13% BE.

These numbers aren't necessarily correct, but they are the latest from the lit, and do suggest that if we have a problem working out Basal Eurasian %s, this has a knock on of presenting questions for whether South->North geneflow before the Neolithic may have been more extensive than first thought.)

Kristiina said...

Samuel, where is your Mesolithic mtDNA evidence from the Eastern Mediterranean and Bulgaria/Romania? How do you know that H arrived to Bulgaria from Anatolia during the Neolithic and not any earlier for example from Anatolia or even from Caucasus?

batman said...

@ Samuel

Congratulations on some great work. Though, your last statement was a bit confusing.

"Neolithic Europeans traced most of their ancestry(70-100%) to Neolithic Anatolia/Near East."

Is there clear-cut proves of a downstream transport of y-dna and mt-dna, respectively, that proves such a migration?

Or is your statement based on mere opinions, borrowed from some old archaeological textbooks - where this opinion used to be presented as "a widely held view", rather than a "unsubstantiated hypothesis"?

AFAIK we can't separate y-dna R1a/b or mt-dna HV (and possible some U5?) from the spread of lactose-persistance, milk-drinking and diary-industry throughout Eurasia, from Iberia to Siberia.

The older spread of herders with smaller cattles - reindeer, goats, sheep and pigs - can be found among y-dna G, H, I, J and N alike, who seem to have occupied the different climate-zones of Mesolithic Eurasia.

Since the remians of 7.300 yeras old chickpeas and barley was found in the proximity of the 7.500 years old R1a from Oleni Ostrov (Carelia) it's hard to argue that the introduction of the agricultural communities in NW Eurasia was NOT carried (mainly) by R1a and R1b.

Contemporary to the 7.500 yrs old Careila R1a and Samara R1b we find the spred of the oldest pottery i Europe - from the Sperrings ceramics, to which the clay-ladden waterways of the Finnish Gulf and the asbestos-mines of the Ladoga was central. Around this period, as the Holocene climate reached its 4.500 year long optimum, the Maglemosian y-dna I started developing the Ertebolle ceramics - from which the TBK, the BB, the CWC, the Yamna and the Maikop seem to have developed.

Moreover - the spread of the early branches of R1a seem to point towards the spread of cattle-farming on the open fields of southern Sweden and Norway, from where it found the Orkneys and Scotland - to complement the older isle- and woodlanders who kept goats and pigs in their barns and y-dna I in their genes...

mt-dna H seems to be dominant among the cattle-farmers, which indicates that this hg had (and still has?) an edge in the adaptibility ('fitness') that came from the "evolutionary pressure" that occured as the Holocene climate created lush grassglands from the old tundra and taiga of Eurasia, openeing the 'biotope' where horse-breeding and cattle-farming would be preferable - if growing larger communities, cultures and 'civilizations' was to be developed - along all major archipelagos and water-ways of the Eurasian continent.

jv said...

@John Smith, where do you think H6 originated from? I'm very interested in H6 and H6a. Seems she was in Pontic- Caspian Steppe Cultures, but where before that???

jeanlohizun said...

@Krefter/Samuel Andrews

Enough with the bullshit! This very paper shows that H3 is nonexistant in Northwest Africa.


In addition, we have found mt-DNA Haplogroup H in:

La Chora(H6), La Pasiega(H), Linantzeta(Hx(H1,H3)), and Vestonice42(H). Why the hell are those samples not in your databases? Do you have any proof that they are not legit, because otherwise, you are parading selective data, not the actual data.

jeanlohizun said...

Correction for Table-S9, since last picture was missing the bottom part.

Samuel Andrews said...


Try this. H3 is abundant in NorthWest Africa. If you want me to give you my sources I will..

"Samuel, where is your Mesolithic mtDNA evidence from the Eastern Mediterranean and Bulgaria/Romania? How do you know that H arrived to Bulgaria from Anatolia during the Neolithic and not any earlier for example from Anatolia or even from Caucasus?"

What matters is H migrated from somewhere far away in the east into 90% of Europe. Whether that somewhere far away in the east is in Asia or Europe or the Caucasus doesn't matter. At this point we have no where to know exactly where this somewhere far away in the east is.

jeanlohizun said...

@ Samuel:

Your sample size for Africa is 641, and it has a grand total of 16 samples with mt-DNA H3. That gives a total of 2.5% mt-DNA H3. Notice that if one removes the two samples from Tunisia which account for a total of 97 samples, yet contribute 7/16 H3 lineages, then the rest of Africa will have 9/544 or 1.65% Mt-DNA H3. This very study, featured per Table-S9 a compilation of 979 samples, from all over Northern Africa, with the exception of Tunisia, and failed to yield any mt-DNA H3, including in samples of Algeria(n=82), Libya(129), Egypt(551), and Morocco(217).

So is it any wonder that mt-DNA H3 is found in Tunisia, of all places, you do know the history of Tunisia right?

So no, you have proven anything. Mt-DNA H3 is ubiquitously European.

Samuel Andrews said...


I said...
"H3 is abundant in NorthWest Africa."

I didn't say H3 is abundant in North Africa or Africa as a whole. Here are NorthWest Africa H3 frequencies I have gatherered from studies.

Algeria(Multi-Ethnic): 3.65%
Tunisia(Multi-Ethnic?): 10%
Tunisia(Berber): 5.9%
Morocco(Berber): 1.7%

Samuel Andrews said...

IMO, it's *possible* H3 is of Mesolithic European origin. I don't deny the possibility. H1 though clearly is not of Mesolithic European origin. Also JT, K, X, W, N1 which makeup usually 30% of European populations mtDNA are undoubtly of Mesolithic Western Asian origin.

Some of you have to learn to understand that...

>Massive population replacement occurred before the colonization of America.

>Location with highest Haplogroup Diversity=/=Place of Haplogroup origin. Sometimes when migration occurs entire populations move and they bring with them all of their homeland's haplogroup diversity.

>A population can have ancestry from a region in which the modern people living there aren't a good proxy for the people who lived there when their ancestors did.

You guys need to understand those things because...

Modern European mtDNA is diverse enough to be the place of origin for most European mtDNA because Neolithic Western Asians who migrated to Europe brought all of their homeland's mtDNA diversity.

Modern Western Asians' mtDNA makeup doesn't support the idea most European mtDNA is from Mesolithic Western Asia because Western Asia mtDNA has changed since the Mesolithic. Modern Western Asians aren't a very good proxy for the Western Asian ancestors of Europeans.

jeanlohizun said...

@ Krefter.

And you my friend need to learn math. Here is what you said:

Try this. H3 is abundant in NorthWest Africa. If you want me to give you my sources I will..

You posted this again as some sort of source that proves the abudance of mt-DNA H3 in Northwest Africa in spite of my comment above.

Algeria(Multi-Ethnic)(n=192): 3.65%
Tunisia(Multi-Ethnic?)(n=29): 10%
Tunisia(Berber)(n=68): 5.9%
Morocco(Berber)(n=121): 1.7%

So, that means that there are 7/192 mt-DNA H3 lineages in Algeria, 3/29 i the Multi-Ethnic Tunisian sample, and 4/68 in the Berber Tunisian sample. Whereas Morocco Berbers had 2/121 mt-DNA H3 lineages. Anyone can clearly see that Tunisia is an outlier here. I mentioned above that the study in this thread featured 979 samples per Table-S9 from all over Northern Africa with the exception of Tunisia, yet they failed to yield any mt-DNA H3 lineages. Therefore, you statement that mt-DNA H3 is abundant in Northwest Africa, does not follow through. In addition, the Berbers per Table-S9 had a total sample size of 201, and yet no mt-DNA H3.

You claim that no mt-DNA H1 has been found in Mesolithic European genomes, but you went silent about the presence of H in La Chora(H6), La Pasiega(H), Vestonice42(H). So Europe did have the precursors. Moreover, you know where H1 has not been found either? In Early Neolithic genomes from Anatolia, Levant, Iran, etc. We have found mt-DNA H, H13, H32 and H5, which a lot of them coincidentally have a modern day distribution that aligns with those results.

Samuel Andrews said...


"And you my friend need to learn math."

I didn't measure my description("abundant") of H3 frequencies in NW AFrica with numbers. So no I don't need to learn math. The H3 frequencies in NW Africa I presented are similar to H3 frequencies in most of Europe. That's why I described H3 frequencies in NW Africa as abundant.

Samuel Andrews said...

"You claim that no mt-DNA H1 has been found in Mesolithic European genomes, but you went silent about the presence of H in La Chora(H6), La Pasiega(H), Vestonice42(H)."

I don't include those samples in my spreadsheet, or do include them but with warning to viewers, because the studies they come from are unreliable. For example there's a little known study from 2000 which sampled Villabruna(he had U5b)'s mtDNA. Guess what result they got? H! But now we know he had U5b. Also the Red Lady from El Miron was said to have H by a 2015 study and then in 2016 another study found she had U5b

Unreliable ancient mtDNA studies' results are mostly H and U-crs(something which unhread of), very little U5 or K or JT. Crazy.

Ask an author of Lazaridis 2016 if Vestonice 16 truly had H. I'm 99% sure he didn't. Another 17 mtDNA genomes from Paleo Europe don't have H anyways so one H wouldn't be a game changer anyways.

I'll examine studies which claim to have found H in pre-Neolithic France and Spain sometime in the future. I keep those results in the back of my mind but I doubt them for good reason.

jeanlohizun said...

@ Sam

Look at Table-S9, the frequency of mt-DNA H3 in Southwestern Europe is >5%. The only sample with 10% here is from Tunisia, and with a total sample size of 29. That's insanely small. Notice that Table-S9 featured 979 North Africans (217 Moroccans, 82 Algerians, and 129 Libyans), and 0 belonged to Mt-DNA H3. Therefore, the frequencies you present are inflated due to low sample size in the Tunisian sample, and its outlier status. Take the frequencies between these two countries:

Algeria(Multi-Ethnic)(n=192): 3.65% 7
Morocco(Berber)(n=121): 1.7% 2

That is 9/313 or 2.88% mt-DNA H3 without Tunisia. 16/410 or 3.9% with Tunisia included.

So, you posting percentage without taking into account sample size speaks to either a) An ignorance of the meaning of percentage or b) A biased agenda to inflated the amount of mt-DNA H3 in North Africa. I hope is a), so my recommendation: Go back, and take some Math classes, the liberal arts are doing you no good.

jeanlohizun said...

El Miron came from two different samples. The one from was done from a tooth, the one from was done from a toe bone. I mentioned this before:!!!!&p=191245&viewfull=1#post191245

Vestonice42, not Vestonice16, and yes, she did have it, here you go:

In addition, neither La Chora, nor La Pasiega were from 2005, they were sequenced in 2012.

Nirjhar007 said...

Thanks Sammy,Will it be possible for you, to have separate posts, focusing on each main Mtdna groups of Europe , where you give your view , on their origins and history?.

Ryan said...

I second JV's point. Lot's of talk about where H didn't go but not as much on where it did. Could people maybe outline what they think did happen? (And as an H6a person myself wouldn't mind on that point either lol).

Gioiello said...

@ jeanlohizun
"We have found mt-DNA H, H13, H32 and H5, which a lot of them coincidentally have a modern day distribution that aligns with those results".

One sample of H32 has been found in the Sardinian samples, and I'll study it next, but H32 was found in Natufian and Iranian aDNA. Sam "halleluyed", but I demonstrated him that all the precursors and the sister clades were in Europe, some only in Tuscany, thus that was a proof that a migration after the Younger Dryas to Middle East happened. I had no answer from Sam. The paper clearly said that there have been Refugia in Iberia and Italy from where these haplogroups may have come. When I began to write 10 years ago, nobody spoke of a Refugium in Italy, and it is clearly demonstarted now that R1b (and I think also R1a) were in Italy and not in the Franco-Cantabrian Refugium, whose people went extint (at least from the male side) after the Younger Dryas.

Unknown said...

Samuel Adams To answer the question at 11:45am in ancient remains in Northern Africa U5 and U2 were found thus any claim that H in these remains must have had a near eastern origin is invalid (or highly unlikely) of course these remains did not really have H1 or H3 if I remember correctly but they were verified H so H1 and H3 (possibly H3 in historical times and maybe H1 due to unfortunate events) may have entered latter. Ancient dna does not suggest H originated in Near East if it does please advise me of this data. I dont care where it originated I just want to know where it orignated and ancient dna does not present a strong case really for near east or Europe so its all just speculation now even though from what I can tell adna seems to suggest Europe. Its important to note at ancient times not all Europeans were the same some may have been dominated by H in the mesolithic while others definitely had none (which is a fact). The reason I think it originiated in Europe is because 63.16% of R omanians had it 7500 years ago while it was not common in the near east at least at that time in the remains AFAIK. Plus the verified paglicci HV* which keeps getting ignored. Of course the above poster has knowledge of H in pre 10kya near east? so maybe that could be used as evidence of near eastern orgin if it is true (dont have time to research)

Davidski said...

@John Smith

You have to discriminate between old school ancient DNA results produced with PCR and new generation results produced with shotgun sequencing and SNP capture techniques.

PCR is inherently unreliable, even when done in association with very stringent ancient DNA protocols, like no handling of the ancient remains without full body cover.

Now here's the problem: almost all of the mtDNA H results from pre-Neolithic Europe come from studies that used PCR.

Gioiello said...

Something about how these scholars calculate the age of the haplogroups. JN573371 is Barbara Ann Lewis, and adopted American I found her parents to (also Ian Logan collaborated) through 23andMe autosomal test. She interested me just for her hg. H with 12 extra mutations, the highest found so far. She resulted full blooded Italian, and her mother came from Central Italy. She is put now in the hg. H105 for the mutation G9921A. JX153454 confirms the Italian origin of this subclade, but in the same time that Barbara's ancestresses had 12 mutations, this other survived line had much less.
H105 G9921A
1. JN573371 FTDNA H105 23-AUG-2011
A263G 309.1C 309.2C 315.1C 523.1C 523.2A 523.3C 523.4A A750G A1438G T4002C C4029T T4232C A4769G T4907C T6975C A8860G G9921A C12378T A14133G A15326G A16041G T16189C C16287T G16526A
2. JX153454(Italy) Raule H105 10-JAN-2014
A263G 315.1C C522- A523- A750G A1438G A4769G A8860G G9921A A15326G T16491C T16519C

Onur Dincer said...

@John Smith

Its important to note at ancient times not all Europeans were the same some may have been dominated by H in the mesolithic while others definitely had none (which is a fact). The reason I think it originiated in Europe is because 63.16% of R omanians had it 7500 years ago while it was not common in the near east at least at that time in the remains AFAIK. Plus the verified paglicci HV* which keeps getting ignored. Of course the above poster has knowledge of H in pre 10kya near east? so maybe that could be used as evidence of near eastern orgin if it is true (dont have time to research)

7500 years ago Romanians were already Neolithic Near Eastern farmer immigrants. mtDNA H and HV were already widespread in the Near East before 7500 years ago as can be seen from the Near Eastern Neolithic and pre-Neolithic ancient mtDNA results we have. Interestingly, of the three pre-Neolithic Near Eastern ancient mtDNA results we have (two from Georgia and one from Iran), one of the Georgian ones is H (H13c) and the Iranian one is HV (HV2).

Gioiello said...

@ Onur Dincer

You have a wide meaning of Near East. There there are at least three or four different regions: Caucasus, with Caucasian languages; Anatolia (at least divised in two regions: a Northern Western one linked to South Europe and to WHG) and a Southern one; the region with Afro Asiatic /Semitic languages (where we found old Natufians) and at least Iran/Zagros. I think you cannot ignore that, and it is weird that you (whose Y came very likely recently from Europe) seems to believe that you are linked with this old reality. Perhaps in some part of your autosome, because your language (Turk) is recently arrived and the Turk autosome only a little part of Anatolians to-day (not more that 5%, except some paricular places).
Thus we shoudl be able to speak of these different regions, of these different origins and of the period we are considered.

Alberto said...


I'm quite surprised that you know that people from modern Turkey have a small amount of Turkic ancestry, and that you can even give a number that is in the ballpark: 5%.

This means that you're actually not completely ignorant of autosomal data.

Could you give an estimate of the amount of WHG in modern Italians too?

Gioiello said...

@ Alberto

I'm quite surprised that you know that people from modern Turkey have a small amount of Turkic ancestry, and that you can even give a number that is in the ballpark: 5%.

This means that you're actually not completely ignorant of autosomal data.

Could you give an estimate of the amount of WHG in modern Italians too?"

Alberto, as I posted in another thread, I lasted three years (with the help also of Ian Logan, who lost his bet with me) for finding the parents of Barbara Ann Lewis, an adopted American, just for her weird hg. H with 12 extra mutations. Autosome has a meaning in short times, after it becomes "ideology", as the three Jewish authors of DNA.Land, as Poznik now at 23andMe etc. They have an agenda. My agenda is only the truth. I recomended to Davidski to use only SNPs very slow mutating, what Shi Huang and colleagues did in their papers, ridiculizing the "out of Africa".
About how much WHG is in me, my friend Dartraighe says very much, but I ask me how may we calculate what remains from 14000 years ago if after the closest relatives we have that everything becomes dubious.

Ariel said...


On the issue with basal %, I fear that a poor understaing of the eurasian part of Natufians and Neolithic Anatolians might come in to play there. Reading the Lazaridis 2016 paper you have the impression that natufians have no connection with the older Vestonice/Goyet cluster, but they are closer to WHG, too me that make absolutely no sense, since WHG is related to Vestonice/Goyet (or has some admixture from that group). Unless you are willing to say that there are two side of WHG, one closer to the eurasian in natufians, and one is just the older european side.

Onur Dincer said...


Don't worry. I use the terms "Near East" and "Near Eastern" only in a geographical sense. The Near East has no genetic unity, neither in the present nor in the past. We can talk of a kind of genetic unity when it comes to Europe, but for the Near East hardly any.

idurar said...

Disappointing paper and conclusion. Mtdna H3 is often the second most common mtdna lineage after H1 in Northwest African populations (ignored in the paper), the frequencies in privately tested Maghrebis are in line with peer-reviewed papers at around 1/4 of the total belonging to H1 or H3, particularly high in Moroccans (sample size is small at less than 50 though, but still interesting). Thus H3 looks as abundant there as in Western Europe as a whole: Samuel's statement is accurate, I don't see what the problem is. It is true that only some NW Africans populations reach frequencies of 10% or over of mtdna H3 (like in some SW Europeans) but keep in mind NW Africa is still understudied compared to Western Europe.
This, together with the very stricking domination of the H1+H3 pair in both modern Western Europeans and NW Africans and the absence in mesolithic Europeans makes the conclusion about possible mesolithic European continuity plain wrong. For now, it looks more like recently successful neolithic West Asian lineages that may or may not have been present in pre-neolithic North Africa and in the nearby fringes of [Mediterranean] Europe. Since Northwest Africans are autosomally predominantly descending from Ancient Near Easterners and with only a small pre-neolithic European contribution, the most likely mtdna lineage to fit mesolithic European (namely mesolithic Iberia) continuity in NW Africa is the less frequent but as widespread geographically as H1&H3: mtdna U5b.

truth said...

When a Autosomal study on Corsicans ? I've never seen one. We already know pretty much everything about Sardinians.

Unknown said...

I don't think Corsicans are particularly interesting. They're similar to other neolithic islander farmer populations of near eastern origin, although they might have more CHG and/or North African/Natufian admixtures than Sardinians, similar to Cypriots.

truth said...

@ Uknown

So, if you do know about Corsicans, where do they plot on a PCA ? my guess is between Sardinians and North-Central italians, depending on what regions of the island (being the south more similar to Sardinians), but we still don't know. Also haplogroup studies on them are scarce, but the ones i've seen they have 20% of I2 (the one which is upt to 40% in Sardinians)

Matt said...

@ Ariel, my take away from the Fu 2016 paper was that the Near East (presumably Natufians inc.) were closer to WHG (Villabruna cluster) than the Vestonice and Goyet clusters, but would still be closer to Vestonice and Goyet clusters than say, MA-1. But I haven't read it for a while. Is this different than what you are talking about?

Unknown said...

@Ryan and JV

I too am very interested in H6 but I think it probably had a different path from H1 and H3. Not a lot of evidence about but at the moment it looks like:

(1) Might be older in Europe than the main H1/H3 cohort (La Chora and a potential sample from paleolithic Morocco of H6a1a8).
(2) H6 is a much older, divergent haplogroup than H1/H3 which seems to have expended off the trunk much later (sequence diversity and dating).
(3) Modern distributions strongly favour expansion by a northern (steppe) route from central asia into western Europe with a side prong heading south down into the Sinai.
(4) Looks like there is a Sinai/Mesopotamia group, a northern European/Chuvash group and a Western European group (that may be split in two). The Western European group has expanded back throughout Europe. This is my own observation from modern and ancient data.

Davidski said...

(3) Modern distributions strongly favour expansion by a northern (steppe) route from central asia into western Europe with a side prong heading south down into the Sinai.

H6 is from the Near East originally. It made its way to Europe during the Neolithic, and then from Europe into Central Asia during the Bronze Age.

Unknown said...

I can't tell where Corsicans would be plotting, as I've never had seen any Corsican PCA, but they might be plotting closer to Sicilians.
The difference being that Sicilians have recent North African and Sub-Saharan ancestry whereas I think Corsicans may have less WHG and more basal eurasian than Sardinians as Corsica has been occupied by near eastern-rich populations such as Carthaginians, Etruscans, Acient Greeks and Romans but they shouldn't have much or any recent medieval or modern(from French migrant North Africans/SSAs) MENA ancestry like Sicilians have from invasions. But all of it depends on how much are modern Corsicans mixed with the mainland French.

Regarding WHG ancestry in Italians, I think it's 20-40% depending on a region with around 30% on average but Sicilians only have around 20% just like Algerians, Tunisians and Moroccans.

By the way, so what do you think about the feasibility of that Jomon and Devil Gate calculator Davidski, maybe you could make another fundraiser like the one for Basal-Rich K7? Unless of course those genomes are too low coverage for that?

Davidski said...

It's not feasible. Those genomes are too limited.

We won't know what the story really is in Eurasia until we see some good coverage Paleolithic genomes from Central and East Asia. Maybe then I can come up with a new calc. Although the present one works fine for showing the population history of West Eurasia, which is my focus.

Unknown said...

I think Paleolithic and Mesolithic genomes from Africa are very important too, as african skulls from back then are nothing like the modern ones whereas Vestonice skulls are similar to modern European skulls despite substantial MENA admixtures in modern Europeans which suggests quite a population shift in Africa. Bantus might have wiped many divergent populations during their expansion.

>Modern Western Asians aren't a very good proxy for the Western Asian ancestors of Europeans.

Actually some less admixed Bedouins kinda are. The reason why your typical urban Arab is a bad example is due to significant Iranian, Caucasian and South Asian admixtures in modern Arabs in addition to the SSA. Many modern urban Arabs are more CHG than ENF(Bedouin/Natufian cluster) while others who are more ENF are heavily SSA admixed(Yemenis, North Africans, some Bedouins).

EEF seems to be basically Bedouin without SSA or at least much less than modern Bedouins+WHG along with the regional drift and isolation.
Modern Sardinians actually share more drift with Natufians than Saudis and Bedouins because SSA, and Iranian/South Asian(Pakistani, Indian, Boloch etc) admixtures in modern Gulf Arabs broke their genetic continuity with Natufians.

Ariel said...


EEF have way more WHG than modern bedouins...

Ariel said...


Corsicans plot with tuscans and have a tuscan dialect and tuscan surnames, go figure.

Ariel said...


There is no tables in that paper (or in the lazaridis one for that matter) that will let you think that Natufians are closer to Vestonice than ANE.

Gioiello said...

@ Ariel

Why take you care of someone who is "senza nome né parte"?

Gioiello said...

@ Davidski
"H6 is from the Near East originally. It made its way to Europe during the Neolithic, and then from Europe into Central Asia during the Bronze Age".

From how long are you an expert also of haplotypes beyond the autosome?

Spain La Chora (Cantabria) [CH-1] 5346 ± 90 BC Date from Strauss 2002 ? 73G, 16093, 16362, reported as H6 Hervella 2012
Israel Wadi el Makkukh [WM 30] M 4240-4065 BC H6 16311C, 16362C, 16482G Salamon 2010
Yamnaya Russia Kutuluk I, Kutuluk River, Samara [I0444 / SVP 58] M 3300-2700 BC 551,461 R1b1a2a2 (Z2103) CTS1078/Z2103+, L150+, M415+ H6a1b
Haak 2015; Mathieson 2015; Sergey Malychev
Corded Ware Germany Esperstedt [ESP 15]
2465-2395 BC

H6a1a G73A C146T C152T C195T T239C A247G A769G A825t A1018G G2706A A2758G C2885T T3594C G3915A G4104A T4312C A4727G T7028C G7146A T7256C A7521G T8468C T8655C G8701A G9380A C9540T G10398A T10664C A10688G C10810T C10873T C10915T T11253C A11719G A11914G T12705C G13105A G13276A T13506C T13650C T14766C A16129G T16187C C16189T T16223C G16230A T16278C C16311T T16362C A16482G Brotherton 2013; Brandt 2013
Poltavka Russia Kutuluk III, Kutuluk River, Samara [I0126 / SVP 51] M 2867-2484 BC 339,812 R1b1a2a2 Z2105 H6a2 Mathieson 2015
Okunevo Russia Verkhni Askiz Village [RISE516] F 2201-2036 BC

H6a1b 239C, 263G, 750G, 1438G, 3107T, 3915A, 9380A, 10589A, 15326G, 16362C, 16482G Allentoft 2015
Unetice Germany Leau 2 [LEA 4]
2200-1550 BC H6a1b3 G73A C146T C152T C195T T204C T239C A247G 309.1C 315.1CA16129G T16187C C16189T C16193T A16219G T16223C G16230A T16278C C16311T T16362C Brandt 2013
Unetice Poland Polwica [RISE145] F 2188-1958 BC

H6a1b 239C, 750G, 1438G, 3107T, 3915A, 4288T, 9380A, 10090T, 10589A, 15326G, 16362C, 16482G Allentoft 2015
Srubnaya Russia Spiridonovka IV, Samara River, Samara [I0358 / SVP 6] F 1913-1629 BC 273,856 H6a1a Mathieson 2015 Srubnaya
Andronovo Russia Solenoozernaïa IV, kurgane I, burial 4 [S13] ? 1800-1400 BC H6 16187T, 16362C Keyser 2009

Matt said...

@ Ariel, do you mean there are tables in Lazaridis 2016 that will make you think that Natufians are equally close to Vestonice and MA1, or that it just doesn't have any data on that either way?

Gioiello said...

The oldest H6 found so far is an European woman, and, even though only partially tested with 23andMe, we may see if she is negative for the sunclades:
CRS A263G A750G A1438G A4769G A8860G A15326G
H6 T239C T16362C A16482G
.. 23andMe('Jennifer Grannis'-huB2C416) H6 239C 263G 750G 1438G 4769G 10373A 15326G 16362C 16482G

@ Davidski
when you have some problem with haplotypes, both Y and me, ask me.

Ariel said...


That dont' have any data either way, but too me it's clear omission, it's like they can't model Natufians with WHG, let alone older europeans.

Ariel said...


This graph alone should pose some questions in that regard.

Matt said...

@ Ariel, thanks. Though I would say for whatever it is worth, there is some modeling of Natufians in Supplementary Information 4 and 7. Figure Figure S4.10 and Figure S4.11 which fits an ADMIXTUREGraph tree with Iran_N and WHG - Natufian takes a pre-West Eurasian ancestry edge from post ANE divergence, then post K-14 divergence, pre-WHG divergence (

I think their timescale was too tight to use the other samples from Fu et al 2016 is all, but there may be a deeper reason.

Natufians were also used in model free admixture tests in those sections above, though for me those seem harder to place weight on for me as they do not co-model ancient together populations

(Also in the models that use them, Natufians also notably do not fit as a sum of positive values of "The Four" - Levant_N, Iran_N, EHG, WHG - and have to go above 1 for Levant_N and particularly go negative for EHG and Iran_N.)

(As a tangent re: the Fst matrix you posted, you may be interested to see a couple of plot I have done to embed the Lazaridis Fst matrix for ancients in an MDS.

First with unadjusted values:

Then with population differentiation scaled so that the differentation from Africans is about the same at 0.183:

This is to adjust for WHG, SHG, EHG, Natufian probably having gone through more population bottlenecks and having less diversity while the other populations pool a lot of diversity from admixture.

There are commonalities between the two - Steppe_EMBA is about the same distance between EHG and CHG, Europe_LNBA about the same between EuropeMnCHL and Steppe_EMBA. But you really see in the adjusted version that the Euro_HG are way tight together relative to their genetic diversity, while the Near Eastern populations are very differentiated relative to genetic diversity. (Which is exactly what we find in f3 statistics).

Ariel said...

Thanks, the tree clears a lot of things, they have indeed a "western source" for Natufians, and it's not WHG, that explains the high fst values between them(and many d-stats and nmonte stuff that didn't make much sence if WHG was the western population admixed in Natufians). My speculation is that since Vestonice or Goyet don't fit either, we can imagine that there were a lot of non basal western european HG-like people in the middle east in the mesolithic. That then admixed with Natufians on a later date.

Matt said...

In terms of the timeframe, Natufians and earliest CHG (Satsurblia) are both at around 14000 ybp, late Upper Paleolithic by the European age system, at around the timeframe of Villabruna and Bichon.

So I think it's hard to think that Basal Eurasian wouldn't have pervasive in the intervening regions at this time (between the South Levant to Caucasus), before the European Mesolithic (wiki 12000 ybp - Neolithic).

Though agree there could have been people (at least in pockets) without Basal Eurasian ancestry around as well. Plus you may have also been talking about the Near East Mesolithic which I think is dated older.

Ariel said...

Yes, I think that the middle eastern mesolithic is 20000-10000 BC. I don't think that admixture event could be much older, especially if WHG came from the middle east. Many suggested that the WHG group might have originated in anatolia before the LGM, while BE was hiding in north Africa or the Arabian peninsula, a sister clade too WHG could have being hiding in the levant at this time (early mesolithic) before the spread of BE. If BE was in the middle east before the mesolithic than WHG could be native to Europe, and there was a magic forcefield that prevented people from going into Europe from Anatolia. But at the same time one should wonder how EHG got no admixture from the caucasus for so long (almost 10.000 years!).

Unknown said...

Ariel, I said that Anatolia Neolithic is basther basically Bedouin+WHG or rather Natufian+WHG as modern Bedouins have CHG, SSA and small ANE admixtures but they are close enough.

According to this thread WHG shares surprasingly quite a lot of drift with Natufians, more than with Iran Neolithic.