February 17, 2010

Basques not genetically distinct from other Iberians

Human Genetics doi:10.1007/s00439-010-0798-3

A genome-wide survey does not show the genetic distinctiveness of Basques

Hafid Laayouni et al.


Basques are a cultural isolate, and, according to mainly allele frequencies of classical polymorphisms, also a genetic isolate. We investigated the differentiation of Spanish Basques from the rest of Iberian populations by means of a dense, genome-wide SNP array. We found that F ST distances between Spanish Basques and other populations were similar to those between pairs of non-Basque populations. The same result is found in a PCA of individuals, showing a general distinction between Iberians and other South Europeans independently of being Basques. Pathogen-mediated natural selection may be responsible for the high differentiation previously reported for Basques at very specific genes such as ABO, RH, and HLA. Thus, Basques cannot be considered a genetic outlier under a general genome scope and interpretations on their origin may have to be revised.



DagoRed said...

This is a clear example of as Clothes doesn't make the monk and the language doesn't make the ethnos.

Heraus said...

The very notion of Basque culture being a cultural isolate is dubious. How can Basque culture be an "isolate" when it's the basis of vernacular cultures in NW Spain and SW France, from Burgos to Bordeaux, an area greater than Dutch-influenced Europe ? That the Basque language was preserved in Western Pyrenean valleys is a fantastic miracle but that doesn't make Basque culture different from its immediate romanicized Cantabrian, Aragonese and Gascon neighbours.

There is a trend in academic sectors to make Basque people some "good wild men". For instance, it now cannot be doubted that inscriptions in Veleia were believed to be forgeries because they did not suit some well-known researchers' conception of the Basque people.

visvakarman said...

The Veleia inscriptions were clearly forgeries, no matter what they suited.

Andrew Oh-Willeke said...

Basque country was one of the first industrial centers in Iberia. This doesn't necessarily make the case for it being a cultural isolate, but it does make the case for the Basque being culturally distinct in material ways in addition to language.

We also know from history that Basque culture assimilated less from the Romans, the Moors and the Fascists than other parts of Iberia. Thus, the case that what is distinctive about Basque culture derives from having less of its pre-Roman culture obliterated is strong, and the fact that Basque is not an Indo-European language supports the theory that it was also not influenced as much as much of the rest of Europe by a hypothetical pre-historic Indo-European wave of migration.

The fact that Iberia was a refugia for European populations of animals (like bears) during the LGM also suggests that it may have served a similar purpose for ancestors of the Europeans of 45kya to 18kya, which suggests that to the extent that distinctly Iberian pre-Moorish genetics can be parsed from Vandal, North African and Roman influences, that they it's reasonable to suppose that its inhabitants are closer to European ancestral genetic makeup than the rest of Europe.

The lack of Basque genetic distinctness despite its cultural uniqueness points to the power of even low levels of admixture over long periods of time acting on a small group immersed in a larger one.

Heraus said...

I've long believed that the Veleia inscriptions were fake ones (because I'm highly skeptical of some Basque sectors which might be keen to emphasize on Basque exceptionalism) but newer studies have just proven that linguistic analyzes were at least very biased. Gorrotxategi and the likes were quite dishonest when analyzing linguistic elements.

Franco-Galician linguist, Hector Iglesias will soon testify in court in defence of Veleia archeologists (a French interview) :


His paper :


Another paper :


Heraus said...

Bad link :


Maju said...

It's contradicted by Bauchet 2007, which found a Basque-specific cluster in a large European study at K=4 or K=5, where Basques were not oversampled. Spaniards/Iberians are also a distinct group from other Europeans/Mediterraneans.

It's difficult to evaluate the validity of just an abstract that talks only of Fst (a criticized measure) and of old kind of genetic research such as ABO, Rh and HLA. Also the researchers work for Spanish institutions, what may imply ideological bias.

aargiedude said...

If somebody can obtain the supplementary material, could you tell us which 10 populations were studied, samples per population, and FST distances (there should be a table somewhere).

Maju, can you tell me more about the criticisms regarding FST?

Structure said...

In my last STRUCTURE IntraEuropean analysis that uses HGDP samples and data from DEcodeme/23andme, Basque appear in their own cluster at k=3. But this cluster may represent a Iberian cluster as well.


Basque group is 51 in mostly blue
Red are CentralNorthEuropeans
Green are SouthEasternEuropeans (Ashkenazims, SouthItalians, Turks...)

onur said...
This comment has been removed by the author.
onur said...

Structure, can you name the populations yuu used with reference to their test number and population size? Also I would be very grateful if you provided the whole STRCTURE analysis diagram with additional information about test methods and criteria.

onur said...

Erratum: STRUCTURE analysis, not STRCTURE analysis

Kepler said...

I thought Maju was going to say all Northrern Spaniards are in reality Basques or Proto-Basques who lost their language :-)

Actually: couldn't this be the case?
Vascoiberismo? Iberian language was somehow related to Euskera?

Dienekes said...

It's contradicted by Bauchet 2007

Not really, since that study tested only a sample of 8 Basques and Spaniards only from Valencia

and of old kind of genetic research such as ABO, Rh and HLA

What they're saying is that the supposed distinctiveness of Basques in these individual traditional systems doesn't hold up when we look at dense microarray data (thousands of markers)

eurologist said...

As others have alluded to, seems to me that the "Basque element" is simply largely spread throughout Iberia - and why shouldn't it be?

eurologist said...

"Pools from the following 10 regions were prepared: Galicia, Cantabria and Asturias, Catalonia, Castilla-León, Castilla-la Mancha, Valencia, East Andalusia and Murcia, West Andalusia, Extremadura and Balearic Islands (see Figure S1). All pools were created using 30 individuals..."

F_STs seem to be largely between 0.013 and 0.017. There are a couple of 0.012 and 0.018, one 0.019 - all outside of the Basque comparison. Basque comparison to others is between 0.014 and 0.017 - that is, slightly closer than non-Basques among themselves.

Maju said...

"Maju, can you tell me more about the criticisms regarding FST?"

Not really, I'm just echoing previous criticisms appeared at comments at this blog in relation to other studies (not about Basques), which seemed reasonable to me when I read them. You will have to search them yourself.

"I thought Maju was going to say all Northrern Spaniards are in reality Basques or Proto-Basques who lost their language"...

It's possible in some cases but I just don't know or could back it with hard evidence.

Anyhow, I understand that the historical Basque core area is Aquitaine and hence comparing with Iberians is generally going to produce rather distinctive results. That's also why I'm interested in more extensive and detailed sampling of French, specially Occitans (South French).

"Vascoiberismo? Iberian language was somehow related to Euskera?"

As I was just mentioning in a private email conversation this relation is not well established and in any case should date to the Epipaleolithic/Neolithic period.

"Not really, since that study tested only a sample of 8 Basques and Spaniards only from Valencia".

I did not notice before that the Spaniards of that study are from Valencia but that doesn't make them least valid, specially as they show only limited relation with Greece in a secondary component and instead they make up their own distinctive cluster. Valencians even show some amount of "Pyrenean" R1b1b2a1a2c, which is almost null among Western Iberians.

Valencians are also the archetypal Iberians, not just in the historical Iberian sense but also in the Paleolithic, Epipaleolithic and Neolithic sense: all these phenomenons are centered around the Southern Valencian Country in the distinctive Iberian province (Iberian peninsula minus the Cantabrian strip, more related to Southern France instead, at least most of the time).

So I think that the difference stands, though maybe can be less clear cut if you take many samples from areas near the Basque Country/Pyrenees.

"What they're saying is that the supposed distinctiveness of Basques in these individual traditional systems doesn't hold up when we look at dense microarray data (thousands of markers)".

Well, a Basque-only autosomal cluster at K=5 in a pan-European study (Bauchet again: 10,000 autosomal SNPs) is a clear signal of distinctiveness to me. And even more as Basques were only 8 individuals (small sample, which can't bias the results).


Anyhow, I was looking at Tian 2009, who also uses Fst distances (and also fails to sample any French) and while the closest population to Basques by this measure are Spaniards (0.0060), Basques are not the closest population to Spaniards but Germans (0.0015), Tuscans (0.0023), East Europeans (0.0033), Irish (0.0037) and Swedes (0.0065).

So it looks like, in comparison, Basques are still somewhat apart.

Maju said...

Erratum: "... and Swedes (0.0065)" should read "... and Swedes (0.0055)".

Also, I notice that the second closest group to Basques are Germans (0.0079), not much more distant than Spaniards, reinforcing my impression that French, in particular SW Fench (Gascons and others) should be much closer than Iberians.

eurologist said...

Actually, "closer" is not the right word: I should have said "more homogeneous distance" or something like that.

Extremadura has the largest F_STs tot he others, averaging about 0.017 (between 0.015 and 0.019).

Structure said...

onur said...
"Structure, can you name the populations you used with reference to their test number and population size?"

ID with population size in parenthesis

40 (61) Caucasians North Americans (non Ashkenazis)
41 (6) Finns
50 (30) French - HGDP
51 (24) Basques - HGDP
52 (4) South Italians
53 (9) Central Italians (Tuscans) - HGDP
54 (13) North Italians (Bergamese) - HGDP
55 (16) Orcadians - HGDP
56 (25) Russians - HGDP
57 (17) Adygei - HGDP
58 (4) Turks
60 (38) Ashkenazis

The SNPs used for this anaysis are the ones selected by Tian 2009 in his last study "European Population Genetic Substructure"


These 3519 Euro SNPs were selected by Tian because of their relatively large allele frequency differences between different European groups. They are currently the most accurate for Intra European analysis and much more effective than thousands of random SNPs used by previous European studies. They also appear to be almost as effective than 500.000 SNPs, but their bigest avantage is that we can use much larger samples in STRUCTURE.

Kepler said...

Actually, I always thought I could recognize many Basques from other Spaniards, even Northern Spaniards: there is a certain nose form, eyes and eyebrows that tend to come up very frequently there unlike in Galicia or Asturias or even less in places as castilla or elsewhere.

onur said...
This comment has been removed by the author.
onur said...
This comment has been removed by the author.
onur said...
This comment has been removed by the author.
onur said...
This comment has been removed by the author.
onur said...

Thanks, Structure. Looking at the graph, if we assume that Basques are a good representative of whole Iberians, three major clusters appear in Europe: North European (red), West Mediterranean (blue) and East Mediterranean (green).

Iberians on one side and Turks (and probably also Greeks) on the other side constitute the opposite ends of the West Med and East Med spectrum respectively, being almost totally composed of their main clusters.

North and Central Italians are somewhere between West Meds and East Meds with a nonnegligible North European admixture. South Italians are almost indistinguishable from East Meds, confirming historical narratives of Greek colonialism there.

As to West and Central North Europeans, they are mainly North European with significant Med admixture (mainly West Med). Northeastern Europeans, on the other hand, having less means of access to Southern Europe than West and Central North Europeans, have very insignificant Med admixture, and as such, they represent the extreme end of the North European cluster.

Ashkenazis seem to preserve most of their original East Mediterranean (not exclusively Middle Eastern, they may have significant Northeastern Mediterranean roots (google Jewish proselytism in the Roman Empire (especially in the east) in pagan era)) genetic heritage, with a relatively small European admixture during their long stay in Northern Europe.

miz RAND BLOWTON said...

I never thought the Basques were different from many other authentic Spaniards,I just thought the Basques had wanted to maintain their own culture.And culture can make a person very different...depending on what it is you like to do or have been doing for many years,or choose to do.