Monday, May 28, 2012

My Oracle results (Population Sharing)

I'll come back soon with new samples when I get time to upload them. For the moment, just a tiny gadget one can find on GEDmatch. The results obtained are rather interesting as it gives sharing distances with many populations plus it creates "ghosts" made of two populations which constitute a basic approximation of one's genomic characteristics.

The various algorithms don't possess the same populations, I used two of Dienekes' calculators : K12b and World9. Here are my results : check here to know about my Gascon/Pyrenean background.

- Oracle "Dienekes' K12b calculator" :

- Oracle "Dienekes' World9 calculator" :

- Comments :

1. Single Population Sharing :

The comparison of both results clearly proves something I had long anticipated : the French Basque sample from HGDP, the one used in most studies found on the Internet (including 23andme I believe), is pretty specific, quite aberrant actually and might be very inbred or quite isolated. The second study, with another French Basque sample, shows results one can expect : the French Basques are my second best match (with a distance of 2.87) after the Spanish Basques (at 2.53 : the difference is negligible).

Other results are pretty consistent : after the Basques, I share most with Aragonese people then Cantabrians. There is nothing strange about me - a Pyrenean Gascon guy - sharing high with (High) Aragonese people who are our neighbours and with whom we share the same "Vasconic" history. 

Cantabrians are more of a surprise : I had come to accept that they had long ceased to share the destiny of "Vasconian" people and some genetic results seemed to show they were more akin to people from Western Iberia (Galicia, North Portugal, ...). My anthropological amateurish research also had made me believe they were strandard Iberian-looking phycally wise, not that Basque-looking. Add that their "Celtic" substrate has long been over-emphasized. It looks like things are blurrier.

And then the usual populations : Catalan people (with whom Gascon people might share an ancient history as far as Y-DNA studies are concerned : quite probably Catalan people were then subject to subsequent migrations which differentiated them hence Iberian which appears to be a Basque-based pidgin) then other Iberians and the French ("French" is a stupid category : it doesn't mean anything, we really lack regional studies in the "Hexagone").

2. Mixed Mode Population Sharing :

The "ghosts" created by the first analyzis are rather interesting. They confirm that the French Basque sample from HGDP is very specific as I'm only about 60% French Basque (and 40% French) when compared to them (as I appeared on charts on 23andme). On the contrary, I'm about 87% "Spanish Basque" (the remainder being generally a far-away population which hints to the fact the algorithm doesn't really know what to do with what remains of my genetic variation). Other interesting ghosts is that I could be analyzed as being 63% French Basque and 37% Catalan which is quite on par with where I was born.

The "ghosts" created by the second calculator (with many global populations) are less interesting but maybe more accurate in a West Eurasian context : I happen to be mostly a Basque (either from Iberia or France) with greater admixture from the Levant (to whom the Kalash people are a proxy).

In all cases, I believe I may be differentiated from Basque people by the fact we, Béarnais people, may have received more influences from Neolithic migrations but not that much actually.


  1. this is really cool. how does one get a test like this done?

  2. Excepting that some, including me, may find the terms "French Basque" and "Spanish Basque" rather offensive, imperialist (we mostly use "Northern Basque" and "Southern Basque" respectively, as you know well), I think you are right with most of your observations, some of which I find interesting:

    1. The "French Basque" HGDP sample not being good is interesting to know.

    2. The genetic reality of Cantabrians being quite more interesting than a mere extension of generic Iberians is something I think I can agree with.

    The elements of North African affinity in West Iberia may well be very very old after all because there's no obvious archaeological reason to think that Cantabrians or Asturians were replaced at any moment since the Upper Paleolithic, hence their differences with Basques may be something formed for example in the Magdalenian... or who knows but it's not something clear nor too clear cut either.

    Whatever the case, I do not feel that Asturians and Cantabrians are typical Iberians in phenotype either.

    3. Catalans and Gascons (and maybe some Occitans-Languedocines but they are ill-researched) may share some genetics. However I thought that the bulk of the coincidence was Y-DNA and not so much mtDNA. I wonder why do you say mtDNA.

    4. Bearnais may be a tad more Neolithic than Basques but not much. Emphasis in not much. Among Basques also it depends where you take the sample.

    In any case, the data is always in reference to your individual case, which may not be too useful of a reference to infer too much about populations (i.e. your individual peculiarities may and probably do distort the matter overall a bit).

  3. Sorry, I wanted to write Y-DNA studies for Catalan people, I had the many subclades of R1b in mind.

    You're also right "French Basque" and "Spanish Basque" are indeed quite offensive (and meaningless) but I used them only because they're the ones used in most studies found online. I'm quite aware of the implications behind such terms. We could have a discussion about whether or not "Northern Basque" is a valid term though : North Basques are actually East of "South Basques" as one Basque nationalist made us notice (I cannot remember his name now). The old terms in the 9th century used to be "SpanoVasconia" and "Aquitania" but they're out of fashion I think.

    I may be a bit hasty when stating that the "French Basque" HGDP sample is not that good, a fact is that two studies with two different "French Basque" samples show quite distinct results. But who knows ? I cannot say for sure. My intuition is that the HGDP sample is very localized.

    Eventually, I must agree that those are just individual results, I can only lament the fact ...

  4. @Anonymous :
    You've got to get your genome tested by commercial companies (it may be quite expensive, I took advantage of a discount on 23andme) then you just have to submit the file you get from the aforementioned companies on GEDmatch.

  5. Now I remember the name of the Basque patriot : that's Txillardegi. An interesting book :



I've chosen to let people comment freely on my posts. Nevertheless, you'll lose your time taunting me and calling me a fascist (which I'm really not) : I pray you to read my introduction which will reassure that my intentions genuinely aim at achieving amateurish knowledge. I understand that you may not share my passion for the history of the peopling of the World, just don't let me know as clear conscience gained by bashing a humble documentary work is useless.