October 25, 2010

More detailed analysis of Eurasian populations (K=10)

I have removed some populations from the previous run (such as Moroccan Jews and Samaritans) that tended to generate mini-clusters due to the presence of close relatives and/or inbreeding in the sample. I have removed some redundant populations to even out the dataset, and I have also added North Kannadi and Gujarati, which helped reveal the gradient of ancestry in South Asia.

ADMIXTURE results:

Admixture proportions:


Some interesting observations:
  • The occurrence of 3.8% South Asian in Romanians may signify its Roma population. Indeed, almost all of this comes from a 25% South Asian individual, almost certainly a Roma.
  • The small African component in Spaniards which was revealed in a previous K=8 run turns out to be East African (0.5%) rather than West African (0.1%). If this holds up in larger sets then it might signify that its origin is from East African admixed populations from the east, rather than Sub-Saharan Africans.
  • The multiplicity of ancestries of the Uygur is made evident, in agreement with the extensive craniometric and genetic data on prehistoric and extant populations from the area.
  • The proportion of the two East Eurasian components in Turkic populations is interesting. It seems that the earliest departures from the Turkic homeland (such as the Chuvash and Yakut) have a predominance of the NE Asian component, the Anatolian Turks are intermediate, and the Uygurs, the only ones to have stayed close to the homeland, have experienced an increase in the E Asian component.
  • The absence of the West African component in Ethiopians is striking. Here are the individual results for Ethiopians, illustrating the variability of the Southwest African vs. East African components. The Ethiopian sample consists of a number different ethnic groups of the country, some of which (like the Amharas) are of Western Eurasian linguistic origin.

I am currently running K=11 and K=12 on the exact same data to see how the LogLikelihood and Bayes Information Criterion will move and whether new mini-clusters will appear, or if the mega-components (such as the "West Asian", "South European", and "North European") will split informatively. I will update this post with information on what actually happened, and with additional plots -- if I get robust results.

62 comments:

Onur Dincer said...

Why did you remove Iranians? They are a pivotal population to understand the transition from West Eurasia to South Asia and Central Asia.

Onur Dincer said...

Wish you also included Uzbeks.

ashraf said...

Thank you for including the south Indian Kannadi population and I do think that most likely the light blue western Asian cluster is the marker of "indo-european" component.
As this component is highly present amongst all the populations linked with the earliest indo-european speaking biocultures(Pathans, Turks, Cypriots...) and is also the biggest Caucasoid component amongst south Indians, besides its expectable scarcity amongst the linguistically recently indo-europeanised fringe regions such as Lithuanians, Spaniards, Sardinians+its complete absence amongst the non indo-european speaking (vasconic speaking) French Basques.
Proto indo-europeans introduced indo-european languages as well as patriarchality,ie religions and cremation to Europe but frm those 4only indo-european languages still remain nowadays(though the consequent european indo-european languages were much the result of local non ie speaking peoples adopting ie[an example could be slavo-iranian linguistic symbiosis]dialects with much local non ie loanwords and local phonetics=>gave birth to the ancient ie European languages)

Dienekes said...

ashraf, it will be interesting to see what non-admixed Finns look like in that component, another non-IE population of Europe. Also, Sardinians, who show continuity with their Bronze Age ancestors who were similar to Iberians, and perfectly in sync with the idea that they are a southern European relic population, relatively untouched by continental European phenomena.

Onur Dincer said...

BTW, Vologda Russians, Lebanese, Yemenese, Palestinians, Bedouin, Yemeni Jews, Iranian Jews and Druze aren't included either. Wish you included them too.

Onur Dincer said...

BTW, Vologda Russians, Lebanese, Yemenese, Palestinians, Bedouin, Yemeni Jews, Iranian Jews and Druze aren't included either. Wish you included them too.

Also Moroccans.

Lank said...

Thank you!

In case you did not know, many Amharas underwent a massive language shift recently from Cushitic to Semitic, even though there have been Semitic speakers in the area for thousands of years. The origins of their language is not an explanation for their higher Eurasian affinities, I would say.

Behar sampled Oromos in southern areas, which explains the differences. Amharas were actually pretty similar to Ethiopian Jews (Cushitic speakers from around the same area as Amharas).

Onur Dincer said...

If Ashraf's hypothesis is correct, proto-Indo-Europeans (who were living in Anatolia according to Ashraf), in concordance with geography, were genetically somewhere between Caucasian-speakers and Semites with somewhat more NW Mediterranean (or rather Balkan?) affinities.

BTW, I also wonder how the non-Caucasoid admixture of Hungarians would appear in this analysis. I guess it would be like Belorussians'.

AWood said...

Ashraf, doesn't your hypothesis pretty much fall apart when you realize Georgians peak the West Asian component and do not speak an IE language? Followed by Lezgins and Adygei who also do not speak an IE language? You cannot relate ADMIXTURE to the spread of language. The most isolated communities will only have 1 or 2 components because of a lack of gene flow. If there was an Indo-European invasion (which I don't think there was), we wouldn't see this much gene flow from West Asian. I suspect a good chunk of the genes that don't fall into either N. or S European are the result of neolithic settlement into Europe, not invasion. These people lived side by side with the pastoralists for thousands of years. Over time they have mixed in, in particular in the south of Europe. The Etruscans come to mind as some of these people in Europe, perhaps the sedentary cultures of the Danubian as well.

Dienekes said...

A particular (physical-genetic) type of people may spread a language; this does not mean that ALL people of that type spoke that language. For example, Spaniards, who are quite similar to Basques and French in terms of their ancestral proportions spread Spanish into the New World, but that does not mean that all people with Spanish like admixture proportions ought to speak Spanish.

It is also possible that a yet-to-be-identified subset of the major clusters may signify particular language family speakers.

Dienekes said...

Dienekes, Where are the Greeks ?

There are some in POPRES, which, unfortunately, is not publicly available. Anyone willing to point me to a Greek dataset, feel free to do so.

Lank said...

What are the East/North African clusters like at lower K values, and at what K values do they appear?

Onur Dincer said...

Ashraf, doesn't your hypothesis pretty much fall apart when you realize Georgians peak the West Asian component and do not speak an IE language? Followed by Lezgins and Adygei who also do not speak an IE language?

As I and Dienekes have explained, that doesn't necessarily follow Ashraf's hypothesis. Proto-Indo-Europeans might have been between Caucasian-speakers and Semites in their distribution of the "West Asian component". Correlation doesn't necessitate absolute association; in fact, it would be ludicrous to propose such an absolute genetic association among racially so close people (all of the relevant populations are racially Caucasoid irrespective of the language family).

ceraul said...

please eliminate the roma individual(gipsy) from Romanian set
for a accurate picture of romanians, i imagine that al european population have noneuropean minorities but this is not the object of present analysis

ceraul said...

maybe it's better to present in a separate column with other roma samples

ashraf said...

Dear mr's Aaron and Onur
The hypothesis of the Halaf culture of southern Anatolia-northern Syria being the ie homeland, has been advocated as soon as the 80's by Vyacheslav Ivanonv&Tamaz Gamkrelidze's book Indo-Europeans and Indo-European languages and it's not my own amateur theory of course=>it's a very voluminous and serious book.


"when you realize Georgians...IE language?"

This point does not stand, Turks actually do speak an Altaic language so what...!?[LANGUAGES ARE NOT STATIC AND DID NOT APPEAR IN SITU=>indo-european,kartvelian and afro-asiatic do descend from a very ancient hypothetical proto language wich in turn do descend from a more remoted(and primitive)"proto caucaso-nostratic" wich in turn do descend from the(more primitive)language of the first out of Africa modern human...]
Also who spoke about indo-european INVASION? I think the bronze age indo-european migration was similar to the older paleolithic&neolithic human migrations to Europe(the first brought first modern humans to Europe and the second brought first farmers)


The fact that ethiosemites do have more Caucasoid admixture than ethiokushites is most likely due to them having received at least 2 back-to-Africa Caucasoid migrations (the first one brought early Afrasan Kushitic and the more recent one brought Afrasan Semitic)while more inland and southern afrasan speaking ethiopians have their caucasoid admixture essentially from the earliesy Caucasoid back migrations that brought Afro-Asiatic languages to eastern Africa.

Back to the topic I think it would be also interesting to include kurds as well as the endogamous Yezidi Kurds.

Onur Dincer said...

neolithic settlement into Europe, not invasion

Aaron, Ashraf associates the spread of the Indo-European language family in Europe and Central Asia with the spread of Neolithic and agriculture in these two regions.

please eliminate the roma individual(gipsy) from Romanian set
for a accurate picture of romanians, i imagine that al european population have noneuropean minorities but this is not the object of present analysis


He/she is right, Dieneke.

Dienekes said...

The populations are as published, and I am not going to re-categorize them with different labels than they appear in the source data. I may occasionally detect and report on the presence of unusual individuals as I did with the Romanian Roma.

Onur Dincer said...

I think the bronze age indo-european migration was similar to the older paleolithic&neolithic human migrations to Europe(the first brought first modern humans to Europe and the second brought first farmers)

Ashraf, I understand from this statement of yours that you are completely differentiating between Neolithic migrations to Europe and IE migrations to Europe, am I right on that?

So my statement "Ashraf associates the spread of the Indo-European language family in Europe and Central Asia with the spread of Neolithic and agriculture in these two regions" should be false, at least for the European part even if not for the Central Asian part.

pconroy said...

In terms of Romanians and South Asian admixture, with the exception of that one individual of Roma descent, the other Romanians may have a slight amount of ancient South Asian admixture. The reason I say this is that in a similar analysis by Polako/Davidski shows South Indian admixture even in Irish. Here is the Global K=10 Ireland

Dienekes said...

In terms of Romanians and South Asian admixture, with the exception of that one individual of Roma descent, the other Romanians may have a slight amount of ancient South Asian admixture. The reason I say this is that in a similar analysis by Polako/Davidski shows South Indian admixture even in Irish. Here is the Global K=10 Ireland

I can't comment on other people's estimates as I am not familiar with their methodology.

In my estimate, there is one individual with 25.1% South Asian, one with 1.2%, one with 0.6%, and the remaining 13 ones have 0% South Asian.

eurologist said...

Ashraf,

Much of the associations you are trying to spell out could well be of much earlier origin than IE. Also, I have always found it very difficult to pinpoint a mechanism that would spread autosomal DNA with new languages in highly-established, diverse, and successful agricultural societies across continents - many millennia before organized warfare.

Lithuanians, Basques, Scandinavians - in the end, language matters little, isolation matters a lot.

Fanty said...

Polakos Global K12 (+ 23andMe project members) just came in.

There is a single Greek member in it, whos breakdown reads there like this:

Mediteranean: 0.360067
Northern European: 0.333892
West Eurasian: 0.253155
North African: 0.026986
East African: 0.007235
Central American: 0.006571
Oceanean: 0.005509
East Asian: 0.005421
North Eurasian: 0.001135
West African: 0.00001
South Indian: 0.00001
South American: 0.00001


There is also 3 Fins, with like 70% Northern European and virtually nothing of Mediteranean" but about 17% "West Eurasian" and 8% "North Eurasian".

(Lithuanians are still the top in Northern European in this here, but at only 72% this time)

An Estonian, 70% northern European.
21% West-Eurasian, 2% Mediteranean, 3% North-Eurasian.

A single Dutch with 65% NOrthern European and 4 Norse with almost 70%

4 Poles with Northern European 66%
22% West Eurasian, 9% Mediteranean.

3 Swedes also 66%ish

Hmm....a Slovenian.
60% Northern. 13% Mediteranean and 22% West Eurasian.

An Ukrainian... 62% North, 24% West Eurasian, 10% mediteranean, 3% Northern Eurasian.

2 Swiss at around 52% northern European but like 25% mediteranean.

3 Germans with northern European vary between 55% and 67% and their Mediteranean from 12%-20% and 19% West Eurasian.

Dienekes said...

There is a single Greek member in it, whos breakdown reads there like this:

Interesting breakdown; I can't comment on my own Greek participants, as I haven't received 5 Greeks yet.

His test seems to include Oceanian, and American groups and to group West Asians, whereas mine is focused on Eurasia and splits West Asians into "West Asian" and "Southwest Asian" categories.

Also, his definition of "South Indian" may be different from my "South Asian", as the latter is well-represented in Central Asia and may even be retermed Central/South Asian.

ashraf said...

Thnak you for you enrichening comments.part1
"many millennia before organized warfare."
Indo-European migrations to Europe and India(3500 bc-1500 bc)occured in a very period characterized by dramatic language shifts(see,as an example,the dissapearance of the Khatti language from Anatolia)and of course the freshly arrived(by 4000 bc)[light blue]"indo-european component" can not outnumbers the well established mesolithic and neolithic Europeans[orange and light green components!]
At hg Y-DNA level,neolithic Anatolian farmers(who,probably,were the builders of the great balkanian&iberian civilisations and probably introduced languages such as Iberian,Aquitanian,"Pelasgian",Eteocretan...)entering Europe fits well with current R1b distribution in Europe and more recently diffusing[most likely by the Balkans=>secondary ie homeland]Indo-Europeans fits well with J2[&perhaps also G or/and R1a or/and E1b1b]hg distribution patern in Europe.
To avoid lenghty OT please take a look at the threads below:
http://anthrocivitas.net/forum/search.php?searchid=422313
http://anthrocivitas.net/forum/showthread.php?p=86523#post86523
http://anthrocivitas.net/forum/showthread.php?p=105442#post105442
http://anthrocivitas.net/forum/showthread.php?t=6022

ashraf said...

part2/Back to the topic,I think it should not be forgetten that those autsomal studies are dealing with DNA wich is roughly speaking is a bunch of 4 different aminco acids wich in Turn are made of merely 4 elements: carbon, hydrogen, nitrogen&oxygen (but in various combinations)and there is a lot to be understood about this staff one of wich is the mechanismas that lead to the emergence of new isolated intrarcial clusters (such as the [endogamous]Druze&Mozabite clusters)as well as the fact that some clusters get absorbed/"eaten" by other clusters at higher K values.
Perhaps the Druze cluster ,as an example,emerged after some biocultural milieu in western asia (perhaps the hassuna culture...??) has long been isolated and perhaps due to religious taboos they did not mix with ancient neighbor biocultural milieus such as the matriarchal[probably Khatto-Caucasic speaking]çatal höyük culture IE it could be that habitants of the X biocultural milieu being patriarchal and worshipping,let's say,a male thunder god[Hadad,Zeus]would not intermarry with neighbors from the matriarchal Y bioculture who do worship,let's say,a mother godess instead, and this very long interbreeding process would perhaps shows itself at a Genomic level resulting perhaps in the emergence of that Druz cluster.
This all I can say, please post this(probably)my last comment and I am sorry for it being a bit lengthy and OT.

Onur Dincer said...

Fanty, what does Polako's West Eurasian component mean and in which populations does it peak? Also I wonder in which populations his Mediterranean component peaks. Thanks in advance.

Fanty said...

"Fanty, what does Polako's West Eurasian component mean and in which populations does it peak? Also I wonder in which populations his Mediterranean component peaks. Thanks in advance."

The Mediteranean peaks in Sardinians.
But different to the values I am used from his FNTDNA Files. In those, Mediteranean has over 90% in Sardinians while this time, it peaks in Sardinians, but with "only" 54-62%

It also spreads different to what I was used to.

The second highest in this "Mediteranean" are Druze (52%ish and Palestinians (48%ish).

Then Cypriots, Syrians, Jordanians, Egyptians, Ashkinazim and Spephardim all in in the 40-45%

Then the Italian project members and the Tuscans somewhere in the 38%

The Greek member with his 36%

The "North Italians" with like 36%
followed by French Basque with 32-35%

At 30% the Spanish, Turks and Morrocans start kickin in.

French and Romanians between 24-30%

Then Ethiopians, Adygei, more French, Germans and Hungarians in the 13-20% region.

Dutsh, British Isles, Scandinavia and the single Slovenian with 10-13%

The Ukrainian with 10%
Then Poles with 9%

hmmm
---------------------------------

WEST EURASIAN peaks in... Brahiu and Makrani. All 55-70%

Thats Pakistan.

At 54% Georgians and Adygei kick in.

Iran at 47%
Armenians 45%
Pathans, Turks 43% and downwards...

40%.. now there are some Sindhi...
38... lots of Druze, more Sindhi,

35% Burusho and Syrians start coming

34... Palestinians
32... Jordan
26%... AJews and SJews
25% the single Greek
24%... the Egyptians, Romanians and some Chuvash show up.

23% Russians and Poles make their apearance.

22% Lithuanians, Belorussians, Hungarians, the Slovenian and some Tuscans

21% the bulk of the North-Russians, more Russian project members, Gujaretti (Northwest India), the Estonian.

20% Germans, Swiss, Swedes and Irish show up. And it seems now comes the bulk of the Hungarians.

16%-19% UK, Norway, North Italians

15% and the French show up
12% first Spanish
8% Ethiopians, Sakilli
5% Yakut, Mongols, Mozabites
2% first French Basque

aargiedude said...

Dienekes, which of the Romanian samples is the possible Gypsy sample, using their Behar labels (Romanian1, Romanian2, ... , Romanian15, Romanian16)?

Would it be possible to post a huge image file with all the individual samples?

I agree with the samples you eliminated. I only think you should consider reincluding these guys: Iran, Yemen (not Jewish I presume), and Morocco.

The Gypsy sample from Romania, with 25% purple, would indicate that Gypsies are 50% "Gypsy" (Pakistani), since Pakistanis have 50% purple. The rest of Gypsies' ancestry would then be a combination of Iranian, Turkish, and southeast European.

It would be interesting to include the Mexican samples and see how their West Asian ancestry is clustered.

R1a and clusters. The purple cluster is present at 50% in Pakistan and 70%+ in India, but it's virtually inexistent in the 2 R1a-heavy populations of Europe: Lithuanians and Belorussians. Even what few purple exists in Europe seems to bear no relation to the diffusion of R1a: France has the most, followed by a tie between English, Belorussian, and Sardinian.

R1b1b2 L11+ (ht15) and clusters. No cluster apparently connected to the supposed recent expansion of R1b1b2 L11+.

Great work. This is very interesting.

Dienekes said...

Dienekes, which of the Romanian samples is the possible Gypsy sample, using their Behar labels (Romanian1, Romanian2, ... , Romanian15, Romanian16)?

Now that I've seen it again, it's two, one with 0.25 and one with 0.3.

GSM536589 and GSM536593

aargiedude said...

GSM536589 and GSM536593

These terms don't exist in the 12 GB and 3 GB matrix txt files.

Each single sample in the Behar matrix file has these labels:

Romania1.GType
Romania1.Score
Romania1.Theta
Romania1.R
Romania1.B Allele Freq
Romania1.Log R Ratio

Dienekes said...

I don't know what 3 and 12GB file you are looking at, these individuals definitely exist, e.g., http://www.ncbi.nlm.nih.gov/projects/geo/query/acc.cgi?acc=GSM536589

Look in the series matrix files as I told you.

aargiedude said...

Thanks. That link helped. They're Romania3 and Romania7.

Onur Dincer said...

Fanty, you wrote that the single Greek member has 33% Northern European component in Polako's ADMIXTURE analysis. What about other South European populations, and also French? Also what about West Asian, North African, South Asian, Central Asian and Siberian populations? Thanks in advance.

Anonymous said...

I would have liked Dienekes to have included Russians. I assume that their East Asian component is mainly NE Asian. Which Russians are they anyway?

Does the Cypriot set include Turkish Cypriots or only Greek Cypriots? Is it possible to compare these two groups?

On the Indo-European "invasion" / "settlement" question - All I1 descend from a single ancestor who lived less than 5000 years ago. http://www.eupedia.com/europe/origins_haplogroups_europe.shtml That implies that the rest of north central European males were wiped out by IE invaders. IE culture was associated with the warrior use of horses, chariots and bronze age weaponry. IE gods are warrior (eg. Odin) unlike the agricultural Neolithic gods. The Germanic pantheon is a combination of the two sets of gods, IE and Neolithic.

princenuadha said...

I'd have to say maybe the most interesting result I find is how nearly European specific the "north European" element is.

If you look at the European populations you can see that the NE component is significant in each population and relatively diffuse it is, but for the exception of Italy. You can also see how quickly the component vanishes outside of Europe with the exception of central Asia.

On thing I find extremely hard to believe is that the French Basque are composed solely of two components. How could they have zero north African and west Asian when the French and Spanish don't. I know there was limited gene flow but still...

Lastly does anybody have data on Switzerland. I really want to know what the Swiss are in relation to other Europeans. Sore far I've seen too many studies only looking at Genevans. I want to know where Swiss cluster.

Man I love this kind of stuff

eurologist said...

Fanty,

Thanks so much for your contributions (and of course, many thanks to Dienekes, as well!).

3 Germans with northern European vary between 55% and 67% and their Mediteranean from 12%-20% and 19% West Eurasian.

That's very close to my guesstimate ;) - especially, given that Polako uses additional categories (that tend to reduce the remainder).

WEST EURASIAN peaks in... Brahiu and Makrani. All 55-70%

Thats Pakistan.

At 54% Georgians and Adygei kick in.

Iran at 47%
Armenians 45%
Pathans, Turks 43% and downwards...


This is why I suggested Pakistan and NW India (and perhaps Iran and Afghanistan) need to be included. Either you get a slightly different vector and weighting, or at some K-level you will get additional separation into two W Asian groups.

Fanty said...

Additional thought on that "West Eurasian" from Polako:

The spread reminds me strongly of the suposed microsatelite diversity of R1a1a. In the way of: So more of that, so higher the diversity in R1a1a haplotypes. But maybe thats coincidence. -.-

Dienekes said...

I would have liked Dienekes to have included Russians. I assume that their East Asian component is mainly NE Asian. Which Russians are they anyway?

They are ethnic Russians from Vologda, and I have included them in many previous analyses where they invariably turn out to be part N European + minority NE Asian primarily.

Does the Cypriot set include Turkish Cypriots or only Greek Cypriots? Is it possible to compare these two groups?

The are sampled in the Republic of Cyprus, hence probably only Greek Cypriots. If any Turkish Cypriots want to send me their data, I'll be more than happy to include them in the Dodecad project.

On the Indo-European "invasion" / "settlement" question - All I1 descend from a single ancestor who lived less than 5000 years ago. http://www.eupedia.com/europe/origins_haplogroups_europe.shtml That implies that the rest of north central European males were wiped out by IE invaders.

You are assuming that the rest of N/C European males (non-I1 ones) are descended from Indo-European patrilineages.

If you equate Proto-Indo-Europeans with steppe groups (I don't), then the physical anthropological evidence is pretty clear that they did not penetrate far into Central Europe. They may have modified Central Europeans, but the local Neolithic element was more important.

There is simply no evidence for massive replacement of populations in Europe as a result of steppe migrations.

Fanty said...

I just NOW realize, I missed some posts.

"Fanty, you wrote that the single Greek member has 33% Northern European component in Polako's ADMIXTURE analysis. What about other South European populations, and also French? Also what about West Asian, North African, South Asian, Central Asian and Siberian populations? Thanks in advance."

I just finnished putting the averages for "Northern European" onto a map.

http://img204.imageshack.us/img204/9031/northerneuropek12neu.jpg

Some rather strange things show up. Like... the French Basque having a higher northern average than the 3 Germans. Or the French.

Also, the quiet high "northen" values on the Iberian peninsular compared to the Italian peninsular.

Hmm. I mean, ok... there is a strange mtDNA connection between Iberia and the Baltic region. With H aswell as with V. Even the Origin of U5 (over 50% in European Hunter/Gatherer corpses) is thought to be Iberia.

Dienekes said...

I just finnished putting the averages for "Northern European" onto a map.

What's the difference between numbers in red and in black?

Dienekes said...

What's the difference between numbers in red and in black?


Nevermind, it seems the numbers in red are from 23andMe members. It's hard to swallow that in Spain the "Northern European" element is the majority (51%). It'd take a lot of Celts and Visigoths to make it so.

Dienekes said...

Or 39% in Sardinia vs. my 0.8%

GrIQ said...

Not surprised at all about the 52% north-european of spaniards. After all Spain has 70 % of R-P312 and the main ancestry of spaniards were the Celt-Iberians. And Spain has the lowest middl-eastern input of all southern Europe. It's a myth that mediterranean are genetically similar.

Dienekes said...

After all Spain has 70 % of R-P312

R-P312 is not northern European specific.

Fanty said...

"Or 39% in Sardinia vs. my 0.8%"

The other calculation, wiich he did with FTDNA members included has less than 0.5% northern European in Sardinians and Basques both.

it looked like this:
(it was a K7 and limited to Europe, middle east and northern African populations)

http://img233.imageshack.us/img233/7219/balticnortherneuropean.jpg

But there had been different "anchor" populations.

In the one he made for FTDNA users the anchors that mainly affected most EUROPEANS had been:

Lithanians ("Baltic"), Chuvash("Ural"), Georgians ("Anatolia-Caucasus"), Saudi Arabians ("Middle East"), Sardinians ("Mediteraneans") and "Basque" ("Atlantic").

Now its totaly different.
All Europeans are only affected by:

Lithuanians "Northern European"
Sardinians "Mediteranean"
Pakistanis "East Eurasian"

yes ok, some 1-4%ish North Asian and African sometimes.

The effect had been that "Northern European" seemed to unite with what was "Atlantic" and "Mediteranean" united with former "Middle East".

The new "East Eurasian" spreads almost like former "Ural", but with greater numbers involved.

princenuadha said...

"What's the difference between numbers in red and in black?"

I'm still not sure what the black is but almost all the black populations are exactly the ones you have.

"It's hard to swallow that in Spain the "Northern European" element is the majority (51%). It'd take a lot of Celts and Visigoths to make it so."

How would you know? Do you know when, where, and how the north european component evolved and how it spread? I would actually like to know, but I don't think you could give evidence for the certainty required to know not only that 51% would be unusual but that if true it must have come from the Celts and Visigoths.

Fanty's map shows NE to be much more diffuse and has less dramatic drops in percentage than dienekes' (largely due to the differences in Sardinia and Turkey). I hope dienekes' is correct as it seems much more interesting.

princenuadha said...

Also does anyone have more detailed analysis of the Swiss population as a whole? I am also curious about the German Swiss separately and wonder where they belong in Europe.

Fanty said...

"I'm still not sure what the black is but almost all the black populations are exactly the ones you have."

Exactly.

Numbers in black are from various scientific gathered profiles.

Numbers in red are project members that send him their profiles. Some of these base on only a single person (like Ukraine and Greece for example.

I also made another mistake.
The Portuguise must be red, since its a project member.

AND...I did it again...
"Pakistanis "East Eurasian""----> WEST Eurasian... :(

I seem to have a serious problem with calling something east of myself WEST. He he he

Anonymous said...

Anything about the portuguese in this study?

Best regards

Fanty said...

"Anything about the portuguese in this study?"

Its 3 of them.

Their averages:

Northern European (Lithuania): 48% (47,48,49)
Mediteranean (Sardinia): 29% (33,27,29)
WEST-Eurasian (Pakistan): 13% (12,15,11)
North-African (Mozabite): 6% (5,6,8)
West-African (Yoruba): 2% (1,3,2)

The first has 2% NORTH-Eurasian (Yakut), the others 0%

Fanty said...

Hmm.
He did a new multi-dimensional-scaling with Greeks (now 2)

http://img220.imageshack.us/img220/6810/europe12labeled2.png

princenuadha said...

@dienekes

I noticthed that the 0.1% northeast Asian in Spaniards from the chart does not show up in the bar graph.I know I should be able to see 1/10 of a percent on the bar graph since I can see the 0.1% east Asian in the CEU.

I'm sure the tiny NE Asian element in Spaniards is just noise but I wanted to know if the chart or the bar graph was wrong on that detail.

Fanty said...

hmm. OK, I have all these 3 big European clusters from Polakos K12 in one picture.

As I have expected, the mediteranean and middle east clusters united. The result however is worth to be called "mediteranean" as its just that: it surounds the mediteranean sea.

That "WEST Eurasian" cluster has a pretty interesting shape.

http://img28.imageshack.us/img28/381/threebigclusters.jpg

Dienekes said...

I noticthed that the 0.1% northeast Asian in Spaniards from the chart does not show up in the bar graph.I know I should be able to see 1/10 of a percent on the bar graph since I can see the 0.1% east Asian in the CEU.

Probably the chart was done with the raw numbers, while the table is rounded in the first decimal point.

The East Asian in Utahns is:

1.321400e-03

And, the Northeast Asian in Spaniards is:

0.0008084167

Both round to 0.001 or 0.1% but the former 63% higher than the latter.

Onur Dincer said...

I think Polako should use more Greek samples (preferably Greeks from different parts of Greece, including Anatolian Greeks) in his studies.

Fanty said...

"I think Polako should use more Greek samples (preferably Greeks from different parts of Greece, including Anatolian Greeks) in his studies."

The problem is, if there is no Greeks who either tested or are willing to send their raw-data profiles, then you dont get them.

There is a thread in the FTDNA Forum and the 23andME Forums about this project, that anyone who wants to join, can join and send in their profiles.

I also ordered a 23andMe kit now, because it seems to work better with that, plus, there seem much more Europeans.

GR1 (the one in the map) is from Xios (father) and Evia (mother) btw. *looked on the project member list*

Onur Dincer said...

looked on the project member list

Is that list open to public and where?

Fanty said...

Is that list open to public and where?

Its a "google docs" document where project members can write backgrounds.

Its seperated in 23andMe members and FTDNA members.

https://docs.google.com/document/edit?id=1BesUZ4T0mTczn-H5kommfYfI57Rzbba4OuYK_2uQOxY&hl=de#

horacioh said...

Likely Ethiopian Jews are the result of the confluence between Jews of Elephantine and Meroe (south Egypt-Kush) and the ancient kingdom of Sabah, located in both banks of the Red Sea, -with stepped migrations waves to the land- and transferred to the Quara region when the Nilo waterfall begin
at the North of The Tana lake for preserve religion life ; It is extraordinary to know that the months description and Jewish celebrations have a clear prop, in the Hebrew language that they must have known in former times and that they lost except these liturgics words by oral tradition into Quara Cushitic tongue or including in "liturgy Geez" -very related to Hebrew, but a dead language today, with only liturgy use- and the Amharic Semitic spoken language, or the Agaw Cushitic "Quara", (Quara was the region where Jewish tradition, coming from these two vias, the Elephantine and The Sabah Kingdom, was kept). Spliting also the speaking tongue from the sacro liturgyc language like was common in Jewish tradition.
If the Ashkenazim with only 1200 years ancient, likely born in massive coming from Alexandria Egypt –first in number- and the Levant –second in number and before in times towards II century AE- To Central Italy in VII century AE and after the North and the East of Europe, they possess maternal mitochondrial markers mtDNA L1,L2 (4%) from Abyssinia, and the sibling N1 and M linked with L3 from Abyssinia – Ethiopia and Eritrea also Somalia-, this proves the existence of Abyssinian people in Israel or, it proves the fact that the former Israelite people had Abyssinian numerous masculine markers -E3b etc.- as well enough known, as and feminine markers like L1L2 etc., showing a feedback between people into Ethiopia and Israel in ancient times.
Remember that the Ethiopians in whole are linking with Semitics languish, traditions and genes also with the Horn of Africa; and the Ethiopians Jews are very close to Ethiopians Christians Amhara and others Semitic people of Abyssinia.
The Greek, as far as my theory, had seen that singularity in two ancient Jewish communities about IV century BC, one in the banks of the Red sea, from there, perhaps, the famous Greek legends of Andromeda about the Ethiopian kings of Jaffa (Joppe) – nowadays seacoast of Israel-. Come, in a mixture of religions and traditions, and the other Jewish community in Israel and Judaea according to the times. Also Sambation - also Sabath one- legend about the Jews living in the Nile falls. The find of the Adulis Stone in 1960 – near to Asmara - with inscriptions is a display from a king (likely a Melik Jew king) towards the Ptolemy IV Philopator Faraon, III century BC. All I wrote is Supported in current archeological and genetics studies.

horacioh said...

Permanent villages and towns around Asmara predate, and were also contemporaneous with, even the pre-Aksumite settlements in the highland of Southern Eritrea and northern Ethiopia.
Dating –C.14- from 800 BC, (it is the walls discovered) -- not sites in Arabia -- that were the vital precursors to urban developments in the southern highlands of Eritrea and northern Ethiopia later in the first millennium BC
Likewise, students of evolution and distribution of languages now believe that Semitic and Cushitic At their foot survive. Languages are of African origin. Yes a lot of revision and rewriting of textbooks is called for!.
Knowing now that Geez are so or more ancient than Hebrew and Geez was spoken and written at first in ancient Semitic Hebrew figures since 3000 B.C.
A short track leads past even more pre-Aksumite ruins at Metera, one of Eritrea's World Heritage sites. Like Sembel, they are adjacent to productive agricultural land. Dressed stonework, artistically designed courses, a covered tomb chamber –where now is testing bone and teeth DNA samples- and several substantial walls of this early settlement At Metema, a unique pre-Christian inscribed stele (or obelisk) once stood 4.68metres above the ground (and another meter below it). But today its base, 80 X 28 cm, lies shattered and scattered up to 40 meters by high explosives clearly deliberately and carefully attached by the retiring Ethiopian army; such a tragedy cannot reflect well on Ethiopia's ongoing campaign to recover an Aksumite obelisk which Benito Mussolini had shipped to Rome and erected near the Gate of Constantine.
Certain of these earlier sites, including Metera and Kohaito in modern Eritrea and Yeha and Aksum just across the border in northern Ethiopia, have been designated world Heritage Sites by UNESCO -- but their supposed origin as Sabaen colonies established in pre-Islamic times has not previously been questioned.
The fact is that recent surveys reveal that far from being an insignificant forerunner of Aksum, the ruins on the Kohaito plateau in Eritrea indicate very extensive settlement. documented archaeological sites there to date number more than 1, 000. The names of many of its emperors are known from inscriptions and a coinage and there are near contemporary accounts of its adoption of Christianity in the 4th century
Likewise, ancient African harbors on the Red Sea and Gulf of Aden -- including Suakin in modern Sudan, Adulis on the Bay of Zula east of Massawa in Eritrea, and Zeila and Berbera in Somaliland -- are well known from inscriptions in Egypt and elsewhere and from mention in manuscripts.

Onur Dincer said...

Fanty, thanks for the link. The list is incomplete (I have seen more project members in Polako's genetic analyses), but still it is a good source since it provides valuable information about the backgrounds of some project members (like GR1).