Dodecad Project identifies Indo-European autosome from North Caucasus

Maciamo

Veteran member
Admin
Messages
9,970
Reaction score
3,273
Points
113
Location
Lothier
Ethnic group
Italo-celto-germanic
Dienekes just posted a new autosomal analysis from his Dodecad Project : Fine-scale admixture in Europe (Dagestan/Basque/Sardinian components)

He managed to split the West Asian component in two new components :

- West Asian (read Caucasus/Anatolia/Assyria/Persia), which peaks in Georgia and Assyria (71%), followed by Armenia (68%), Iran (66%), Turkey and the North-West Caucasus (55%).

- Dagestan (read North Caucasus), which peaks among the population sample from Urkarah in Dagestan (93%), the Lezgian speakers of southern Dagestan (48%), and the Dagestani from Stalskoe (39%), but is also well represented in the North-West Caucasian Adygei (16.5%) and the Georgians (12.5%).


The West Asian component decreases progressively with the distance from West Asia, and is virtually absent from Nordic countries as well as among the Basques. It seems to correlate well with the spread of agriculture. The frequency diminishes along the two migration routes followed by neolithic farmers. The southern one would have passed through Crete and southern Greece to reach South Italy (42% in Sicily), then up to central Italy (28% in Tuscany) then North Italy (19%), France (7%) and Germany (5%), or from Sicily, following the North African coast, reached Spain (11%) and Portugal (13%), but bypassing Sardinia (2.5% probably of later Phoenician and mainland Italian origin). The Danubian route started from Greece (36%), and would have reached first the Balkans, then Romania (24%), and Hungary (11%), while an eastern branch from Romania continued to western Ukraine and Belarus (9%).

The main breakthrough, though, is that the Dagestan component is remarkably strong is northern Europe, particularly in Germanic countries. The highest frequencies observed in Europe are found respectively among the Orcadians (12.5%), Scandinavians and White Utahns (11%), Germans (9%), French (8%) and Hungarians (7.5%). The lowest level of Dagestan admixture is found among the Basques (0.7%), Sardinians (1.5%) and Cypriots (2%).

Better still, the Dagestan admixture is found as far as South Asia (see admixture), while the West Asia one is not. This Dagestan DNA represent 19% of the genomes of the Pathans, 10.5% of the Sindhi, 7% of the Gujarati, 11% of the Tamil Brahmins, 10% of Andra Pradesh Brahmins, but is rare among Dravidian people and almost completely absent from the lower castes and tribal populations. I cannot see a better evidence that this Dagestan component in fact represents the autosomes of the Indo-Europeans.

Note that the components used for the European and South Asian admixtures are not identical, and as Dienekes warns, we shouldn't compare the percentages between the two runs. However they both represent the same ancestry from the North Caucasus, which is precisely where I hypothesised that the Indo-European migrations originated. There is also a triple correlation with Y-DNA, since haplogroups R1b1b, R1a1a and G2a3b1a are all present in the North Caucasus, Europe, Central Asia and South Asia. The North-East Caucasus is a region that has been found to harbour a particularly high frequency of each haplogroup (up to 40% of R1b1b; 92% and 100% of haplogroup G in the Dargin and Kubachi of southern Dagestan respectively, although as low as 6% of the Nogai of northern Dagestan).

I would think that this Dagestan component represents more G2a3b1a and some R1b1b, while the North European component would correlate especially (but not exclusively) with R1a1a. The Basque and Sardinian admixtures, found in high densities as far north as Scandinavia, are probably the genetic contributions of Paleolithic Europeans (haplogroup I). The admixture should probably be refined as I cannot find any useful pattern for these new Basque and Sardinian component, except that Basque is a bit more West European, and Sardinian a bit more South-East European.
 
I find this very interesting since autosomal DNA has always seemed the "800 pound gorilla in the room" to me. I don't know enough to add anything except that I hope that those who know more than me can help me understand this.

The Lemba in Zimbabwe have very high Semetic DNA, and in some cases, the Cohenim Modal, but their autosomal must be sub-Saharan African surely.

So the question that I ask myself is how meaningful is the Y-DNA and mtDNA analysis after all?
 
The "Daghestan component" as you call it could also be remnants of the IJ macro-group seen in autosomal mode.
The south-west asian component correlates most strongly with M267 thus pointing it as a north-afrasan marker.
 
As I understand from above, Daghestan group is defined more strongly in Urarkah. That is the region where Dargins or Dargwa people represent great majority of population (over 90%). As we know from recent studies Dargins have 58% of haplogroup I (Nasidze paper from 2004). How does this results fit in all this story about indoeuropean caracter of Daghestan group as we know that haplogroup I is not Indoeuropean by origin and that language of Dargins are also not Indoeuropean.
 
The "Daghestan component" as you call it could also be remnants of the IJ macro-group seen in autosomal mode.

If Dagestan correlated with haplogroups I and J, it would be over 50% in the Middle East, at least 30% in Scandinavia, but be tiny in India. It's not at all what is observed.

The south-west asian component correlates most strongly with M267 thus pointing it as a north-afrasan marker.

There are already Northwest African and East African markers that correspond well to E1b1b. Southwest Asian is the Arabian peninsula admixture of E1b1b, T and J1 (and perhaps some rare haplogroups like C5 or Q*).

Egyptians, for instance are 15% East African and 8% Northwest African and 38% Southwest Asian.

Saudis, on the other hand are 76.5% Southwest Asian, 12% West Asian, 4.5% European, 4% sub-Saharan African, 2% South Asian and 1% Northwest African. Based on this study of Saudi Y-DNA, the 12% of West Asian should corresponds to the 16.5% of hg G1, G2 and J2. The 4.5% of European would be the 6% of R1a and R1b. The 2% of South Asian are obviously the 3.5% of H and L. The 4% of sub-Saharan are the 11% of B and E(xE1b1b). This leaves us only with 42% of J1/J*, 7.5% of E1b1b and 5% of T, so 54.5% of "native" Southwest Asian. The higher proportion of native autosomal DNA can be explained by the fact that, most of the time, the foreign haplogroups that entered Saudi Arabia were already mixed autosomal with neighbouring countries, and also by the higher proportion of native women than men (something that happens almost everywhere, explaining the great phenotypical differences between R1a countries, for instance).
 
As I understand from above, Daghestan group is defined more strongly in Urarkah. That is the region where Dargins or Dargwa people represent great majority of population (over 90%). As we know from recent studies Dargins have 58% of haplogroup I (Nasidze paper from 2004). How does this results fit in all this story about indoeuropean caracter of Daghestan group as we know that haplogroup I is not Indoeuropean by origin and that language of Dargins are also not Indoeuropean.

The Nasidze study is old and not very reliable because it failed to test many key markers.

The Marchani study (2008) of Dagestan Y-DNA found 100% of hg F or G among the Dargins and 92% of F or G among the neighbouring Kubachi. They didn't find any haplogroup I among the 5 Dagestan populations studied.
 
The Nasidze study is old and not very reliable because it failed to test many key markers.

The Marchani study (2008) of Dagestan Y-DNA found 100% of hg F or G among the Dargins and 92% of F or G among the neighbouring Kubachi. They didn't find any haplogroup I among the 5 Dagestan populations studied.

Does it means that all haplogroup I presented in that study (among Ossetians, Abkhazians) is false by origin. This is new information for me. I didn't know for new study for Daghestan. So, the presence of I haplogroup in Caucasus region is less than symbolic.
 

This thread has been viewed 12440 times.

Back
Top