What can autosomal calculators tell us about ancient Greek admixture?

Alexandros

Regular Member
Messages
66
Reaction score
10
Points
0
Ethnic group
Greek
Y-DNA haplogroup
R1a1a (R-L1029)
mtDNA haplogroup
H14
There are plenty of autosomal calculators around (http://gedmatch.com/) and I have used practically all of them in order to identify any hint that would point me to the ancient Greek admixture. I will use data from the Dodecad project (k12b calculator) and concentrate on 3 populations who, I believe, may be giving us some hints: the modern Greeks, the Greek Cypriots and the Lebanese. A question that automatically arises is 'what do the Lebanese have to do with the ancient Greeks?'. An easy answer would be 'not much', which is quite accurate. The Lebanese can thus be seen as an eastern Mediterranean population not related to the Greeks; a sort of 'negative control' if you like. Below are the normalized median proportions for the admixture of these 3 populations.

Population
Gedrosia
Siberian
NW_African
SE_Asian
Atlantic_Med
North_European
S_Asian
E_African
SW_Asian
E_Asian
Caucasus
Sub_Sah
Greek
3.3
0.3
0.6
0
28
20.2
0.1
0
10.1
0
37.4
0
Cypriot
5.8
0
2.1
0
20.5
4.5
0
0
17.7
0
49.3
0
Lebanese
10.8
0.2
4.7
0.6
11.8
3.7
0
2.4
23.5
0
41.3
1

If we start from left to right, Cypriots seem like an almost perfect match between Greeks and Lebanese, that until we reach the North European admixture. The North European admixture is astonishingly high among the mainland Greeks compared to the Greek Cypriots. One could easily conclude that then basically Greek Cypriots are not at all Greek, but merely a West Asian (i.e. Lebanese) population with a bit of European admixture. There is a problem though. If you look at the Atlantic_Med admixture (a West European component), Cypriots have twice the frequency to that of the Lebanese. This admixture could not have been introduced to Cyprus by any nation other than the ancient Greeks, simply because no other nation from mainland Europe migrated 'en masse' to Cyprus during ancient or modern times. There is strong historical and archaeological evidence for mass migrations from the Mycenaeans to Cyprus between the 12th and the 5th centuries BC, followed by practically no major migration from Europe afterwards. What's with the North European component though? Shouldn't the Cypriots have received a substantial component from those migrating Greeks? The answer is yes, if it was there.. Interestingly, we find archaeological evidence only for Mycenaean culture in Cyprus but practically no evidence for Dorian culture. I speculate here that the Dorians are a major source of North European admixture among Greeks, which failed to reach hellenized places like Cyprus. This North European component among modern Greeks probably also received a later boost from the Germanic migrations of the 2nd-6th century AD and to a larger extent the Slavic expansion from the 6th century onward, which again did not reach Cyprus. Of course these are speculations, but I would be very interested to read your views.
 
Last edited:
Makes a lot of sense; Greeks have the largest European component whereas Lebanese are middle eastern and Cypriots are a conservative mix more similar to Crete.
 
Sounds about right. I would like to see though modern day Greece broken up into more regions so your theory could become even clearer.
 
I would not trust much in the results of calculators at Gedmatch since they are made by amateur individuals with no samples approved by academics. Genealogy is a nice hobby for them to do in spare times.

I read many comments in internet forums complaining about their calculators. Dienekes and Davidski themselves confess that they have limited samples.
 
I would not trust much in the results of calculators at Gedmatch since they are made by amateur individuals with no samples approved by academics. Genealogy is a nice hobby for them to do in spare times.

I read many comments in internet forums complaining about their calculators. Dienekes and Davidski themselves confess that they have limited samples.

They are interesting to compare to other people, but they are very speculative and I they are sort of just guessing at what markers mean what and then seeing how the results compare to actually geography, which does make some sense, but Im not sure how much in reality it represents "ancient populations".
 
They are interesting to compare to other people, but they are very speculative and I they are sort of just guessing at what markers mean what and then seeing how the results compare to actually geography, which does make some sense, but Im not sure how much in reality it represents "ancient populations".

Dienekes, at least, lists all the sources for the samples he uses in his runs...He hasn't done one in a while, but when he creates a new one, it always includes all the publicly available academic samples.

One thing I don't know is where the Greek mainland samples were taken...people who submitted their data to him were encouraged to state their specific origins in a dedicated thread, but most people didn't do it.

From what I 've seen at 23andme, there are differences depending on geography. The best thing to do would be to try to locate other Greed testers and then compare results.
 
Thanks to everyone for their comments..


For the people who have some concerns regarding the validity of autosomal calculators, I would say that I do not worry at all about their validity as the methodology used (software, statistical analysis techniques, etc.) is exactly the same as the one used in proper research publications. In fact, these people (i.e. Dienekes, David, etc.) are not amateurs at all, but they are scientists themselves and I am pretty sure that the majority of them are working professionally on admixture analyses in their proper jobs, otherwise believe me they wouldn't have been able to construct genetic admixture calculators!


I agree with Angela in that the main limitation of these analyses is that it is not always known where exactly the participants originate from and this is particularly important for the Greek sample, which seems to be quite heterogenous. For example, I would not expect a Greek from Crete to be very similar genetically with a Greek from Drama (i.e. Northern Greece), yet they are all pooled together under 'Greek'. In fact, something that I have commented in other posts as well is that Cretans are surprisingly absent from the admixture calculators, while Sardinians, Sicilians and Cypriots are quite well represented. Another limitation I see is the small sample size for some populations (i.e., n=7 for the Lebanese), but nothing other than that. In general I think these calculators are pretty accurate, if the proper reference populations are used.


I would urge any Greek person who has used the Dodecad k12b calculator (or any other) to post their results here and mention their EXACT place of origin (not where they reside now, otherwise 50% of individuals would be labelled as 'Athenians'..). This way we will start getting a clearer picture of the geographical differences in admixture among Greeks.
 
I suppose... But is there any testing company that it's samples are approved by academics? Is this the case with 23andme? If not, then how can I diverge studies that are reliable from those that aren't?
 

This thread has been viewed 11998 times.

Back
Top