Appearance of Sicilians relative to other Mediterraneans? (poll included)

[Knovas]

oreo cookie: But doesn't appearance correlate more with autosomal DNA?

Correct. Too sad this seems to be a nightmare for some individuals.

And acording to the autosomal data, althought Sicilians and Southern Italians are very close to Greeks, they must be a bit darker in average.

Haplogroups have nothing to do with phisical appearence. Tell those Italians tested in haplogroup T that they are Ethiopians, and they'll be laughing all day. It's that simple.

 

[oreo_cookie]

oreo cookie: But doesn't appearance correlate more with autosomal DNA?

Correct. Too sad this seems to be a nightmare for some individuals.

And acording to the autosomal data, althought Sicilians and Southern Italians are very close to Greeks, they must be a bit darker in average.

Haplogroups have nothing to do with phisical appearence. Tell those Italians tested in haplogroup T that they are Ethiopians, and they'll be laughing all day. It's that simple.

How would they be darker? I figured maybe Slavic influence could lighten Greeks but the Normans were in southern Italy too so I thought it'd balance out. The thing I notice is, lighter Greeks often look Slavic, while lighter Sicilians often look Germanic.

 

[Knovas]

You are right about the slavic influence. Southern Italians and Sicilians have a little bit more West European, but Greeks have much more East European. And also, more European autosomal DNA in average. Here you have from Dodecad:

Southern Italians/Sicilians: WE (14.9) + EE (3) + Medit. (44.6) = 62.5

Greeks: WE (13.5) + EE (11) + Medit. (41.9) = 66.4

The Norman influence in Southern Italy was quite insignificant. Not relevant, except for the higher presence of some haplogroups (I1 or Q), but not apreciable on autosomal results.

 

[oreo_cookie]

Doesn't Greece have more West Asian as well but southern Italians and Sicilians have more North African and SW Asian/Arabian?

 

[Knovas]

The main reason is that Southern Italians have less Northern European and less total European score. But checking the non European results, it's apreciable that Southern Italians have also more African (sub-saharan and not) and more Southwest Asian too.

I'll try to vote. My account doesn't work properly as I pointed.

 

[oreo_cookie]

The main reason is that Southern Italians have less Northern European and less total European score. But checking the non European results, it's apreciable that Southern Italians have also more African (sub-saharan and not) and more Southwest Asian too.

It does surprise me that the Normans didn't have more of an autosomal affect.. but I want to know about where parts of Greece cluster on autosomal plots.. do Cretans, who are 40%+ J2, cluster differently than northern Greeks who have much higher frequencies of R1a and I2? I'd guess the (North)Eastern European components are much higher in these Greeks than in Peloponnesian, Cretan, and islander Greeks.

 

[Wilhelm]

It does surprise me that the Normans didn't have more of an autosomal affect.. but I want to know about where parts of Greece cluster on autosomal plots.. do Cretans, who are 40%+ J2, cluster differently than northern Greeks who have much higher frequencies of R1a and I2? I'd guess the (North)Eastern European components are much higher in these Greeks than in Peloponnesian, Cretan, and islander Greeks.

In this autosomal plot there are greeks from all parts (dark blue dots) they float around between southern-italians and pulling towards balkans :

 

[oreo_cookie]

Some of the Greeks appear to be heading toward the Near East while others toward Romania, with most in the middle, near the Italians. ^

 

[Dorianfinder]

Where is the Greek sample from? Are those Cypriots Turkish or Greek-speaking Cypriots?

Looks like they got their sample in the Athens Mall. We need more rural sampling in Greece.

According to the scatterplot the Greeks are not related to the Balkan population.:useless:

 

[oreo_cookie]

Some of the Greeks are deviating in the Balkan direction though.

 

[Dorianfinder]

Some of the Greeks are deviating in the Balkan direction though.

Deviating is the correct term however it appears that all the Greeks are deviants, makes me wonder where their median lies. Looks like it could be just above the Italians from Central Italy?

 

[Knovas]

It's curious. I supose that they also used Greeks from distant areas, not only the main Greeks. If you see the Romanians, it happens the same, since they surely used normal Romanians and others with substantial gipsy ancestry.

For the moment, the autosomal Greek sample is very homogeneous checking Dodecad, so it's clear that they are typical Greeks. Checking Romanian individual results (downloading the population portraits), it's easy to see some individuals who are very different from the main population, matching a gipsy profile. So probably as more samples come, different types of Greeks will join the project and some differences will appear.

 

[Dorianfinder]

Here are the figures for antosomal DNA within various populations represented graphically http://4.bp.blogspot.com/_Ish7688voT0/TShZrL1hQLI/AAAAAAAADKM/f2VuiF4Fm0E/s1600/ADMIXTURE_10.png

 

[Dorianfinder]

it's easy to see some individuals who are very different from the main population, matching a gipsy profile. So probably as more samples come, different types of Greeks will join the project and some differences will appear.

Many individuals are willing to participate because they have an unresolved background or look different to others in their community. Results are scewed by these participants. Dienekes even mentions that he decided to exclude two Gypsy Roma and two Armenian outliers.

 

[Knovas]

But this is not rare since the objective is to get realistic aproximations to concrete ethnic groups. He probably had serious reasons to do it, because there are specific groups for Armenians and Gipsies (quite for them).

One of the reasons to exclude some people are plain lies coming from the participants, who clearly didn't present their real ancestry just to get in to the project, and when the results appear, Dienekes' noticed the fraud inmediatly. This is very usual in some South Americans who pretend to be Spanish for example. But sometimes, when there was no fraud, he decided to accept the sample even if the person was of mixed ancestry but, of course, he had to exclude quite of those participants in some runs.

Probably as more samples come from distant places of Greece, and other countries, he will think in the creation of a group for them, like others that already exist in the project for minorities.

 

[Dorianfinder]

But this is not rare ...
One of the reasons to exclude some people are plain lies coming from the participants, who clearly didn't present their real ancestry just to get in to the project

My point exactly, the problem with sampling is that you are relying on the participant's word or belief. Neither are very scientific!

 

[Dorianfinder]

Often what participants express is their desire not what they know to be true. Studies show that heterogeneous populations have a higher sensitivity to racial or cultural stereotypes meaning that they are more likely to lie about their ethnic background.

 

[Knovas]

Don't underestimate Dieneke's, he has powerful tools to detect fraudsters. I'm pretty sure none of them got into to the project, even Dienekes created this post to warn about this: http://dodecad.blogspot.com/2011/02/truly-despicable-behavior.html

And some of the population samples have quite individuals, so even if there are one or two fraudsters in them, the final average report does not change drastically. But as I said, I highly doubt you can find fraudsters listed in the populations. What he calls outliers is another thing, just a few people with some special characteristics, wich make those individuals different from the main population asociated. Nothing else.

 

[Cambrius (The Red)]

You find people misrepresenting their backgrounds even in highly structured genetic research projects run by well-known academic institutions. Drastic individual deviations from any indigenous population's average scores are normally discarded.

Anyway, professional autosomal clustering research essentially confirms Dodecad (Dienekes) and Eurogene scores.

 

[Wilhelm]

Actually the two Romanian gypsies were part of the sample of the professional academic study of Behar et al. , but Dienekes already excludes the two of them when averaging the Romanian sample. Also, there are part Canarian people who want to make part of the spanish sample. They are going to screw up the results, obviously. But the Greek sample is 100% greek, Dienekes is greek, he surely takes care of the sample. He has tools to detect frauds, such as the Clusters Galore.