How to group Italian samples in order to obtain good references?

Hi,

NE Italy can be easily done by grouping Italian_Veneto + Italian_Trentino-Alto-Adige + Italian_Northeast (Triveneto region).

For C. Italy, I'd go for Italian_Lazio + Italian_Tuscany + Italian_Umbria + Italian_Marche.
(Abruzzo is "ethnically" Southern Italian)

S. Italy: Italian_Abruzzo + Italian_Apulia + Italian_Basilicata + Italian_Calabria* + Italian_Campania + Italian_Molise* + Sicilian_East
I didn't put Sicilian_West bc they're abnormally plotting with Central Italians (lack of samples?).

*Some here mentioned Calabrese avg isn't well represented because there are three samples only, but honestly I've being using it and IMO I don't think they're too outlying, there are some Campanian individual samples very similar to them. If you think it's better to remove the Italian_Calabria avg, consider also removing Italian_Molise, since it's composed of only two samples and it's quite central-italian shifted, otherwise the "northern" part of South Italy would be overrepresented. To summarize, I'd put both Calabria and Molise or remove both, but I'd certanly remove Sicilian_West.

NW Italy is a problem. Swiss_Italian is acting almost as a Tuscan avg for some reason and Italian_Piedmont is pretty shaky.
I've read people saying Italian_Piedmont samples are from Val Borbera, which is "ethnically" Ligurian, I'm not sure if this is the case, could someone here confirm? :confused:
Anyways, a NW Italy sample would have a huge hole because it's missing the piece that would practically be its protagonist: a consistent Italian_Piedmont avg.
I'm not sure how I would group them together :icon_ask:

Quote:

---

Originally Posted by Ethel

Hi,
*Some here mentioned Calabrese avg isn't well represented because there are three samples only, but honestly I've being using it and IMO I don't think they're too outlying, there are some Campanian individual samples very similar to them. If you think it's better to remove the Italian_Calabria avg, consider also removing Italian_Molise, since it's composed of only two samples and it's quite central-italian shifted, otherwise the "northern" part of South Italy would be overrepresented. To summarize, I'd put both Calabria and Molise or remove both, but I'd certanly remove Sicilian_West.

---

They are very outlying, I've seen probably dozens of Calabrian results on Gedmatch, the vast majority of them coming from Reggio Calabria/Vibo Valentia, and these G25 samples are even more outlying than what your average southern calabrian would be.

Sicilian West has 3-4% WHG admix for some reason. Probably recent Norman admixture?

You should probably just use this.

Italian_South samples are from Naples (ITS2 ; ITS4), Belvedere Marittimo (BEL57) and Crispiano (ITS7) :

Code:

---

Italian_South_BEL57,0.108132,0.150298,-0.009051,-0.04199,0.009232,-0.003347,-0.004935,-0.002769,0.002659,0.019864,0.001949,0.003447,-0.002081,0.003441,-0.009229,-0.012198,0.005346,0.003547,0.007416,-5e-04,-0.002496,-0.006925,0.00456,0.010724,0.002036
Italian_South_ITS2,0.113823,0.137096,-0.004525,-0.038114,0.00277,-0.013945,0.00282,-0.002538,-0.001841,0.016948,-0.004384,0.006894,-0.005798,-0.006055,-0.007872,0.002387,-0.000782,0.002914,0.006159,-0.001876,-0.004742,0.000866,-0.001725,-0.007109,0.002754
Italian_South_ITS4,0.100164,0.153345,-0.010182,-0.039406,0.010463,-0.015339,0.003995,-0.004384,0.001227,0.023873,-0.003085,0.006744,0.000595,-0.000688,-0.004343,-0.002784,0.001043,-0.004434,0.005154,0.006878,-0.00287,0.003833,0.003451,-0.000361,-0.000958
Italian_South_ITS7,0.10927,0.14319,0.004525,-0.036499,0.02462,-0.013666,-0.002585,-0.002077,0.003681,0.016766,-0.002436,0.003447,-0.007582,-0.003991,-0.002443,-0.002519,-0.00352,0.00228,0.00176,-0.005878,-0.003743,0.005688,0.001602,-0.000482,0.002515

---

Code:

---

Italian_South,0.1078472,0.14598220018232,-0.0059717,-0.0037785,0.0005218,0.0010768,0.0051222,-0.000344,-0.0034628,0.0008655,0.001972,0.000693,0.0015868

---

Total Euro score (WHG+Steppe+farmer) for pops:

Apulia: 80
Campania :78,4
Lazio:85
Molise: 84,2
Calabria: 77,8
Basilicata: 80,4
Abruzzo 82,4
Italian_South: 80.

Quote:

---

Originally Posted by Deusex99

Sicilian West has 3-4% WHG admix for some reason. Probably recent Norman admixture?

You should probably just use this.

Italian_South samples are from Naples (ITS2 ; ITS4), Belvedere Marittimo (BEL57) and Crispiano (ITS7) :

Code:

---

Italian_South_BEL57,0.108132,0.150298,-0.009051,-0.04199,0.009232,-0.003347,-0.004935,-0.002769,0.002659,0.019864,0.001949,0.003447,-0.002081,0.003441,-0.009229,-0.012198,0.005346,0.003547,0.007416,-5e-04,-0.002496,-0.006925,0.00456,0.010724,0.002036
Italian_South_ITS2,0.113823,0.137096,-0.004525,-0.038114,0.00277,-0.013945,0.00282,-0.002538,-0.001841,0.016948,-0.004384,0.006894,-0.005798,-0.006055,-0.007872,0.002387,-0.000782,0.002914,0.006159,-0.001876,-0.004742,0.000866,-0.001725,-0.007109,0.002754
Italian_South_ITS4,0.100164,0.153345,-0.010182,-0.039406,0.010463,-0.015339,0.003995,-0.004384,0.001227,0.023873,-0.003085,0.006744,0.000595,-0.000688,-0.004343,-0.002784,0.001043,-0.004434,0.005154,0.006878,-0.00287,0.003833,0.003451,-0.000361,-0.000958
Italian_South_ITS7,0.10927,0.14319,0.004525,-0.036499,0.02462,-0.013666,-0.002585,-0.002077,0.003681,0.016766,-0.002436,0.003447,-0.007582,-0.003991,-0.002443,-0.002519,-0.00352,0.00228,0.00176,-0.005878,-0.003743,0.005688,0.001602,-0.000482,0.002515

---

Code:

---

Italian_South,0.1078472,0.14598220018232,-0.0059717,-0.0037785,0.0005218,0.0010768,0.0051222,-0.000344,-0.0034628,0.0008655,0.001972,0.000693,0.0015868

---

---

How can we consider representative a sample for all Sout Italy with only 4 individual averages?

The only problem that could be is between NorthEast and NorthWest because they are ambiguous zones.

And you posted wrongly coordinates of your South sample.

https://i.imgur.com/7s5hB5N.png

Edit: I´ve posted the results of Ancien Components Model of Celtibero for the references you posted ancient composition.

Quote:

---

Originally Posted by Samnium

They are very outlying, I've seen probably dozens of Calabrian results on Gedmatch, the vast majority of them coming from Reggio Calabria/Vibo Valentia, and these G25 samples are even more outlying than what your average southern calabrian would be.

---

Gedmatch results and G25 not always show exactly the same, better compare gedmatch results vs gedmatch results and G25 ves G25 ones.

Quote:

---

Originally Posted by gixajo

How can we consider representative a sample for all Sout Italy with only 4 individual averages?

The only problem that could be is between NorthEast and NorthWest because they are ambiguous zones.

And you posted wrongly coordinates of your South sample.

https://i.imgur.com/7s5hB5N.png

Edit: I´ve posted the results of Ancien Components Model of Celtibero for the references you posted ancient composition.

---

OK that's good for me.

Quote:

---

Originally Posted by gixajo

How can we consider representative a sample for all Sout Italy with only 4 individual averages?

The only problem that could be is between NorthEast and NorthWest because they are ambiguous zones.

And you posted wrongly coordinates of your South sample.

https://i.imgur.com/7s5hB5N.png

Edit: I´ve posted the results of Ancien Components Model of Celtibero for the references you posted ancient composition.

---

OK that's good for me.

Anyway I've checked the list of Campanians from Raveane et al on G25 and 4 of them have like significant recent north African admixture. It's obvious that they are not natives since none of the other S. Italians or Sicilians have any of that. It looks like this mess of study have lumped for some reason a lot of mixed outliers inside the Campania cluster. A total shitshow.

Notice that the 8 samples from 23andme are similar to Apulians in their ancestry proportions (total Euro 79,8 vs 80).

For comparison 8 Campanians from 23andme.

https://i.ibb.co/DLW6BK6/Screenshot-20200711-012413.jpg

Total Euro score: 79,3

Apulia and Basilicata.

https://i.ibb.co/jkrRGZY/Screenshot-20200711-012524.jpg

Total Euro score: 80,2

It doesn't really surprise me. After all, nowhere in Raveane et al they claim that they have checked the ancestry of their samples.

Double post.