The genetic origin of Daunians

ORD010 … about 1078 - 1156 AD, around the time of the County of Apulia and Calabria:

“later the Duchy of Apulia and Calabria, was a Norman state founded by William of Hauteville in 1042 in the territories of Gargano, Capitanata, Apulia, Vulture, and most of Campania. It became a duchy when Robert Guiscard was raised to the rank of duke by Pope Nicholas II in 1059 …”

… “Bohemond of Taranto, son of Robert Guiscard, born in Calabria, in 1098 AD became Bohemond I of Antioch, he was the most experienced military leader of the First Crusade” ….

… speculating :) maybe ORD010 or one of the Parents originated in Antioch, …

k00nh8o.jpg


https://en.m.wikipedia.org/wiki/County_of_Apulia_and_Calabria

https://en.m.wikipedia.org/wiki/Bohemond_I_of_Antioch
 
Salento said:
ORD010 … about 1078 - 1156 AD, around the time of the County of Apulia and Calabria:

“later the Duchy of Apulia and Calabria, was a Norman state founded by William of Hauteville in 1042 in the territories of Gargano, Capitanata, Apulia, Vulture, and most of Campania. It became a duchy when Robert Guiscard was raised to the rank of duke by Pope Nicholas II in 1059 …”

… “Bohemond of Taranto, son of Robert Guiscard, born in Calabria, later became Bohemond I of Antioch, was the most experienced military leader of the First Crusade” ….

… speculating :) maybe ORD010 or one of the Parents originated in Antioch, …

k00nh8o.jpg


https://en.m.wikipedia.org/wiki/County_of_Apulia_and_Calabria

https://en.m.wikipedia.org/wiki/Bohemond_I_of_Antioch
Click to expand...

It is possible, there's a number different scenarios that could explain who he was. He could have had ancestry further from the east, from the more cosmopolitan era prior to the middle ages.

He could have had some Saracen/Moorish input.

The South underwent a series of ethnic cleansing in middle ages, where you saw the destruction of places like Lucera, the final Moorish stronghold. There was also persecution of Jewish people by Catholics, in the middle ages as well. That's what drove former Roman Jews into northern Europe.

Maybe some stayed behind in secret and were mixed out of existence, by the larger native Southern population.

The possibilities are endless.

At any rate, I don't think this person should have been used as a being representative of medieval Puglia in terms of autosomal proportion. The Apulian samples from the Viking paper are different from him as well, and they are from roughly the same era.
 
ihype02 said:
Since Pax Augusta mentioned it is true that David only uses academic samples.
Click to expand...

That doesn't matter, his calculators within and of itself were modified to produce results to control for a "calculator effect".

Here is an old post where Dienekes addresses this issue of "calculator effect", and in the comment section, has a debate with Davidski about it.

http://dodecad.blogspot.com/2012/08/on-so-called-calculator-effect.html

I myself have tried to make calculators in the past, with other collaborators. All of our attempts were unsuccessful in trying to replicate the results of academic studies. While trying to do that, I realized making a calculator itself is vulnerable to a high possibility of error, even when using the same exact samples used for studies.

Using samples to model population X has a high possibility of error.

Calculator X within and of itself could have something erroneous with it.

This is why I take these calculators with a grain of salt.
 
Jovialis said:
That doesn't matter, his calculators within and of itself were modified to produce results to control for a "calculator effect".

Here is an old post where Dienekes addresses this issue of "calculator effect", and in the comment section, has a debate with Davidski about it.

http://dodecad.blogspot.com/2012/08/on-so-called-calculator-effect.html

I myself have tried to make calculators in the past, with other collaborators. All of our attempts were unsuccessful in trying to replicate the results of academic studies. While trying to do that, I realized making a calculator itself is vulnerable to a high possibility of error, even when using the same exact samples used for studies.

Using samples to model population X has a high possibility of error.

Calculator X within and of itself could have something erroneous with it.

This is why I take these calculators with a grain of salt.
Click to expand...

I completely agree.

Fwiw, I don't know if he only uses academic samples today. Has he ever published his complete set? I absolutely know he didn't when he first created his calculators because he solicited samples from his followers. So did Dienekes, of course, but he showed you the clusters and posted his complete methodology for the sake of transparency. So far as I know Davidski has yet to do that.
 
I want to add that some of K12b samples are not perfect either. The Ligurian one is too Tuscan-like, probably an outlier and the Greek Macedonian one is from the assimilated Slavophone minority, they are genetically closer to Bulgarians than to Peloponnesians, which is not true for other Greek Macedonians. Davidski himself knows that some of his modern samples are not good either. But overall the calculator is not amazing but it's decent.
Distance to:Italian_Liguria
0.01994164Italian_Lombardy
0.02259627Italian_Piedmont
0.02374994Italian_Bergamo
0.02490093Italian_Trentino-Alto-Adige
0.02570667Italian_Veneto
0.02748159Swiss_Italian
0.02848505Italian_Tuscany
0.03041169Spanish_Menorca
0.03137781French_Corsica
0.03167126Italian_Northeast
0.03182874Spanish_Eivissa
0.03237525Greek_Thessaly
0.03267611Spanish_Mallorca
0.03379182Italian_Marche
0.03387672Spanish_Baleares
0.03449545French_Provence
0.03564830Spanish_Terres_de_l'Ebre
0.03616846Spanish_Murcia
0.03625818Albanian
0.03679837Spanish_Andalucia
0.03701475Italian_Umbria
0.03718578Spanish_Castilla_La_Mancha
0.03756287Greek_Macedonia
0.03762135Portuguese
0.03818783Spanish_Extremadura

 
ihype02 said:
I want to add that some of K12b samples are not perfect either. The Ligurian one is too Tuscan-like, probably an outlier and the Greek Macedonian one is from the assimilated Slavophone minority, they are genetically closer to Bulgarians than to Peloponnesians, which is not true for other Greek Macedonians. Davidski himself knows that some of his modern samples are not good either. But overall the calculator is not amazing but it's decent.
Distance to:Italian_Liguria
0.01994164Italian_Lombardy
0.02259627Italian_Piedmont
0.02374994Italian_Bergamo
0.02490093Italian_Trentino-Alto-Adige
0.02570667Italian_Veneto
0.02748159Swiss_Italian
0.02848505Italian_Tuscany
0.03041169Spanish_Menorca
0.03137781French_Corsica
0.03167126Italian_Northeast
0.03182874Spanish_Eivissa
0.03237525Greek_Thessaly
0.03267611Spanish_Mallorca
0.03379182Italian_Marche
0.03387672Spanish_Baleares
0.03449545French_Provence
0.03564830Spanish_Terres_de_l'Ebre
0.03616846Spanish_Murcia
0.03625818Albanian
0.03679837Spanish_Andalucia
0.03701475Italian_Umbria
0.03718578Spanish_Castilla_La_Mancha
0.03756287Greek_Macedonia
0.03762135Portuguese
0.03818783Spanish_Extremadura
Click to expand...
These samples were not made by Dienekes. They are very recent additions from apricity members. We aren't sure where they come from.
 
Jovialis said:
These samples were not made by Dienekes. They are very recent additions from apricity members. We aren't sure where they come from.
Click to expand...

I agree with that. Amazing how people feel free to comment when they're not in possession of the facts

As for the Ligurian sample mentioned above, the only way anyone could know if it's an "outlier" and "not representative" is if it could be compared to a large sample of Ligurians, preferably elderly people, all of whose grandparents were from Liguria. That's because migration to Liguria began in the late 19th century. Also, it would be important to gather samples from both western and eastern Liguria, because despite being such a small region, the history and genetic influences were quite different depending on the area.

It would also be useful to compare that one lone sample added by apricity members to the so-called "Piemonte" samples, who are really mountain Ligures speaking a Ligurian dialect who were only very, very recently made a part of Piemonte.

That's unless, of course, someone has a crystal ball, and I hear they're in short supply. :)
 
ihype02 said:
I want to add that some of K12b samples are not perfect either. The Ligurian one is too Tuscan-like, probably an outlier and the Greek Macedonian one is from the assimilated Slavophone minority, they are genetically closer to Bulgarians than to Peloponnesians, which is not true for other Greek Macedonians. Davidski himself knows that some of his modern samples are not good either. But overall the calculator is not amazing but it's decent.
Distance to:Italian_Liguria
0.01994164Italian_Lombardy
0.02259627Italian_Piedmont
0.02374994Italian_Bergamo
0.02490093Italian_Trentino-Alto-Adige
0.02570667Italian_Veneto
0.02748159Swiss_Italian
0.02848505Italian_Tuscany
0.03041169Spanish_Menorca
0.03137781French_Corsica
0.03167126Italian_Northeast
0.03182874Spanish_Eivissa
0.03237525Greek_Thessaly
0.03267611Spanish_Mallorca
0.03379182Italian_Marche
0.03387672Spanish_Baleares
0.03449545French_Provence
0.03564830Spanish_Terres_de_l'Ebre
0.03616846Spanish_Murcia
0.03625818Albanian
0.03679837Spanish_Andalucia
0.03701475Italian_Umbria
0.03718578Spanish_Castilla_La_Mancha
0.03756287Greek_Macedonia
0.03762135Portuguese
0.03818783Spanish_Extremadura
Click to expand...

The one you posted is the G25 and the Ligurian average on G25 is based only on 1 individual (ALP099), as anyone can check in the datasheet. Apart from the fact that the K12b average is accurate or not, the Ligurian average in G25 is really unlikely to be accurate, it is not even an average but a single individual (unclear from what area of Liguria, AFAIK the only sampled area of Liguria in the previous studies is the Savona area, which linguistically contains Western Ligurian areas, Genoese areas, and transitional area between Ligurian and Piedmontese).

The only Ligurian individual on the G25 datasheet

Italian_Liguria:ALP099,0.113823,0.146236,0.027153,-0.011951,0.029236,-0.002231,0.00235,0.001154,0.009613,0.025331,-0.009743,0.002847,-0.013082,0.001101,-0.001086,-0.012463,-0.00665,-0.00038,0.003771,0.00075,-0.00025,0.000989,0.003944,0.000723,0.000599

The Ligurian average on G25

Italian_Liguria,0.113823,0.146236,0.027153,-0.011951,0.029236,-0.002231,0.00235,0.001154,0.009613,0.025331,-0.009743,0.002847,-0.013082,0.001101,-0.001086,-0.012463,-0.00665,-0.00038,0.003771,0.00075,-0.00025,0.000989,0.003944,0.000723,0.000599

As you can check for yourself, there is no difference.

So, the Ligurian Dodecad K12b average is not accurate? Maybe yes, but you certainly can't use the Ligurian G25 average to prove it as it is definitely not even an average. But Davidski is not to blame, it is some Italian geneticist who plays games.
 
Pax Augusta said:
The one you posted is the G25 and the Ligurian average on G25 is based only on 1 individual (ALP099), as anyone can check in the datasheet. Apart from the fact that the K12b average is accurate or not, the Ligurian average in G25 is really unlikely to be accurate, it is not even an average but a single individual (unclear from what area of Liguria, AFAIK the only sampled area of Liguria in the previous studies is the Savona area, which linguistically contains Western Ligurian areas, Genoese areas, and transitional area between Ligurian and Piedmontese).

The only Ligurian individual on the G25 datasheet

Italian_Liguria:ALP099,0.113823,0.146236,0.027153,-0.011951,0.029236,-0.002231,0.00235,0.001154,0.009613,0.025331,-0.009743,0.002847,-0.013082,0.001101,-0.001086,-0.012463,-0.00665,-0.00038,0.003771,0.00075,-0.00025,0.000989,0.003944,0.000723,0.000599

The Ligurian average on G25

Italian_Liguria,0.113823,0.146236,0.027153,-0.011951,0.029236,-0.002231,0.00235,0.001154,0.009613,0.025331,-0.009743,0.002847,-0.013082,0.001101,-0.001086,-0.012463,-0.00665,-0.00038,0.003771,0.00075,-0.00025,0.000989,0.003944,0.000723,0.000599

As you can check for yourself, there is no difference.

So, the Ligurian Dodecad K12b average is not accurate? Maybe yes, but you certainly can't use the Ligurian G25 average to prove it as it is definitely not even an average. But Davidski is not to blame, it is some Italian geneticist who plays games.
Click to expand...

Ligurians in an Academic PCA come up as, strictly, Northern Italian not Central Italian. The paper was probably discussed here, even though I cannot find it.

Also the Ethnic Macedonian sample were very Greek-shifted compared to Bulgarians, some that I have seen. Davidski has updated the (ethnic) Macedonian samples and they are Romanian/Bulgarian-like now. In Academic PCA ethnic Macedonians are pretty much like Bulgarians but a little more South-western shifted.

There is one Macedonian t.roll saying that South Slavs are 80% Ballkanic, I would not be surprised if he twisted his own samples that he gave.
 
ihype02 said:
Ligurians in an Academic PCA come up as, strictly, Northern Italian not Central Italian. The paper was probably discussed here, even though I cannot find it.
Click to expand...


Ligurians do not come out as Central Italian in K12b, regardless of whether it is really accurate as an average. The most comprehensive paper on Italy so far, from which that sample used on G25 came from, is Raveane 2019. The PCAs have been posted several times.

The Northern Italian is the biggest Italian cluster.

In the Raveane 2019 paper in green are Ligurians and Emilians (NItaly3) with a minority of Piedmontese and Venetians, in pink and red are Tuscans (NCItaly 1, 2, 3). Mainland Italy this time was divided in two parts, not into three clusters as in previous studies. Ligurians came out together with Emilians as the southernmost part of the Northern Italian cluster, while Tuscans ended up just behind Ligurians and Emilians in a cluster called Northern-Central Italian grouped with the Northern Italian clusters.

A sharp north-south division in cluster distribution was detected, the separation between northern and southern areas being shifted north along the peninsula (Fig. 1B) (12). The reported structure dismissed the possibility that the Central Italian populations differentiated from the Northern and Southern Italian groups (Fig. 1A) (13). Individuals from Central Italy were, in fact, assigned mostly to the Southern Italian clusters, except for samples from Tuscany, which grouped instead with the Northern Italian clusters (Fig. 1, A and B) (12).
Click to expand...

Do you need me to show you better?

aaw3492-f1.jpeg


ihype02 said:
Also the Ethnic Macedonian sample were very Greek-shifted compared to Bulgarians, some that I have seen. Davidski has updated the (ethnic) Macedonian samples and they are Romanian/Bulgarian-like now. In Academic PCA ethnic Macedonians are pretty much like Bulgarians but a little more South-western shifted.

There is one Macedonian t.roll saying that South Slavs are 80% Ballkanic, I would not be surprised if he twisted his own samples that he gave.
Click to expand...


Which calculator are we talking about now? The K12b or the G25? Likely the updated K12b averages may have been made by several different users. The G25 updated averages are based on available academic samples. But not all available academic samples are accurate and exhaustive.
 
Pax Augusta said:
Which calculator are we talking about now? The K12b or the G25? Likely the updated K12b averages may have been made by several different users. The G25 updated averages are based on available academic samples. But not all available academic samples are accurate and exhaustive.
Click to expand...

The G25 ones, but they are different now though.
 
Re: Jovialis Models

New here, but these models are interesting. From what I read of the Daunian paper it does seem that ancient Daunian samples carry more WHG-related and Anatolia_N (likely EEF-related ancestry) so some of these results indicate that imo.


But you say that the fits are bad, do you mean those with distance >2.0? Also, would these be scaled or unscaled coordinates?


I think it's worth adding Anatolia_BA to at least check if any of these samples have Anatolian-related ancestry distinct from Minoan or if bad fits are more of a consequence of low coverage.
 
SandbagsIA said:
New here, but these models are interesting. From what I read of the Daunian paper it does seem that ancient Daunian samples carry more WHG-related and Anatolia_N (likely EEF-related ancestry) so some of these results indicate that imo. But you say that the fits are bad, do you mean those with distance >2.0? Also, would these be scaled or unscaled coordinates? I think it's worth adding Anatolia_BA to at least check if any of these samples have Anatolian-related ancestry distinct from Minoan or if bad fits are more of a consequence of low coverage.
Click to expand...
Here you go:https://www.eupedia.com/forum/threa...Italians/page5?p=638321&viewfull=1#post638321
 
Jovialis said:
Here you go
Click to expand...

Thank you, these results make sense to me, and I suspect this Anatolia_BA may reflect some antiquity (Hellenistic/Roman) events of Anatolia-South Europe interactions (possibly due to the Greek colonies)

Btw, which population are you using as Greek? Is it one of of the Global_25 presets like Greek_Macedonia or Greek_Peloponnese or a composite?
 
SandbagsIA said:
Thank you, these results make sense to me, and I suspect this Anatolia_BA may reflect some antiquity (Hellenistic/Roman) events of Anatolia-South Europe interactions (possibly due to the Greek colonies)Btw, which population are you using as Greek? Is it one of of the Global_25 presets like Greek_Macedonia or Greek_Peloponnese or a composite?
Click to expand...
The Greek sample comes from Dodecad. But I believe it is from Central Greece.
 
Jovialis said:
The Greek sample comes from Dodecad. But I believe it is from Central Greece.
Click to expand...

Then the results are even more reasonable imo, though if it is the samples I think of they are likely Central Greek islanders (If it's the Central_Greek labeled-ones from the Dodecad project) rather than Central Greeks from mainland Greece. They should be close to Peloponnesians but likely have some more Anatolia_BA-ancestry.
 
The paper is officially out. Link provided by Lazaridis; many thanks as always.

https://academic.oup.com/mbe/advance-article/doi/10.1093/molbev/msac014/6509524?searchresult=1

I've already said it, but I do wish academics would stop talking about the genetic history of "The Italians". Just look at the space we occupy on the PCA. Everybody else in Europe starts looking like the Han by comparison.

As I probably also said, the Daunians look to me like western shifted Central Italians and Tuscans.
 
Angela said:
The paper is officially out. Link provided by Lazaridis; many thanks as always.

https://academic.oup.com/mbe/advance-article/doi/10.1093/molbev/msac014/6509524?searchresult=1

I've already said it, but I do wish academics would stop talking about the genetic history of "The Italians". Just look at the space we occupy on the PCA. Everybody else in Europe starts looking like the Han by comparison.

As I probably also said, the Daunians look to me like western shifted Central Italians and Tuscans.
Click to expand...

Yes, but that J2b2-L283 was probably brought by the Illyrian part of Daunians. It's interesting that they give an indirect hint that the J2b2-L283 might have been potentially richer in CHG.

Another signal coming from qpAdm analyses is the apparent excess of CHG ancestry in IAA; however, the predominant contribution of CHG to the Steppe-related ancestry that, by the Iron Age, had already spread to the Mediterranean area makes it hard to properly detect a CHG signature independent from the Steppe wave, possibly brought by pan-Mediterranean influxes. When directly investigated with an f4 framework, IAA shows generally more CHG than Mycenaean, less CHG than contemporary Croatian_EIA and, in some cases (ORD019, SGR002 and the Mediaeval SGR001 with Z-scores higher than 2) more CHG than older Croatian samples (_N=Neolithic and _MN=Middle Neolithic) (Supplementary fig. 8).
Click to expand...
 
I think the authors are pretty circumspect about all of this.

"It is not clear whether these connections indicate a movement of people or a sharing ofcultural ideas and a conclusive answer to the origin of the Daunians remains elusive. From aparsimony perspective, the genetic results point to an autochthonous origin (e.g. a geneticcontinuity of Daunians with the population that inhabited the area prior to the examined historicalperiod), here mainly marked by the presence of WHG signature, although we cannot excludeadditional influences from Croatia (ancient Illyria), as described by available historical sourcesand by the material remains (De Juliis 1988; Norman 2016)."

In another part of the paper they say that some contribution from the Balkans is plausible.

". Three of them, which clustered close to modern Italians in the PCA (ORD001, ORD014 and SGR003, Fig. 1C), show higher affinity with the Iron Age Croatian sample (ORD004 followed this pattern too, but with lower f3 values). However, the remaining majority are closest to the Roman Republicans, which can be interpreted as representative of local Iron Age peninsular Italy ancestry, as also indicated by our MDS results."

I think it's interesting, going over the text of this paper once again, to think of the work Jovialis has done with the Balkan samples, as well as the Roman Republican samples.
 
You must log in or register to reply here.

This thread has been viewed 147206 times.

Back
Top