The genetic origin of Daunians

[Salento]

^^hmm, Maybe someone can double-check my work here, but ORD001 doesn't seem anywhere near Mycenaeans. Maybe due to the low-coverage?

Using 1 populations approximation
1 100% Romanians @ 11.535
2 100% Bulgarian_D @ 12.448
3 100% Bulgarians_Y @ 12.601
4 100% O_Italian_D @ 14.207
5 100% N_Italian_D @ 15.513
6 100% TSI30 @ 17.934
7 100% North_Italian @ 18.341
8 100% Tuscan @ 18.947
9 100% Hungarians @ 19.195
10 100% C_Italian_D @ 19.545

I get the same results.

... as you know WGSE3 produces slightly different RawData, and Admixture Studio and GedMatch often produce slightly different coordinates.

... from the Dodecad Ancestry Project:

... below are the WGSE2 and WGS3 coordinates produced with the Original Dodecad K... calculator

Dodecad Ancestry Project: 'K12b' and 'K7b' calculators

Orig_Dodecad_K12_WGSE2_ORD001,0.65,2.21,2.84,0.00,27.53,34.98,0.00,1.37,9.83,0.43,20.16,0.00
Orig_Dodecad_K12_WGSE3_ORD001,2.73,2.34,1.89,0.01,30.09,33.94,0.00,2.16,7.76,0.55,18.54,0.00

… as long as we use the same standard (WGSE2 sometimes ….3 and Admixture-Studio) we’re OK

 

[Salento]

 

[Hawk]

Do we have any Y-DNA breakdown of modern Apuglians?

 

[kingjohn]

nice salento
ORD010 he looks western jewish or greek islander profile
maybe he had recent byzantine ancestery
or muslim ancestery

Admix Results (sorted):

#	Population	Percent
1	East_Med	36.04
2	West_Med	20.38
3	North_Atlantic	15.09
4	West_Asian	12.20
5	Red_Sea	9.29
6	Baltic	5.83

Finished reading population data. 204 populations found.
13 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Sephardic_Jewish @ 4.557816
2 Italian_Jewish @ 5.256803
3 Algerian_Jewish @ 5.670169
4 Ashkenazi @ 6.580314
5 South_Italian @ 7.225772
6 East_Sicilian @ 8.048594
7 Tunisian_Jewish @ 8.227295
8 Libyan_Jewish @ 9.225803
9 Central_Greek @ 10.025041
10 West_Sicilian @ 11.006186
11 Italian_Abruzzo @ 12.665685
12 Cyprian @ 13.848021
13 Greek_Thessaly @ 16.500519
14 Lebanese_Muslim @ 17.823469
15 Tuscan @ 18.530861
16 Syrian @ 18.570854
17 Samaritan @ 19.522041
18 Palestinian @ 20.010542
19 Jordanian @ 20.402901
20 Tunisian @ 20.948820

Using 2 populations approximation:
1 50% Samaritan +50% Tuscan @ 2.571835Using 3 populations approximation:
1 50% Samaritan +25% Tuscan +25% Tuscan @ 2.571835


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 Ashkenazi + Samaritan + Sephardic_Jewish + Tuscan @ 2.227096

 

[ihype02]

ORD009 in G25. Daunians are like Western Shifted Tuscans, so Corsica should be the the first modern group they show greatest affinity to.

Distance to:	Italy_IA_Messapic_Daunian_Ordona:ORD009
0.03375273	French_Corsica
0.03402973	Italian_Veneto
0.03500182	Italian_Piedmont
0.03527467	Italian_Tuscany
0.03556619	Swiss_Italian
0.03639077	Italian_Bergamo
0.03884339	Italian_Trentino-Alto-Adige
0.04026252	Italian_Umbria
0.04182701	Spanish_Baleares
0.04239967	Italian_Northeast
0.04291958	Spanish_Mallorca
0.04303520	Italian_Lombardy
0.04311957	Italian_Aosta_Valley
0.04322355	Spanish_Menorca
0.04353634	Italian_Marche
0.04476102	Spanish_Eivissa
0.04508314	Albanian
0.04542251	Greek_Thessaly
0.04542500	Greek_Central_Macedonia
0.04567742	French_Provence
0.04579698	Spanish_Peri-Barcelona
0.04627010	Italian_Lazio
0.04644970	Spanish_Girona
0.04647612	Italian_Liguria
0.04702300	Spanish_Murcia

Distance to:	Italy_IA_Messapic_Daunian_Ordona:ORD009
0.02874863	HRV_MBA
0.03606988	HRV_Pop_CA
0.03979688	ITA_Rome_MA
0.04035270	DEU_MA_Alemannic_o2
0.04069774	ITA_Broion_BA
0.04160810	ITA_Etruscan
0.04203421	Bell_Beaker_ITA
0.04261694	DEU_Roman
0.04282862	ITA_Rome_Late_Antiquity
0.04285214	ITA_Villanovan
0.04285506	Iberia_Northeast_c.8-12CE
0.04459502	CHE_IA
0.04563672	BGR_EBA
0.04626143	HRV_EBA
0.04762398	HUN_BA
0.04798769	UKR_Cimmerian_o
0.04826724	ITA_Tivoli_Renaissance
0.04907844	ITA_Ardea_Latini_IA
0.04960111	ITA_ReginaMargherita_BA
0.04964653	ITA_Rome_Latini_IA
0.05036041	HUN_MBA_Vatya
0.05122268	HRV_IA
0.05124357	ITA_Etruria_Imperial
0.05186858	VK2020_ITA_Foggia_MA
0.05274986	GRC_Helladic_MBA
0.05344087	ITA_Prenestini_tribe_IA_o
0.05383536	DEU_Lech_MBA
0.05404840	Scythian_MDA
0.05410692	BGR_IA
0.05488058	ITA_Proto-Villanovan
0.05633730	DEU_MA_Alemannic_o1
0.05773184	ITA_Sardinia_IA
0.05902228	ITA_Prenestini_tribe_IA
0.05908047	DEU_Lech_EBA_contam
0.05943041	SRB_Mokrin_EBA
0.05944251	Levant_LBN_MA_o4
0.06017542	TUR_IA_low_res
0.06051290	ITA_Boville_Ernica_IA
0.06057405	TZA_Zanzibar_Euro_outlier
0.06072095	Bell_Beaker_HUN

This one was a bit northern let's wait and see for Messapian genomes to be revealed. I think Iron Age Campania will look like this though.

Distance to:	Italian_Apulia
0.04766329	Iberia_Northeast_Empuries2
0.04995370	GRC_Mycenaean
0.05850706	Italy_IA_Messapic_Daunian_Ordona:ORD009

 

[Riverman]

On a general PCA, Ord009 is also closer to the J2b cluster from the British paper, and one of the more Northern-Western shifted samples. The others tend to have more Neolithic ancestry, which pulls them towards BGR_IA/EBA in comparison. The connection between Proto-Villanovan and Middle Danubian and G?va/Channelled Ware groups being also noted in the pottery.

 

[Angela]

All I see is more reason to distrust analyses using Eurogenes.

Corsica is represented by ONE sample, "collected", or rather "chosen" by a notorious Italian Nordicist, and that sample is not at all "typically" Corsican like, if there is such a thing, but rather resembles a French admixed Corsican.

That's why "Ligurian" fits, beyond the fact that coastal Corsicans, like Napoleone Buonaparte, have Ligurian (his mother) as well as Tuscan (his father) ancestry.

 

[Riverman]

All I see is more reason to distrust analyses using Eurogenes.

Corsica is represented by ONE sample, "collected", or rather "chosen" by a notorious Italian Nordicist, and that sample is not at all "typically" Corsican like, if there is such a thing, but rather resembles a French admixed Corsican.

That's why "Ligurian" fits, beyond the fact that coastal Corsicans, like Napoleone Buonaparte, have Ligurian (his mother) as well as Tuscan (his father) ancestry.

We are all happy about these tools and their immense potential and we all know that not all samples are perfect, some individuals being admixed. But do you know what, in a recent study samples of "British" being presented among which were people of Levantine origin, most likely Ashkenazi Jewish. So whatever you have to say about G25, many of the "scientific samples" are by no means better and the same applies to the commercial testing companies, if they miss regional differences and ethnic minorities, or obvious individual outliers in their reference samples. Its all work in progress, usually of people with good will, but nobody is perfect and having admixed samples is sometimes better than having none.
If there are "pure Corsican" samples around, a lot of people would be happy to get them into G25, including myself (I can't do it myself though :embarassed: )

 

[Pax Augusta]

All I see is more reason to distrust analyses using Eurogenes.

Corsica is represented by ONE sample, "collected", or rather "chosen" by a notorious Italian Nordicist, and that sample is not at all "typically" Corsican like, if there is such a thing, but rather resembles a French admixed Corsican.

That's why "Ligurian" fits, beyond the fact that coastal Corsicans, like Napoleone Buonaparte, have Ligurian (his mother) as well as Tuscan (his father) ancestry.

Eurogenes almost always uses academic samples (although he chooses which ones to use). The notorious Italian Nordicist has nothing to do with Eurogenes. It is the academic sample from Corsica (at least a dozen individuals) that is composed of some Corsicans that may be partly of French ancestry. While the Ligurian one is always an academic sample and is represented by one individual, unikely representative for the whole Liguria. It is the academic samples that are often not fully representative.

 

[Angela]

Eurogenes almost always uses academic samples (although he chooses which ones to use). The notorious Italian Nordicist have nothing to do with Eurogenes. It is the academic sample from Corsica (at least a dozen individuals) that is composed of some Corsicans that may be partly of French ancestry. While the Ligurian one is always an academic sample and is represented by one individual, unikely representative for the whole Liguria. It is the academic samples that are often not fully representative.

Well, that's news to me. I was specifically told by someone I consider reputable that the samples in the Vahaduo updated list for Dodecad K12b modern Italian populations were from his collection, and that indeed many different samples in that list were submitted by internet "amateurs". If any of those samples were submitted by him from among samples of his acquaintances and given that his main purpose in life seems to be turning Italians into Nordics, then they should be removed from any analyses.

As for the bolded comment I completely disagree. Cavalli Sforza's choice of where to gather samples has stood the test of time although given when they were selected they're not perfect. I realize you don't like the Tuscan samples and what they show; as to whether they're representative or not we'll have to agree to disagree.

If I have to choose between samples submitted by who knows what amateur on the web with who knows what hidden agenda and those chosen by an academician whose identity I at least know and who has a profession to lose I know which I'd choose. You choose whom you want.

Ed. I would be interested to know how it can be verified precisely which Corsican samples were averaged to get the coordinates on Vahaduo and whether they can be definitively shown to come from the paper(s) on Corsicans. My recollection of the latest paper does not at all support them being French admixed, but my recollection may be faulty.

 

[ihype02]

Peloponnesian members of Anthrogenica complained that G25 academic Peloponnesian samples are too southern shifted for them.

If they are academical I guess I pick them.

 

[Riverman]

If I have to choose between samples submitted by who knows what amateur on the web with who knows what hidden agenda and those chosen by an academician whose identity I at least know and who has a profession to lose I know which I'd choose. You choose whom you want.

The good thing with this much larger corpus of samples is that they can be evaluated against each other. Like obvious outliers, either because they are admixed, from ethnic minorities, or some fringe regions, stick out, the more data you have. So everybody is free to "purify" the data and check which difference it makes. Most of the time, the difference is not that big or significant at all.
For Italians in particular its easy to form various smooth gradients which make perfect sense, so whatever people don't like about it, its a minor issue. Instead, you should consider yourself lucky because Italians being one of the best sampled people of all!
Just for comparison, for German speaking people there are only a few non-definitive regional groupings, a lot of outliers and lack of sampling for many regions and ethnolinguistic units. I would be very happy having as much good data, sort by ethnic groups and provinces, like the Italians got it. But I guess the more one gets, the more nitpicking.
I do understand that the samples and method might be not perfect, but its excellent nevertheless and everybody can prove that to him- or herself by using it against known groups and variation. There is nothing which sticks out as being deliberately manipulated with bad intention.
And if there are no official samples, we should be lucky having at least those gathered by amateurs. If they would be completely off, this would show up. And if there are new scientific samples, they won't be held back because of some sort of conspiracy.

 

[Pax Augusta]

Well, that's news to me. I was specifically told by someone I consider reputable that the samples in the Vahaduo updated list for Dodecad K12b modern Italian populations were from his collection and that indeed many samples in that list were submitted by internet "amateurs". If any of those samples were submitted by him from among samples of his acquaintances and given that his main purpose in life seems to be turning Italians into Nordics, then they should be removed from any analyses.

Eurogenes was being discussed. Dodecad K12 is a completely different thing.

As for the bolded comment I completely disagree. Cavalli Sforza's choice of where to gather samples has stood the test of time although given when they were selected they're not perfect. I realize you don't like the Tuscan samples and what they show; that doesn't mean they're not representative.

If I have to choose between samples submitted by who knows what amateur on the web with who knows what hidden agenda and those chosen by an academician whose identity I at least know and who has a profession to lose I know which I'd choose. You choose whom you want.

Some academic samples from Northeastern Italy, from Raveane 2019, end up among the Croatians and Hungarians. But of course, academic samples are always very accurate. Are we sure that it is the amateurs who turn some Italians into Nordics? You should avoid to drag the discussion onto a personal level. For me it's not a research about my ancestry or myself, let alone my identity, I'm not blinded by a research about myself unlike many others. Accuracy is all that ever matters to me. Those who really know me know this.

You are free to believe what you want but time has shown that Cavalli Sforza was wrong about many things, and you seem to forget that Alberto Piazza was Cavalli Sforza's main collaborator.

 

[iluvatar]

Peloponnesian members of Anthrogenica complained that G25 academic Peloponnesian samples are too southern shifted for them.

If they are academical I guess I pick them.

That was regarding the old average, which included only three samples, and understandably didn’t capture the entire variation. The new one is more extensive and from an academic paper as well, though i don’t recall which one exactly. I think it was Sarno et al.

 

[Pax Augusta]

I would be interested to know how it can be verified precisely which Corsican samples were averaged to get the coordinates on Vahaduo and whether they can be definitively shown to come from the paper(s) on Corsicans. My recollection of the latest paper does not at all support them being French admixed, but my recollection may be faulty.

I can only use the G25 at the moment, so as to take everything with due caution. There are 16 Corsicans in the 2019 study as I remember, in the G25 Davidski included 14 of them. Out of 14, 4 are outliers, but others could also be affected by the same problem though to a much lesser extent. The Corsicans are a very small population, and an island population at that. You don't expect this variation to exist, and certainly not that a Corsican could end up between Trentino-Alto Adige and Northeastern Italy, two representative averages of the Alps.


French_Corsica:corsica03708,
French_Corsica:corsica1308,
French_Corsica:Corsica19508,
French_Corsica:Corsica24508,
French_Corsica:corsica29008,
French_Corsica:Corsica29708,
French_Corsica:CorsicaS03308,
French_Corsica:CorsicaS10208
French_Corsica:CorsicaS13808
French_Corsica:CorsicaS29908
French_Corsica:CorsicaS04208
French_Corsica:CorsicaS00708,
French_Corsica:corsica11908,
French_Corsica:Corsica14708,

It is as if they form two clusters.

 

[ihype02]

ORD009 in G25. Daunians are like Western Shifted Tuscans, so Corsica should be the the first modern group they show greatest affinity to.

Distance to:	Italy_IA_Messapic_Daunian_Ordona:ORD009
0.03375273	French_Corsica
0.03402973	Italian_Veneto
0.03500182	Italian_Piedmont
0.03527467	Italian_Tuscany
0.03556619	Swiss_Italian
0.03639077	Italian_Bergamo
0.03884339	Italian_Trentino-Alto-Adige
0.04026252	Italian_Umbria
0.04182701	Spanish_Baleares
0.04239967	Italian_Northeast
0.04291958	Spanish_Mallorca
0.04303520	Italian_Lombardy
0.04311957	Italian_Aosta_Valley
0.04322355	Spanish_Menorca
0.04353634	Italian_Marche
0.04476102	Spanish_Eivissa
0.04508314	Albanian
0.04542251	Greek_Thessaly
0.04542500	Greek_Central_Macedonia
0.04567742	French_Provence
0.04579698	Spanish_Peri-Barcelona
0.04627010	Italian_Lazio
0.04644970	Spanish_Girona
0.04647612	Italian_Liguria
0.04702300	Spanish_Murcia

Distance to:	Italy_IA_Messapic_Daunian_Ordona:ORD009
0.02874863	HRV_MBA
0.03606988	HRV_Pop_CA
0.03979688	ITA_Rome_MA
0.04035270	DEU_MA_Alemannic_o2
0.04069774	ITA_Broion_BA
0.04160810	ITA_Etruscan
0.04203421	Bell_Beaker_ITA
0.04261694	DEU_Roman
0.04282862	ITA_Rome_Late_Antiquity
0.04285214	ITA_Villanovan
0.04285506	Iberia_Northeast_c.8-12CE
0.04459502	CHE_IA
0.04563672	BGR_EBA
0.04626143	HRV_EBA
0.04762398	HUN_BA
0.04798769	UKR_Cimmerian_o
0.04826724	ITA_Tivoli_Renaissance
0.04907844	ITA_Ardea_Latini_IA
0.04960111	ITA_ReginaMargherita_BA
0.04964653	ITA_Rome_Latini_IA
0.05036041	HUN_MBA_Vatya
0.05122268	HRV_IA
0.05124357	ITA_Etruria_Imperial
0.05186858	VK2020_ITA_Foggia_MA
0.05274986	GRC_Helladic_MBA
0.05344087	ITA_Prenestini_tribe_IA_o
0.05383536	DEU_Lech_MBA
0.05404840	Scythian_MDA
0.05410692	BGR_IA
0.05488058	ITA_Proto-Villanovan
0.05633730	DEU_MA_Alemannic_o1
0.05773184	ITA_Sardinia_IA
0.05902228	ITA_Prenestini_tribe_IA
0.05908047	DEU_Lech_EBA_contam
0.05943041	SRB_Mokrin_EBA
0.05944251	Levant_LBN_MA_o4
0.06017542	TUR_IA_low_res
0.06051290	ITA_Boville_Ernica_IA
0.06057405	TZA_Zanzibar_Euro_outlier
0.06072095	Bell_Beaker_HUN

This one was a bit northern let's wait and see for Messapian genomes to be revealed. I think Iron Age Campania will look like this though.

Distance to:	Italian_Apulia
0.04766329	Iberia_Northeast_Empuries2
0.04995370	GRC_Mycenaean
0.05850706	Italy_IA_Messapic_Daunian_Ordona:ORD009

A 0.033 distance means that it is not clustering with anyone like in official PCA despite being close.

 

[Riverman]

A 0.033 distance means that it is not clustering with anyone like in official PCA despite being close.

On the other hand 009 looks like a "pure Northern Illyrian", he is fully in the J2b cluster from the British paper also and his distance to 0.02874863 HRV_MBA speaks for itself. That's a good match. HRV CA and EBA is a much worse match, because they received additional Bell Beaker ancestry through the Middle Danubian Tumulus Culture expansion. The superficial similarity to some French and Italians is primarily due to this Bell Beaker : Mediterranean Neolithic mixture.

 

[ihype02]

On the other hand 009 looks like a "pure Northern Illyrian", he is fully in the J2b cluster from the British paper also and his distance to 0.02874863 HRV_MBA speaks for itself. That's a good match. HRV CA and EBA is a much worse match, because they received additional Bell Beaker ancestry through the Middle Danubian Tumulus Culture expansion. The superficial similarity to some French and Italians is primarily due to this Bell Beaker : Mediterranean Neolithic mixture.

HRV_IA is a pure Northern Illyrian.

 

[Salento]

nice salento
ORD010 he looks western jewish or greek islander profile
maybe he had recent byzantine ancestery
or muslim ancestery

Admix Results (sorted):

#	Population	Percent
1	East_Med	36.04
2	West_Med	20.38
3	North_Atlantic	15.09
4	West_Asian	12.20
5	Red_Sea	9.29
6	Baltic	5.83

Finished reading population data. 204 populations found.
13 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Sephardic_Jewish @ 4.557816
2 Italian_Jewish @ 5.256803
3 Algerian_Jewish @ 5.670169
4 Ashkenazi @ 6.580314
5 South_Italian @ 7.225772
6 East_Sicilian @ 8.048594
7 Tunisian_Jewish @ 8.227295
8 Libyan_Jewish @ 9.225803
9 Central_Greek @ 10.025041
10 West_Sicilian @ 11.006186
11 Italian_Abruzzo @ 12.665685
12 Cyprian @ 13.848021
13 Greek_Thessaly @ 16.500519
14 Lebanese_Muslim @ 17.823469
15 Tuscan @ 18.530861
16 Syrian @ 18.570854
17 Samaritan @ 19.522041
18 Palestinian @ 20.010542
19 Jordanian @ 20.402901
20 Tunisian @ 20.948820

Using 2 populations approximation:
1 50% Samaritan +50% Tuscan @ 2.571835Using 3 populations approximation:
1 50% Samaritan +25% Tuscan +25% Tuscan @ 2.571835


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 Ashkenazi + Samaritan + Sephardic_Jewish + Tuscan @ 2.227096

I think that a 1 to 1 comparison on the chromosome level, reveals things that a distance calculator cannot.

Salento me (from Messapia) VS Daunians ORD009 and ORD010 - One-to-One at 60 SNPs (same as MTA).

regardless of autosomal distances, I still get a Generational distance with all the 16 samples, … most samples are low coverage besides ORD009 and ORD010, … so keep that in mind …

S vs ORD009 ~517 BC

ORD010 ~1088 AD

 

[ihype02]

Dienekes and Davidski both have their own biases. But Davidski's biases are still more down to earth compared to Dienekes.