Moots: Ancient Rome Paper

[Leopoldo Leone]

Leopoldo, they are not, and mostly the same. The only J1 clade in Italians not shared with Levantines so far is this branch J1-BY94, https://www.yfull.com/tree/J-BY94/. Which represents anywhere from 20-30% of J1 in Southern Italy. Then J1-CTS1460 and J1-PF7263, CTS1460 is likely to be a Kura Araxian marker being heavily found in the Caucausus, then PF7263 origin is still not determined but is equally found in Europe and the Middle East, the remaining half of J1 is found under J1-Z1853 which is the marker of Semitic speakers. Of Z1853 the two most common clades found in Southern Italians and Italians in general are J1-YSC76 and J1-L829. Both are Levantine markers YSC76 has been found in several Levantine sites including Beirut, Hazor, Meggido, etc.

J2 is pretty much the same as J1, except the shift of focus is Anatolia and the Caucasus. But all clades are shared, the main reasons are the movements of all Mediterranean Empires, Roman, Byzantine, Ottoman (in the case of moving to Middle East, I know Italy was never under the Ottomans, this is for shared clades in Aegean moving East). So far from Levant in terms of J2 we have J2b-M205 which is found in Italians mostly in Sicily though, and for now there is J2a-M92 unfortunately the samples weren’t strong enough to get further snp analysis, so we’ll have to see when more samples come out, letting you know M92 is the third most common J2 branch for all Italians. But again we’ll see if it’s all M92 or just a portion like CTS2906 or the PF7412.

It also doesn’t end at J1 and J2, there are other markers like E-M123 branches, E-V22, E-V12, etc... which are not trivial in numbers E-V12, E-V22 and E-M84 usually are over 1% throughout.

An entire "citation needed" disclaimer is needed, but let's work with what you say:
As for J1, you yourself say that only half could be "unambigously" connected to Semitic speakers, which would mean that only half of the already few J1-carriers in Italy would descend paternally from men that spoke semitic tongues; furthermore, the Eupedia page about J1 states that it is only the L858 subclade of the J1-P58 subclade, and it also states that P58 itself likely originated from eastern Anatolia, so could be linked to the expansion of CHG in Europe, and lastly it says that MOST of the J1 in Europe, Anatolia and the Caucasus is NOT of the P58 variety.
We are not bound to take it as gospel but often it is reliable enough, and given we are home we could ask Maciano to provide the sources.

As for Jb, you have already said that "the shift is on Anatolia and the Caucasus", so again it could be linked to the Iran_N/CHG geneflow into Europe, and furthermore the Eupedia page on J2 states that the in subclades in Italy are likely of Greek origins, because they belike the varieties that peak on Crete, and the Z435 variety is linked with Romans, so I don't see how Jb can be taken as a Levantine uniparental marker; also, https://yhrd.org/tools/branch/J2b-M205, M205 seems to have nothing to do with the Levant, as does M92, https://yhrd.org/tools/branch/ J2a-M92

As for E, V12 is found also in french Basque, https://en.wikipedia.org/wiki/Haplogroup_E-V12, and the other two are canonically linked to the Balkans and north Africa respectively.

First, provide a reliable source that shows that what you are saying isn't made up, and secondly argue why those are "unambigously semitic markers"; to be thorough, some such markers have been found, not only limited to Italy (and to be precise, the J1 distribution doesn't seem to have any north-south gradient in distribution), but to Europe as a whole, and, and this is what is cogent to the discussion, not in the amount required to back up your claim that uniparental data "shows a not trivial levantine gene flow in Italy".

P.S.
I think we have already said that G25 samples are really" funny" to say the least, as their results are totally at odds with the results of academic papers. The only one, as far as I know, that could give it some support is the Sarno paper about the "east med continuum" but its results have never been replicated, and in his latest paper he used no Levant_N to model Calabrian Greeks, and that paper had the flaw of lumping the Levant with Anatolia, so much of that "near east, not sardinian-like" admixture is likely Anatolian-like rather than Levant-like, and I say so because of the known ancient cline made up of a mix of EEF and CHG/Iran_N that span SE europe and Anatolia. Also, and I am going to repeat myself, is it so reasonable to trust user reported samples when the results are so at odds with the scientific literature and when there is a known history of presenting untrustworthy genetic samples as "south Italians"?

 

[Jovialis]

@Azzurro, Sarno et al. 2021, found this modeling to be most appropriate:

FYI, for Iran_N, they really mean non-steppe related CHG-related. But both are very similar and I believe used in the same cohort of samples for analysis often, such as in Raveane et. al. This ancestry is highly differentiated from Natufian.

Canary Island also doesn't make sense to use, because it already has Natufian admixture in it. Also, another issue is that the eurogenes modeling is not picking up WHG, which is probably being subsumed by corded ware. This is why I never have too much faith is trying to achieve such precise analysis with these calculators. Rather is it better to use them for broader analysis, such as the 2-ways modeling. Not to mention, analyzing very old and distant samples like WHG has a lot of room for error.

 

[brick]

It is enough to add the Neolithic sample Tepecik_Ciftlik_N and both Iran_N and Natufian descend significantly in the Italian samples.

 

[Leopoldo Leone]

Also, the often quoted Fernandes 2019 (a little trivia: it was published exactly 2 years and a day ago) shows the exact opposite of what many people who cite claim: in the PCA Sicilians pre-IE samples are closer to both Anatolia and Greece samples (to be precise between Iberia and Greece, as you'd expect), while Sardinia samples are closer (touching) the samples from Iberia, so the samples do not show that the pre-IE folks in Sicily were Sardinian-like, but already shifted towards the north east mediterranean.

 

[Azzurro]

An entire "citation needed" disclaimer is needed, but let's work with what you say:
As for J1, you yourself say that only half could be "unambigously" connected to Semitic speakers, which would mean that only half of the already few J1-carriers in Italy would descend paternally from men that spoke semitic tongues; furthermore, the Eupedia page about J1 states that it is only the L858 subclade of the J1-P58 subclade, and it also states that P58 itself likely originated from eastern Anatolia, so could be linked to the expansion of CHG in Europe, and lastly it says that MOST of the J1 in Europe, Anatolia and the Caucasus is NOT of the P58 variety.
We are not bound to take it as gospel but often it is reliable enough, and given we are home we could ask Maciano to provide the sources.

As for Jb, you have already said that "the shift is on Anatolia and the Caucasus", so again it could be linked to the Iran_N/CHG geneflow into Europe, and furthermore the Eupedia page on J2 states that the in subclades in Italy are likely of Greek origins, because they belike the varieties that peak on Crete, and the Z435 variety is linked with Romans, so I don't see how Jb can be taken as a Levantine uniparental marker; also, https://yhrd.org/tools/branch/J2b-M205, M205 seems to have nothing to do with the Levant, as does M92, https://yhrd.org/tools/branch/ J2a-M92

As for E, V12 is found also in french Basque, https://en.wikipedia.org/wiki/Haplogroup_E-V12, and the other two are canonically linked to the Balkans and north Africa respectively.

First, provide a reliable source that shows that what you are saying isn't made up, and secondly argue why those are "unambigously semitic markers"; to be thorough, some such markers have been found, not only limited to Italy (and to be precise, the J1 distribution doesn't seem to have any north-south gradient in distribution), but to Europe as a whole, and, and this is what is cogent to the discussion, not in the amount required to back up your claim that uniparental data "shows a strong levantine gene flow in Italy".

Dude common, https://www.yfull.com/tree/J-M205/ has been found in every single ancient Levantine site as well as Ancient Egypt. I have all of it tracked down from Ancient samples, I have an entire thread on Anthrogenica where every single J2 has been found in ancient dna and the corresponding clade, J-M205 is a non argument, there is nothing to say it is a confirmed marker of the ancient Levant and Egypt, the oldest sample we have of J2b-M205 is in EBA Jordan.

J1-P58's downstream clade Z1853 which I stated along with J2b-M205 are the Iran Neolithic markers that moved into the Levant and mixed with the local ANF+Natufian population. The dates are practically identical and both have been found together in every single ancient site. With all fairness I respect Maciamo's genetics section but for J1 and J2 it is highly outdated. Like I said so far we have 2 M92 in Middle Bronze Age Meggido, which at the moment are the oldest J-M92, you didn't understand what I wrote, for M92 specific branches can be Levantine in origin and others could have an Anatolian origin.

J-Z435 is linked with Roman expansion, but is not a Roman marker, has not been found in ancient Italics, and J-Z435 itself is highly contested, many samples are old in the Near East as well. Currently there is 3 theories for it, an Anatolian origin, Levantine origin and a Greek origin, we will have to wait and see what the ancient dna will yield, so far all the ancient samples under J-Z435 are young, we need Iron Age and Bronze Age samples to figure out its origin.

https://www.yfull.com/tree/J-Z435/

Umm no they are not, E-V13 is linked with the Balkans, and I did not mention E-V13. So what if E-V12 is found in the French Basque, their source will ultimately be either a Levantine or Egyptian one, it is a founder effect.

Do me a favour and look at the ftdna projects and see what branches people are falling under.

 

[Azzurro]

It is enough to add the Neolithic sample Tepecik_Ciftlik_N and both Iran_N and Natufian descend significantly in the Italian samples.

It still proves my point that you need a Levantine source, the whole argument of the other side is that there is none.

 

[Azzurro]

@Azzurro, Sarno et al. 2021, found this modeling to be most appropriate:

FYI, for Iran_N, they really mean non-steppe related CHG-related. But both are very similar and I believe used in the same cohort of samples for analysis often, such as in Raveane et. al. This ancestry is highly differentiated from Natufian.

Canary Island also doesn't make sense to use, because it already has Natufian admixture in it. Also, another issue is that the eurogenes modeling is not picking up WHG, which is probably being subsumed by corded ware. This is why I never have too much faith is trying to achieve such precise analysis with these calculators. Rather is it better to use them for broader analysis, such as the 2-ways modeling. Not to mention, analyzing very old and distant samples like WHG has a lot of room for error.

I use Canary Islander to account for North African admix, it would be the best proximate source for Berber ancestry in Sicily and Malta.

 

[brick]

None of these models can be considered definitive evidence. Because of the possible sampling error, because of the errors in G25, and because of the very nature of these tools, which are tools that calculate only possible estimates that depend on the choices made (models, reference samples) and are not definitive proofs of anything.

Again it is enough to change the settings, without Tepecik_Ciftlik_N and ADC: 0.5x RC these are the results, both Iran_N and Natufian descend significantly in the Italian samples.

 

[brick]

It still proves my point that you need a Levantine source, the whole argument of the other side is that there is none.

I am not saying that there is nothing, something is certainly there the further south you go in Italy, I am saying that it is not possible to calculate it with precision, as it is not possible to understand what has happened for example in southern Italy without the ancient samples that are still missing.

 

[Leopoldo Leone]

Dude common, https://www.yfull.com/tree/J-M205/ has been found in every single ancient Levantine site as well as Ancient Egypt. I have all of it tracked down from Ancient samples, I have an entire thread on Anthrogenica where every single J2 has been found in ancient dna and the corresponding clade, J-M205 is a non argument, there is nothing to say it is a confirmed marker of the ancient Levant and Egypt, the oldest sample we have of J2b-M205 is in EBA Jordan.

J1-P58's downstream clade Z1853 which I stated along with J2b-M205 are the Iran Neolithic markers that moved into the Levant and mixed with the local ANF+Natufian population. The dates are practically identical and both have been found together in every single ancient site. With all fairness I respect Maciamo's genetics section but for J1 and J2 it is highly outdated. Like I said so far we have 2 M92 in Middle Bronze Age Meggido, which at the moment are the oldest J-M92, you didn't understand what I wrote, for M92 specific branches can be Levantine in origin and others could have an Anatolian origin.

J-Z435 is linked with Roman expansion, but is not a Roman marker, has not been found in ancient Italics, and J-Z435 itself is highly contested, many samples are old in the Near East as well. Currently there is 3 theories for it, an Anatolian origin, Levantine origin and a Greek origin, we will have to wait and see what the ancient dna will yield, so far all the ancient samples under J-Z435 are young, we need Iron Age and Bronze Age samples to figure out its origin.

https://www.yfull.com/tree/J-Z435/

Umm no they are not, E-V13 is linked with the Balkans, and I did not mention E-V13. So what if E-V12 is found in the French Basque, their source will ultimately be either a Levantine or Egyptian one, it is a founder effect.

Do me a favour and look at the ftdna projects and see what branches people are falling under.

I own up my slip on E-V22 and E-V13, but E-V22 has been found also in Asturians, and it is a bit hard to link that place with any Phoenician or Moorish legacy, given that neither set foot there, and the picture is still muddy to say the least.

Also, J2-M205 is found also overall Europe, and J2-M92 even in India, so it is clear that you can't simply state "it is an ancient (of the last 5,000 years) Egyptian and Levantine marker", and either a CHG/Iran_N origin makes up for a much more parsimonious explanation, given that all these places have some admixture from it.

The problems with your approach is that you take some haplogroups, throw away all the most plausible explanations, and build your own narrative around them discarding all the contrary evidence.
If I am understandin you correctly, it seems you are saying that half of J1 in Italy is "surely of Levantine origins", and also a good chunk of J2. Also, I think it has already been noticed that potentially shared J1 and J2 subclades might be just shared CHG/Iran_N, so I am not sure that the argument "the oldest subclade we have is from place X so it must be a marker of ancestry from that place" is airtight.

 

[Azzurro]

I am not saying that there is nothing, something is certainly there the further south you go in Italy, I am saying that it is not possible to calculate it with precision, as it is not possible to understand what has happened for example in southern Italy without the ancient samples that are still missing.

That is the thing, I always said its range, in the first or second post to Jovialis I said it is somewhere between 5-20%, we going to get another Ancient Roman Italy paper (with various Italic tribes), several Ancient Greek ones and one on the Iron Age Middle East those will papers will be crucial going forward. We can more accurately calculate autosomal and uniparental inheritance.

 

[Leopoldo Leone]

I am not saying that there is nothing, something is certainly there the further south you go in Italy, I am saying that it is not possible to calculate it with precision, as it is not possible to understand what has happened for example in southern Italy without the ancient samples that are still missing.

It is not a good idea to take G25 results as better than academic results, which didn't show any Levant_N needed for south Italians. By the way, Anatolian Farmers had some Levant_N so more EEF ancestry would mean more Levant_N, but the thing is that it seems that north Italians have more EEF than south Italians, who have less IE than the former but also another admixture that is either Iran_N or CHG, and in fact it is all you need to model them ( and a bit of north african in Sicilians and few Calabresi), and it is only G25 samples showing that you need more Levant_N to model south Italians. Why don't we stick to what is reliable?

 

[Azzurro]

I own up my slip on E-V22 and E-V13, but E-V22 has been found also in Asturians, and it is a bit hard to link that place with any Phoenician or Moorish legacy, given that neither set foot there, and the picture is still muddy to say the least.

Also, J2-M205 is found also overall Europe, and J2-M92 even in India, so it is clear that you can't simply state "it is an ancient (of the last 5,000 years) Egyptian and Levantine marker", and either a CHG/Iran_N origin makes up for a much more parsimonious explanation, given that all these places have some admixture from it.

The problems with your approach is that you take some haplogroups, throw away all the most plausible explanations, and build your own narrative around them discarding all the contrary evidence.
If I am understandin you correctly, it seems you are saying that half of J1 in Italy is "surely of Levantine origins", and also a good chunk of J2. Also, I think it has already been noticed that potentially shared J1 and J2 subclades might be just shared CHG/Iran_N, so I am not sure that the argument "the oldest subclade we have is from place X so it must be a marker of ancestry from that place" is airtight.

J-M205 is non argument, we can discuss other clades if you'd like. J-M205 is equivalent to saying R-M269 and Indo Europeans, the parsimonious explanation is Iron Age and Classical Age Levantine people moving across the Mediterranean. It isn't even a top 20 clade in frequency for J2 in Italy.

E-V22 in Austrias could easily be a paternal ancestor from the Levant who arrived there during Roman era and also not necessarily, could have originated with a paternal Phoenician ancestor from Southern Spain who moved up? All countries have internal movement.

 

[brick]

Iran_N and Natufian can also come out in the Balkans with these models using G25. Why is no one asking how they got there?

It is not a good idea to take G25 results as better than academic results, which didn't show any Levant_N needed for south Italians. By the way, Anatolian Farmers had some Levant_N so more EEF ancestry would mean more Levant_N, but the thing is that it seems that north Italians have more EEF than south Italians, who have less IE than the former but also another admixture that is either Iran_N or CHG, and in fact it is all you need to model them ( and a bit of north african in Sicilians and few Calabresi), and it is only G25 samples showing that you need more Levant_N to model south Italians. Why don't we stick to what is reliable?

I am fully aware of this. Read my second post.

 

[Leopoldo Leone]

J-M205 is non argument, we can discuss other clades if you'd like. J-M205 is equivalent to saying R-M269 and Indo Europeans, the parsimonious explanation is Iron Age and Classical Age Levantine people moving across the Mediterranean. It isn't even a top 20 clade in frequency for J2 in Italy.

E-V22 in Austrias could easily be a paternal ancestor from the Levant who arrived there during Roman era and also not necessarily, could have originated with a paternal Phoenician ancestor from Southern Spain who moved up? All countries have internal movement.

So did also iron and classical age Levantine moved to India? I might be recalling wrongly, but isn't J-M205 more common in the Balkans compared to Italy? Especially northern Balkans? did ancient Levantines not only reach those places, but also left such a significat genetic legacy? Don't get me wrong, some haplos are indeed linked to Phoenicians or other Levantines, but not those you claim, which are much more common.

Lastly, earnestly do you put forwards such explanations? E-V22 haplo shows no correspondency with supposed Phoenician or Moorish settlements in the area, so you must posit internal migrations that "reversed" the original distribution.

 

[Azzurro]

Iran_N and Natufian can also come out in the Balkans with these models using G25. Why is no one asking how they got there?

I am fully aware of this. Read my second post.

Who said no one is asking? Greece already had an abundance of Iran Neo since Bronze Age both Minoans and Cycladic culture had a lot of Iran Neo, the Natufian is the bigger the question, my guess is the same way how Italy got theirs, just add a layer with the Hellenistic era. From Hellenistic through Roman through Byzantine you have over 1000 years of opportunity for this component to reach the Aegean and Balkans, and case in point would be the J-M205 clade in the Balkans.

 

[Azzurro]

So did also iron and classical age Levantine moved to India? I might be recalling wrongly, but isn't J-M205 more common in the Balkans compared to Italy? Especially northern Balkans? did ancient Levantines not only reach those places, but also left such a significat genetic legacy? Don't get me wrong, some haplos are indeed linked to Phoenicians or other Levantines, but not those you claim, which are much more common.

Lastly, earnestly do you put forwards such explanations? E-V22 haplo shows no correspondency with supposed Phoenician or Moorish settlements in the area, so you must posit internal migrations that "reversed" the original distribution.

Your right J-M205 is more common in the Northern Balkans than Italy, https://www.yfull.com/tree/J-Y22059/ the clade is 1000 years old, and is probably a Byzantine era founder effect. I am no longer discussing J-M205, if you chose not to believe its Levantine in origin its on you, this is a fact.

Please do me a favour and look at E-V22's phylogentic tree

https://www.yfull.com/tree/E-V22/

Are you thinking of J-L283 instead of J-M205? J-L283 is historically old in Europe, some Bronze Age Caucasus movement into the Balkans, and has been found throughout several Balkan and Italian ancient sites.

 

[Leopoldo Leone]

Your right J-M205 is more common in the Northern Balkans than Italy, https://www.yfull.com/tree/J-Y22059/ the clade is 1000 years old, and is probably a Byzantine era founder effect. I am no longer discussing J-M205, if you chose not to believe its Levantine in origin its on you, this is a fact.

Please do me a favour and look at E-V22's phylogentic tree

https://www.yfull.com/tree/E-V22/

So I guess that the J-M205 in India, Scandinavia and even central asia was ancient Phoenicians settling there, and if you think that postulating founder effects after founder effects based on implausible internal migrations is scientific you might want to review a basic logics handbook.

P.S.

is this outdated as well? How did J-M205 end up in India and central asia spread by Levantine travellers?
https://www.eupedia.com/forum/threads/32953-New-map-of-Y-haplogroup-J2b1-(M205)

 

[Azzurro]

So I guess that the J-M205 in India, Scandinavia and even central asia was ancient Phoenicians settling there, and if you think that postulating founder effects after founder effects based on implausible internal migrations is scientific you might want to review a basic logics handbook.

LOL, it was found in every single Levantine civilization and Egypt not only Phoenicians. So it can be Phoenician, Jewish, Syrian, Assyrian, Edomite, Egyptian, Amorite, Aramean, etc...

The Scandinivan, the Indian and the Central Asian ones all have an ancestor who belonged to one of those mentioned groups, again this is a fact and facts don't care about your feelings.

 

[brick]

That is the thing, I always said its range, in the first or second post to Jovialis I said it is somewhere between 5-20%, we going to get another Ancient Roman Italy paper (with various Italic tribes), several Ancient Greek ones and one on the Iron Age Middle East those will papers will be crucial going forward. We can more accurately calculate autosomal and uniparental inheritance.

I do not think it will be easy to calculate even with the publication of new ancient samples, certainly now it is impossible. Since with the same models using the G25, never forget it is an amateur tool, these components come out even to the Greeks, and the Greeks had an important role in Italy, it is clear that not everything can be attributed to foreign presence during imperial Rome. Moreover, we must consider that the genetics of the Balkans has been greatly altered by the Slavic migrations of the medieval era and that in the past Balkans may have played an important role as a bridge between Italy and the Near East. Finally, the Natufian sample itself is still based on one or two individuals, isn't it? It is highly problematic.

Who said no one is asking? Greece already had an abundance of Iran Neo since Bronze Age both Minoans and Cycladic culture had a lot of Iran Neo, the Natufian is the bigger the question, my guess is the same way how Italy got theirs, just add a layer with the Hellenistic era. From Hellenistic through Roman through Byzantine you have over 1000 years of opportunity for this component to reach the Aegean and Balkans, and case in point would be the J-M205 clade in the Balkans.

Natufian has likely many problems and is the big question, I agree. With the same model used with Italians and Balkans, many Iberian populations (both Spanish and Portuguese) get out significant percentages of Natufian. Only by putting SSA do these percentages decrease but not disappear. The strange thing is that Natufian comes out to some Iberian samples, to others it does not. Does it really make sense? This suggests to me that these results cannot be taken too seriously, for a number of reasons already listed, possible sampling error, because of the possible errors in G25, and because of the very nature of these tools, which are tools that calculate only possible estimates that depend on the choices made (models, reference samples) and are not definitive proofs of anything.