Comparing Ancient Greek populations to modern Greeks and Italians

[Archetype0ne]

Good job! That must have taken a ton of work, I recall replicating the Danubian Limes, was really tedious at points.

Do your results match one to one with the study? Or are there slight differences?

 

[lacreme]

I got Nebula Genomics WGS30x done a couple years ago.

right now you can get it for only $99.

*plus a mandatory lifetime subscription during the checkout process bringing the total to around $300 or so...Surely cheaper than the regular price but still relatively high.

 

[Jovialis]

Good job! That must have taken a ton of work, I recall replicating the Danubian Limes, was really tedious at points.

Do your results match one to one with the study? Or are there slight differences?

The outgroups are a little slapdash, because the ones used in the study, oddly are not present. But I see there are equivalents. I can post the ind file and prompt when I get a chance.

 

[Jovialis]

*plus a mandatory lifetime subscription during the checkout process bringing the total to around $300 or so...Surely cheaper than the regular price but still relatively high.

Frankly, I don't regret getting it. But if the cost is not feasible, it is better to wait.

This process used to cost $1 million in 2007.

 

[Jovialis]

It would be interesting to replicate this model for ancient population as well, such as italics and Greeks

That would be interesting indeed, I actually ran it against C6 samples, and by the Imperial Era, C6 resembles modern Southern Italy. Prenestini_o 437 gets about 10% CHG_IN with more Anatolia_N in place of it.

 

[Idontknowwhatimdoing]

That would be interesting indeed, I actually ran it against C6 samples, and by the Imperial Era, C6 resembles modern Southern Italy. Prenestini_o 437 gets about 10% CHG_IN with more Anatolia_N in place of it.

C6 is not a real population anyway, they are samples that are mixed with everything that coincidentally cluster with Italians. Some are literal Germanic-Levantine mixes.

 

[felipesc500]

Great job!!!

 

[Francesco]

C6 is not a real population anyway, they are samples that are mixed with everything that coincidentally cluster with Italians. Some are literal Germanic-Levantine mixes.

I believe that's true for all clusters from the imperial age. Indeed, one should take them with a grain of salt, taking into account the archeological context and other genetic data. With that in mind, they could probably be used as proxies of actual populations

 

[Jovialis]

C6 is not a real population anyway, they are samples that are mixed with everything that coincidentally cluster with Italians. Some are literal Germanic-Levantine mixes.

How have you come to this conclusion? If it was with G25, wasn't it established that it is a PCA projection that could make that conclusion by coincidence?


They could not be mixes, because they are identified separately. How could they be mixes within and of themselves, if the point of separating them is to demonstrate future generations mix among them? They did not, C5 and C4 left from the scene by late Antiquity; as determined by the study, so I don't see the point in revisiting that.


Eastern Mediterranean stopped immigrating to Rome when the empire in the west fell. Rome was repopulated with the native people who were indeed C6, which we know has existed since the Iron Age. So what you are saying is incorrect.


They could not be mixes, because they are identified separately. How could they be mixes within and of themselves, if the point of separating them is to demonstrate future generations mix among them? They did not, C5 and C4 left from the scene by late Antiquity; as determined by the study, so I don't see the point in revisiting that.

The genetic profile of southern Italians originated by at least the early bronze age, C6 represents that population, as well for much of the center. C5 is the immigrant population that arrived in the Imperial era, which seems to disappear in Late Antiquity.


It is a fact that C6 people from the Imperial age, and even the Iron Age overlap with modern central and southern Italians. This is not a coincidence, it is because they are sourced from those people. Even when comparing all of the cohorts separately in qpadm it was the only viable source.

 

[Idontknowwhatimdoing]

How have you come to this conclusion? If it was with G25, wasn't it established that it is a PCA projection that could make that conclusion by coincidence?


They could not be mixes, because they are identified separately. How could they be mixes within and of themselves, if the point of separating them is to demonstrate future generations mix among them? They did not, C5 and C4 left from the scene by late Antiquity; as determined by the study, so I don't see the point in revisiting that.


Eastern Mediterranean stopped immigrating to Rome when the empire in the west fell. Rome was repopulated with the native people who were indeed C6, which we know has existed since the Iron Age. So what you are saying is incorrect.


They could not be mixes, because they are identified separately. How could they be mixes within and of themselves, if the point of separating them is to demonstrate future generations mix among them? They did not, C5 and C4 left from the scene by late Antiquity; as determined by the study, so I don't see the point in revisiting that.

The genetic profile of southern Italians originated by at least the early bronze age, C6 represents that population, as well for much of the center. C5 is the immigrant population that arrived in the Imperial era, which seems to disappear in Late Antiquity.


It is a fact that C6 people from the Imperial age, and even the Iron Age overlap with modern central and southern Italians. This is not a coincidence, it is because they are sourced from those people. Even when comparing all of the cohorts separately in qpadm it was the only viable source.

The c6 people from the imperial age were mixed with 100 populations. In the C6 cluster you sent me there was a literal half Germanic and half Levantine individual. They just found a bunch of samples that coincidentally plot together on PCA and assumed they must be related.

It is 100% impossible that the profile of Southern Italians originated in the Bronze Age. I explained many times why. Even the picture you sent shows totally different proportions of medieval Italians vs Iron age.

https://i.imgur.com/2fcC9xo.png

The fact that the study completely missed such outliers that only plotted coincentally with other samples proves for the 100th time how clueless these academics are. They are oversimplifying admixture history with irrelevant PCA's.

I explained many times with qpAdm and G25 models how the Neolithic proportions of modern Italians dont match up any Bronze age iron age samples from Italy.

Even the qpAdm models you ran showed P values proving that Minoan + Yamnaya models are 100% impossible while my models with the West Anatolian cluster pass.

They could not be mixes, because they are identified separately. How could they be mixes within and of themselves, if the point of separating them is to demonstrate future generations mix among them? They did not, C5 and C4 left from the scene by late Antiquity; as determined by the study, so I don't see the point in revisiting that.

They were literal mixes, "identified separately",that is exactly the problem with these academics, they just assumed that they have the same admixture based on coincidental PCA PLOTS. When you mix Germanic + Levantine it plots with the C6 cluster on PCA. So they assumed they are the same population when they are clearly 100% proven to be not related.

For example when you simulate 50% Germanic + 50% Levantine it clusters with South Italians and it is almost identical to the C6 sample and this proves that their C5 C6 whatever clusters are all over the place, they simply coincidentally plot close on PCA.

as determined by the study, so I don't see the point in revisiting that.

The only thing that that study determined is that they don't know how to determine real unmixed populations from coincidental clusters.

 

[Jovialis]

First of all, this is central Italy, not the south. C6 probably came from the southern area.

Also, when I made the model, I didn't modify the outputs. Because it merely a learning exercise for me. It is already been shown in studies, in the supplement of Raveane et al. 2019 that you can indeed model Minoan and Steppe for southern Italy. I was only trying to replicate it, yet didn't account for all of the nuance at the time.4

Also, please stop trying to dismiss other theories as impossible, especially when studies have shown they are indeed valid as theories. As well as produced in qpAdm

At any rate, I am actually quite busy and will have to revisit this at a later date. To be continued.

 

[Idontknowwhatimdoing]

First of all, this is central Italy, not the south. C6 probably came from the southern area.

Also, when I made the model, I didn't modify the outputs. Because it merely a learning exercise for me. It is already been shown in studies, in the supplement of Raveane et al. 2019 that you can indeed model Minoan and Steppe for southern Italy. I was only trying to replicate it, yet didn't account for all of the nuance at the time.4

Also, please stop trying to dismiss other theories as impossible, especially when studies have shown they are indeed valid as theories. As well as produced in qpAdm



At any rate, I am actually quite busy and will have to revisit this at a later date. To be continued.

The studies havent proven anything, their theories are absolutely invalidated by the fact that they included confirmed random mixed outliers in their C clusters. The just made arbirtary clusters based on coincidental PCA positions

For example when you simulate 50% Germanic + 50% Levantine it clusters with South Italians and it is almost identical to the C6 sample, this proves that their C5 C6 whatever clusters are all over the place, they simply coincidentally plot close on PCA.

Matter of fact, if you mix Cypriot + North Italian you get identical South Italian profile because that is how it kinda happened. Greeks from Eastern colonies were migrating to Italy for hundreds of years. We know for a fact that such profiles did not exist in such proportions in the iron age and 50% of Rome's samples appears to be Greek between 0-200AD. In sicily the Himera civilians are Anatolian mixed but Bronze and iron age Sicilian samples have zero such admixture, it is just ANF + Steppe + WHG, no excess iran n/natufian in such proportions, a signal of Anatolian Greek admixture

 

[Idontknowwhatimdoing]

First of all, this is central Italy, not the south. C6 probably came from the southern area.

It is already been shown in studies, in the supplement of Raveane et al. 2019 that you can indeed model Minoan and Steppe for southern Italy.

These are unscientific models with unscientific references/outgroups that broke all qpAdm theory rules/standards, just because a model got a good p value it doesn't mean its true. Many papers are breaking qpAdm standards just to push their own biases. Most people reading the paper don't know about qpAdm standards.

Also South Italians have too much Iran N and not enough ANF proportionally to be just a mix of Minoan + Steppe, they require 50% Anatolian Greek.

 

[Jovialis]

These are unscientific models with unscientific references/outgroups that broke all qpAdm theory rules/standards, just because a model got a good p value it doesn't mean its true. Many papers are breaking qpAdm standards just to push their own biases. Most people reading the paper don't know about qpAdm standards.

Also South Italians have too much Iran N and not enough ANF proportionally to be just a mix of Minoan + Steppe, they require 50% Anatolian Greek.

Enough with this hysterical bs, this was conducted by people much more knowledgeable than you, and passed peer-review by people more knowledgeable than you. Flat out rejection of the model just because it doesn't appeal to your sensibility and pet theory is not sufficient.

 

[Francesco]

The studies havent proven anything, their theories are absolutely invalidated by the fact that they included confirmed random mixed outliers in their C clusters. The just made arbirtary clusters based on coincidental PCA positions
For example when you simulate 50% Germanic + 50% Levantine it clusters with South Italians and it is almost identical to the C6 sample, this proves that their C5 C6 whatever clusters are all over the place, they simply coincidentally plot close on PCA.

Some samples may indeed plot in the C6 cluster by mere chance, but others seem quite consistent with the contemporary southern Italian population, even in the G25 model you posted above

 

[Jovialis]

@Idontknowwhatimdoing

You're free to post your models and share your insights, but I recommend that you stop annoying me. Also, cool it with the polemics, these are all just theories.

 

[Idontknowwhatimdoing]

Enough with this hysterical bs, this was conducted by people much more knowledgeable than you, and passed peer-review by people more knowledgeable than you. Flat out rejection of the model just because it doesn't appeal to your sensibility and pet theory is not sufficient.

Yet they failed to detect obviously confirmed Germanic-Levantine mixed outliers that only coincidentally plotted with others on the PCA.

 

[Idontknowwhatimdoing]

Some samples may indeed plot in the C6 cluster by mere chance, but others seem quite consistent with the contemporary southern Italian population, even in the G25 model you posted above

Thats because if you have a big capital cosmopolitan city that has a ton of West Anatolians, Germanics, Italics, Levantines and people mixed with all of these with various proportions then obviously you will find samples that cluster with modern Italians from all regions. The C6 cluster is a cluster mixed with Greek Anatolian, Italic, Germanic and Levantine. So that does not disprove that modern Italians have Graeco-Anatolian ancestry because no such C6 cluster existed in the Bronze and Iron age. Im sure we will find more samples like that in the Iron age if we sample Italy a lot but these will just be relatively recent arrivals.

 

[Jovialis]

How is C6 a mix, when they decern those populations from one another. What you are saying makes no sense, based on that.

C6 is a population separate from east med (C5), Levantine (C4), and Germanic (C7)

C6 existed before and after those groups arrived.

C6 was a native group formed in the EBA, and that is supported by pre-historic migration.

Yet they failed to detect obviously confirmed Germanic-Levantine mixed outliers that only coincidentally plotted with others on the PCA.

What you are zeroing in on is a coincidence of PCA. I already showed you actual qpAdm from professionals that prove you wrong.

When I provide evidence from peer reviewed studies that make my case, you think dismissing their credibility makes you somehow more credible? It doesn't, in fact it does the exact opposite.

 

[Jovialis]

It is likely the city of Rome started to get C6 people during the unification of Italy, from the South.

The CHG/IN came from farmer migrations that went directly:

modern South Italians, can INDEED be modeled with Steppe/Minoan with some augmentation according to academic studies: