Genetic study Ancient DNA of Roman Danubian Frontier and Slavic Migrations (Olalde 2021)

[torzio]

it is damn hard for me to protect his method
but since the lazaridis paper with the marathon sample is not published yet
and we dont have the danubian limes bam files of the balkan bronze age cluster yet
he took the sources that might be resonable to creat for alike sources


his explanations:

Averaged HRV IA and BGR IA under assumption you'll get a fair Balkans IA centroid that way. The Sclaveni probably already had a lot of this ancestry before entering Greece IMO.

Avar Szolad for the northern Euro ancestry in the Sclaveni.

Emporiote and Balkans IA might be somewhat interchangeable here for obvious reasons. BGR IA already had quite a bit of affinity to Mycenaeans/Emporiotes anyway. There might have been a cline in antiquity from Balkans IA-like in most of the Balkans to more Mycenaean-like as you neared Greece itself.

I thought he used the term to represent the ancient Antes and not the north-caucasian Avars ..................which is it ?

 

[kingjohn]

I thought he used the term to represent the ancient Antes and not the north-caucasian Avars ..................which is it ?

it is not avars from caucasus
it is the barbarian avars from szoald grave yard
in g25 they cluster with modern poles
so this refernce it is a good source for slavic admixture influence
on modern greeks

 

[Dushman]

I never understood why the Germanics in the Balkans are completely ignored, when in my eyes they were the fathers of migrating en masse, migrating to distant locations, and settling somewhere else. Not forgetting how the Celts settled even in the Balkans and even as far as Galatia in Anatolia, it?s just that Germanics fit better the time period we?re discussing.

Since that Szolad sample fits Poland so well and oftentimes Poland is used as a proxy for Slavic admixture, does any member here have reliable data on the percentage of Germanic admixture in modern Poland?

I?d be surprised if Poland is less than 50% Gothic/Germanic.

 

[Vallicanus]

I never understood why the Germanics in the Balkans are completely ignored, when in my eyes they were the fathers of migrating en masse, migrating to distant locations, and settling somewhere else. Not forgetting how the Celts settled even in the Balkans and even as far as Galatia in Anatolia, it�s just that Germanics fit better the time period we�re discussing.

Since that Szolad sample fits Poland so well and oftentimes Poland is used as a proxy for Slavic admixture, does any member here have reliable data on the percentage of Germanic admixture in modern Poland?

I�d be surprised if Poland is less than 50% Gothic/Germanic.

I'd be extremely surprised if Poland was anywhere near that figure.:useless:

 

[ihype02]

I'd be extremely surprised if Poland was anywhere near that figure.:useless:

Same.




...

 

[Ralphie Boy]

good post by michaelis from anthrogenica
so i share it
.

Averaged HRV IA and BGR IA under assumption you'll get a fair Balkans IA centroid that way. The Sclaveni probably already had a lot of this ancestry before entering Greece IMO.

Avar Szolad for the northern Euro ancestry in the Sclaveni.

Emporiote and Balkans IA might be somewhat interchangeable here for obvious reasons. BGR IA already had quite a bit of affinity to Mycenaeans/Emporiotes anyway. There might have been a cline in antiquity from Balkans IA-like in most of the Balkans to more Mycenaean-like as you neared Greece itself.

According to this study, mainland Greeks apparently fit better with Empuries than Roman Greek, who is used to model Aegean Islands, Crete and Cyprus.

 

[Dianatomia]

The islanders are closer to the Roman Greek from Attica than the Greek from Empuries. There are no mainlanders for the Roman Greek sample comparison.

 

[Francis Drake]

Why don't these guys use formal tests for their admixture runs, since when are PCAs used for admixture?

Also what is considered a "good fit"; you can get <2% on the Albanian average with Myceanean + BGR_IA/HRV_IA + Slav. Are we just using arbitrary standards of <1.5% now? How is this supported?

 

[Archetype0ne]

It is not supported. In fact there is such a thing as an overfit. And these are not even close to 1%< are they? Did not see any distances mentioned, maybe I have to check again.

What I do see is p-value of .20 for Albanian under that model.
That p value in any scientific enviroment, sends the null hypothesis straight into the trash. Even social sciences require smth like .05, genetics I would expect .01.

 

[Riverman]

I never understood why the Germanics in the Balkans are completely ignored, when in my eyes they were the fathers of migrating en masse, migrating to distant locations, and settling somewhere else. Not forgetting how the Celts settled even in the Balkans and even as far as Galatia in Anatolia, it?s just that Germanics fit better the time period we?re discussing.

Since that Szolad sample fits Poland so well and oftentimes Poland is used as a proxy for Slavic admixture, does any member here have reliable data on the percentage of Germanic admixture in modern Poland?

I?d be surprised if Poland is less than 50% Gothic/Germanic.

They have pre-Slavic, including Germanic admixture, but its not nearly as high. I think its even possible that the post-migration period German settlement had more of a Germanic impact than the ancient one. But this needs to be tested once more ancient DNA samples come out.

 

[Francis Drake]

It is not supported. In fact there is such a thing as an overfit. And these are not even close to 1%< are they? Did not see any distances mentioned, maybe I have to check again.

What I do see is p-value of .20 for Albanian under that model.
That p value in any scientific enviroment, sends the null hypothesis straight into the trash. Even social sciences require smth like .05, genetics I would expect .01.

I was reffering to G25, and the almost arbitrary methodology used by the people on anthrogenica. Afaik they just made it all up.

Hell, the creator of nmonte calls the scaling process davidski used amateurish and disavows, now its funny how people there post scaled calculator runs like they are oracles.

 

[Archetype0ne]

I was reffering to G25, and the almost arbitrary methodology used by the people on anthrogenica. Afaik they just made it all up.

Hell, the creator of nmonte calls the scaling process davidski used amateurish and disavows, now its funny how people there post scaled calculator runs like they are oracles.

I would love the quote about the first part.

Yeah, I think it is just fun and games. People play around with these calculators just like I do.
But the main reason is because people really do not want to get into statistics and use stuff like fstats. For the most part, if my intuition is correct, all you need is to run some lines of code? Not sure.
Think I will pick up that skill one of these weekends.

 

[Francis Drake]

I would love the quote about the first part.

Yeah, I think it is just fun and games. People play around with these calculators just like I do.
But the main reason is because people really do not want to get into statistics and use stuff like fstats. For the most part, if my intuition is correct, all you need is to run some lines of code? Not sure.
Think I will pick up that skill one of these weekends.

For admixtools running it is pretty easy, you just follow one of the main two tutorials on github and type some commands in RStudio. I can download any dataset from Reichlab and run qpadm on any sample, the thing is all the samples im playing with have to be in the dataset already, I have no damn idea how to format the snps and so on, I think that is the difficult part.

 

[Archetype0ne]

For admixtools running it is pretty easy, you just follow one of the main two tutorials on github and type some commands in RStudio. I can download any dataset from Reichlab and run qpadm on any sample, the thing is all the samples im playing with have to be in the dataset already, I have no damn idea how to format the snps and so on, I think that is the difficult part.

I see. Thanks. Would love a link to the said github.
Also, not sure what format the datasets are in, but admixstudio has tools to convert most stuff.
Thing is unless you have some unpublished DNA sitting in your fridge you likely do not need to even bother with the datasets. Just to run commands to try different combinations for the fstats.

There is this Finish guy on Anthro who is a wiz, Ngansakhan, running, visualizing/contextualizing data, like he is pouring cereal. But he is mostly interested in Finno Urgic peoples . The few times Balkans were in his visuals, it was the dopest thing I have seen yet in 6 years on fora.

Just found the images.
This thread has the codes also. https://www.eupedia.com/forum/threa...khan-from-Anthrogenica?highlight=Nganasankhan

If this aint the sickest thing you have seen, show me what beats it.

These are so freaking relevant for our discussion:

Edit: The images are in very high resolution, right click -> open image in new tab to zoom in.

 

[Francis Drake]

I see. Thanks. Would love a link to the said github.

uqrmaie1 . github . io /admixtools/ articles /recipes . html (easiest one to follow)

uqrmaie1 . github . io /admixtools/ articles /admixtools . html

The second one also provides context for what the input files are aswell, for example.

 

[Dushman]

I'd be extremely surprised if Poland was anywhere near that figure.:useless:

Right, Polish is the epitome of being Slavic.


The Goths (not Early Goths or Proto-Goths) were a conglomerate of Southern Scandinavians, locals, possibly other Germanics, and the fact that Poles are pulled towards such populations is shown in every chart I?ve seen. I?m of the opinion they pulled even West Ukraine (at least West) towards Scandinavia after migrating towards the Black Sea.

A similar genetic pull seems to be completely ignored even in the Balkans, so no wonder even North Italy gets the ?realistically? 8-12% Slavic admixture. I think a quick recap of early Slavic history can remind everyone that the Slavic migrations weren?t as large scale as you think, affecting even Italy or even Crete by 10-15%.

Much of that ?Slavic? input was probably already in the Balkans and already Romanised, as indicated by it?s presence is several historically Vlach populations.

Whoever cares about the subject can also read the study below.

https://www.nature.com/articles/s41598-019-43183-w

Pretty evident to me that Poles are a better proxy for Goths (not Early Goths) than Slavs.

 

[Vallicanus]

North Italy gets the �realistically� 8-12% Slavic admixture.

Really?

 

[ihype02]

According to this study, mainland Greeks apparently fit better with Empuries than Roman Greek, who is used to model Aegean Islands, Crete and Cyprus.
View attachment 12912

I think it says Greek islands fit better with the Marathon sample, nothing about mainlanders.
Mountain regions of Peloponnese probably remained unmixed with the Anatolian genetic layer that came in some rural zones. So far we shouldn't assume the Marathon sample as the right one for southern mainland Greece without additional information and samples.

 

[Angela]

Really?

Clearly, to repeat myself once again, based on history and ydna there's a very small amount of Slavic in northern Italians. Clearly, some members don't know that by the time the Slavs were on the move the Langobards were already in Italy, and stopped them at the border, and there is next to no "Slavic" yDna even in northern Italy.

It's Langobard, a people who were from the area around Denmark. We know their route of travel, and we already know what the majority of the Langobards were like from the ancient dna we have from them, and they weren't Slavs autosomally. Also, all of the men so far were U-106.

As for the Goths, they were an extremely small elite group who couldn't control Italy with their small numbers. Italians remained in most administrative positions of all kinds, which is why it was still a relatively civilized period in Italian history. It didn't completely crash until the arrival of the Langobards. See, this is what you understand if you know about the history of an area, instead of just making things up out of nothing.

The Poles are pulled toward the west because of the Germanic in them, not the other way round, if they knew anything about the history of eastern Europe.

 

[ihype02]

Right, Polish is the epitome of being Slavic.
View attachment 12913

The Goths (not Early Goths or Proto-Goths) were a conglomerate of Southern Scandinavians, locals, possibly other Germanics, and the fact that Poles are pulled towards such populations is shown in every chart I�ve seen. I�m of the opinion they pulled even West Ukraine (at least West) towards Scandinavia after migrating towards the Black Sea.

A similar genetic pull seems to be completely ignored even in the Balkans, so no wonder even North Italy gets the �realistically� 8-12% Slavic admixture. I think a quick recap of early Slavic history can remind everyone that the Slavic migrations weren�t as large scale as you think, affecting even Italy or even Crete by 10-15%.

Much of that �Slavic� input was probably already in the Balkans and already Romanised, as indicated by it�s presence is several historically Vlach populations.

Whoever cares about the subject can also read the study below.

https://www.nature.com/articles/s41598-019-43183-w

Pretty evident to me that Poles are a better proxy for Goths (not Early Goths) than Slavs.

Not always, in some academic papers, Poles are significantly more eastern shifted from the Germanic speakers. Just like Tuscans are significantly more Western shifted compared to the Albanians. And it's not even the only one.