Teal people found: Caucasians!

[Alan]

I haven't had time to think this through yet. Why would there be such a disparity between the admixture and the IBD analysis for Italy. It's much less for IBD, plus, the results are reversed north vs south.

Maybe one part of Italy has more post Yamna CHG like ancestry which will show up as CHG like/Descend admixture. This might have come from a post CHG source thousands of years later.

 

[Angela]

Maybe one part of Italy has more post Yamna CHG like ancestry which will show up as CHG like/Descend admixture. This might have come from a post CHG source thousands of years later.

Maybe, except look at the admixture graph again. Not only southern Italy, but also Iberia and part of Aquitaine in France have very low levels in terms of admixture with CHG. It's not something particular to southern Italy. Perhaps, given that the "Indo-European" speaking groups might have arrived in these areas quite a bit later than they arrived in central and northern Europe, their genetics were slightly different, and they didn't have much CHG left. Then, in the IBD sharing map, the southern parts of Italy wind up with higher IBD sharing than northern Italy, but still, particularly in terms of mainland southern Italy, lower than you would think they would have if there was a significant relatively recent input from the Near East. They don't even have the same amount as Greece, not even Sicily, although it has more. I would say that's down to Greek colonies, but coastal southern Italy also had Greek settlements, and it doesn't have the same levels. Plus, in all the old calculators, Sicily and mainland Greece and even parts of the Balkans were just about tied for "West Asian", and southern Italians sometimes had a little less. What was that all about? Why doesn't it correlate? How can Greece have the same amount of "West Asian" as Sicily, but more CHG in admixture? Another interesting thing I noticed is that Cyprus has a lot of CHG, but Crete doesn't. Were those subsequent cultural changes in Crete more a matter of cultural diffusion than a lot of migration. Might that also apply to southern Italy? Are these just "older" populations?

As for northern and central Italy, they have as much CHG in admixture as the French in so far as I can tell from the map, but less in IBD than the French. That's why I think Arvistro may be on to something when he says that some of this may be due to drift. In other parts of Europe there's been continuous mixing, but according to Ralph and Coop Italians haven't mixed with others for a very long time, not even with each other. The Alps were a greater barrier than is sometimes thought, perhaps, and provided enough isolation for a good deal of drift in those areas. The political fragmentation of Italy also played its part, and the mountainous terrain.

Italian genetics are still a puzzle to me even after a couple of years of poring over every paper on the subject and population genetics in general. I have yet to find a really good amateur analysis either.

Even the PUNT one, which seems to work very well for most nationalities, is awful for me. Every other calculator I've ever tried, and 23andme too, have me as midway between Bergamo and Firenze, which makes perfect sense both in terms of my genealogical records and my "location" on the map of Italy. (Some do have better Fsts than others.) Yet the PUNT analysis doesn't give me any Italian population as first choice. (Maybe in this case he just doesn't have enough Italian reference populations.)

I just don't think Italian genetics can be explained until we get a lot of ancient samples.

Anyway, to be continued this week end.

 

[Regio X]

I haven't had time to think this through yet. Why would there be such a disparity between the admixture and the IBD analysis for Italy. It's much less for IBD, plus, the results are reversed north vs south.

Someone in Anthrogenica posted these averages of Eurasia K11 CHG-NAF (GedrosiaDNA in GEDmatch): https://drive.google.com/file/d/0B-ObXiVfL-RzOEpDQ2U3MTJzNTA/view Compared to that autosomal map, they are a bit different.

 

[LeBrok]

As I said in the past, Central and South Europe was more attractive for the early farmers in comparison to Northeast Europe as extreme example, which is colder and less suited than even the green highlands of the Scottland. Must be another reason why the Northwest has more EF than Northeast. Even most of Scandinavia (especially the South) should have been better for farming. Therefore the population density in those regions must have been allot higher. And because of that especially South Europe via contact to Western Asia had early Civilizations and should have been more isolated for nomadic herders. I don't think the Early PIE were really warlike people I imagine them more like the "typical immigrants" leaving their homeland for better land/life in every possible direction, and "overpopulating over time regions of Europe, West Asia and South_Central Asia. And as we know the newcomers are always likely to breed more and sometimes bring disseases only they are "immune" to. I kinda think this is how they seriously expanded their language. Not very different from how the Early farmers expanded in search of new land.

Keep in mind that every one thousand years or so, there were farmer population crashes due to colder climate and crop failure. These were the best time for invasions from the Steppe, better adapted populations to harsher climate over depopulated farmer Central and North Europe, and because they were pushed out of Steppe by even harsher conditions. It could have been the same mechanism which has brought down the Roman Empire and pushed more tribes from Steppe to Central Europe. Likewise pushed Slavs from Eastern Europe to depopulated the Center and Balkans. Almost exactly mimicking Corded Ware expansion.

 

[holderlin]

Eurasian Basil

 

[holderlin]

 

[bicicleur 2]

Same with Balts and Germans.
Technically I understand it this way:
IBD ("identical by descent"?) is some hypothetical degree up to which population might be descendant of CHG (or CHG like) population.
Admixture is what portion of genes could be assigned to CHG (or CHG like) population box versus being put into other reference populations/clusters boxes used in analysis.

If what I think is right, IBD should not change by adding other reference populations into equation.
On other hand Admixture should be very sensitive to what other reference populations (other boxes) are.
For example, if only WHG was used together with CHG to find admixture, then all Euros would show increased CHG (because EEF and ANE portion would split between WHG and CHG), because any units used in analysis would be put by model into either CHG or WHG box.
If however a lot of reference populations were used, then CHG ratio would decrease as it may be eaten by other references.

So, to your question, disparity might be because part of what is considered CHG by admixture is assigned to CHG for lack of better reference population. Apparently some other reference folk is needed, the closest to which (of used folks) is CHG.
Alternatively it is drift - used to be CHG, drifted away enough to become non-recognized as IBD, but still being closest to CHG of all other available choices.

admixture is not an absolute yardstick, it is a relative yardstick

 

[bicicleur 2]

You have confused both Uyghurs, Kyrgyz for Turkmens who are indeed very similar to Tadjiks autosomally and most likely the most "pure " Iranic group being absorbed or turkified.

However even Uyghurs and Krygyz have a decent amount of Iranic and Tocharian admixture.

what about the R1b in Turkmenistan?
was it a safe haven when Turkish tribes conquered the steppe?

 

[Fire Haired14]

What you said is that: "South Europe takes up a small minority of Europe." That isn't true even in terms of land mass, much less in terms of population, as you would discover if you did the math. This kind of imprecise language isn't helpful.

I just proved that land wise it is a small minority. It depends how you define South Europe. In genetics it's three peninsulas. There's no point in arguing about this.

As far as I'm concerned, "Europeans" didn't really exist genetically until about 2000 BC when the major part of this tripartite admixture occurred. So, if the Yamnaya people weren't yet Europeans then neither were the MN or the WHG or the EEF. The definition of "Europeans" is very much time dependent.

Ancient DNA in summary tells us.
>First Europeans(at 14,000yo) were WHG.
>Farmers from the Aeagan set up settlements in every-part of Europe. They mixed with WHG but remained mostly non-WHG. They became the new European.
>People in Russia who themselves were native/West Asian hybrids, immigrated into Europe and mixed with EEF/WHG.

This is why I see WHG in 6000 BC and EEF/WHG in 3000 BC as Europeans. Russia was never apart of Europe genetically. Europe was more so what today we call West and South Europe. Today putting all people that live in Europe under the same genetic title makes more sense than it did back then.

The reason Russians today fit in Europe genetically, is because of at least two back migrations into Russia from Central Europe, one in 2500 BC and one in 500 AD.

 

[Alan]

what about the R1b in Turkmenistan?
was it a safe haven when Turkish tribes conquered the steppe?

Most likely local Iranic R1b. The whole Region from West to Northeast Iran has a significant percentage of R1b too.

 

[Greying Wanderer]

Angela

I haven't had time to think this through yet. Why would there be such a disparity between the admixture and the IBD analysis for Italy. It's much less for IBD, plus, the results are reversed north vs south.

Concentration?

If admixture measures the total amount and IBD measures levels of related DNA then maybe in Germany the DNA got spread out evenly whereas in Italy the mountains divided the population into regional segments.

The north (edit: of Italy) showing high admixture and low IBD in the Po valley vs both high admixture and high IBD in the alps may be a microcosm of this.

I wouldn't be surprised if an IBD map of Italy had separate regional hotspots for most of the disparate components that have been discovered all neatly separated into bowls by the various mountain ranges.

 

[Alan]

Someone in Anthrogenica posted these averages of Eurasia K11 CHG-NAF (GedrosiaDNA in GEDmatch): https://drive.google.com/file/d/0B-ObXiVfL-RzOEpDQ2U3MTJzNTA/view Compared to that autosomal map, they are a bit different.

Interesting looking at the fst tables.


Closest to CHG

1. Kalash 0.074
2. WHG 0.084
3. Neolithic Anatolian Farmers 0.087
4. South Indian 0.089
5. EHG 0.107



Closest to EHG

1. CHG 0.107
2. Kalash 0.112
3. WHG/South Indian 0.122
4. Anatolian Farmers 0.123


Closest to WHG

1. CHG 0.084
2. Kalash 0.101
3. South Indian 0.107
4. Anatolian Farmer 0.109
5. EHG 0.122


Closest to Kalash

1. CHG 0.074
2. South Indian 0.082
3. WHG 0.101
4. Anatolian Farmer 0.106
5.EHG 0.112


closest to South Indian

1. Kalash 0.082
2. East Asian 0.087
3. CHG 0.089
4. WHG 0.107
5. Anatolian Farmer 0.115


Closest to Anatolian Farmer

1. CHG 0.087
2. Kalash 0.106
3. WHG 0.109
4. East African 0.114
5. South Indian 0.115

Conclusions out of this?


CHG is literally the only component that shows significant closeness to any of the other major West Eurasian type components. It shares closest relation to Kalash, WHG, Anatolian Farmer and South Indian. It is also the only component that shows at least some closeness to EHG. Almost like the center of all.


EHG is literally not very close to any component beside some to CHG. It is even closer to Kalash than to WHG. And South Indian is equal close to it as is WHG!


expected or unexpected, WHG shows it's closest relation to CHG. Than followed by Kalash (everything that shows close relation to CHG seems to show also to Kalash), South Indian and Anatolian farmer. EHG again not very close and beside general West Eurasian affinities nothing to show close relationship.
Makes me wonder if some people were right with their theory that "ANE" is not a real component and just hiding some WHG like ancestry in South_Central Asians. So "South Indian" and "Kalash" might indeed hide some WHG like ancestry in it. We know that Anatolian farmers are halfway "WHG-UHG" like no suprise that they show affinities.

Kalash as expected very close to CHG, followed than by South Indian. My explanation for that is, Kalash represents slightly drifted version of CHG which settled in South_Central Asia, and a part of that new Kalash component moved into the Indian subcontinent and merged with the local H&G creating the "South Indian" component. And again opposite to what you would expect geographically, Kalash is closer to WHG as to EHG.


The South Indian component seems to be a hybrid of West and East Eurasian DNA (as we knew for long). It must be something "Proto Dravidian" + Something Proto Southeast Asian. This is why they score as second closest "East Asian". And again despite geographic closeness, even South Indian scores better with WHG than with EHG.


Anatolian Farmer scores best with CHG of course followed by the "CHG like" Kalash component. Than we have WHG next, most likely due to the WHG-UHG like ancestry in them. After that comes East African which is quite easy to explain as we know from the recently tested ancient East African skeletons. The East African component is a hybrid of Levant farmers and pre Neolithic Africans.

 

[Greying Wanderer]

Bunch of stuff

#

Tomenable

So your "massive migration from the steppe to Western Europe which entirely omitted Eastern Europe even though they had to cross it on their way to Western Europe because - hey - that's geography" theory still has some major flaws.

I think this ties in with thinking about PIE and IE as two stages when I think it was more likely three stages, that is PIE was a collection of components which may have come from multiple sources but all somehow ended up together somewhere north of the Black/Caspian Seas - stage one. Then over time these components developed into the full IE package which included lots of tattooed dudes on chariots with bronze weapons - stage two (or stage three imo).

I think there was an interim stage two between PIE and the full IE package which had most of the elements but critically *not* the full military component.

My personal labels for the three stages are PIE, wagon IE (or copper IE) and horse IE (or bronze IE).

So to me wagon IE passing through eastern Europe (and the middle east and central Asia etc) as traders/artisans without conquering anybody (but becoming very influential) seems perfectly plausible.

Stage 1) PIE
Stage 2) wagon/copper IE spreading all over as traders/artisans
Stage 3) horse/bronze IE spreading all over as conquerors

#

Goga

Aryans spoke Iranic... The Medes were Aryans, the ancient Persians were Aryans.

It only occurred to me a few months back but "Eire", "Erin", "Eireann" are old words for Ireland - which was quite mind blowing at the time.

#

Once upon a time, 4,000 to 8,000 years after humanity invented agriculture, something very strange happened to human reproduction. Across the globe, for every 17 women who were reproducing, passing on genes that are still around today—only one man did the same.

Women who have all sons still pass on their mtdna but men who have all daughters don't pass on their y dna. Seems to me this must be a factor somehow.

This process must always have existed but what might cause a sudden increase in the severity of the process with agriculture?

Inheritance?

#

north vs south europe

These definitions vary with context but in the context of ancient migrations I think you need four divisions because of the different physical routes in and out of Europe:

1) coastal south (including southern France)
2) central-danubian - danube, hungary, northern France, southern Germany
3) northern - north of the Carpathians to Baltic and North sea
4) atlantic coast - where the other three streams merged

 

[LeBrok]

Interesting looking at the fst tables.


CHG is literally the only component that shows significant closeness to any of the other major West Eurasian type components. It shares closest relation to Kalash, WHG, Anatolian Farmer and South Indian. It is also the only component that shows at least some closeness to EHG. Almost like the center of all

Well, it is in geographical center of all. Though I didn't expect to correlate with center so nicely, better than expected.

EHG is literally not very close to any component beside some to CHG. It is even closer to Kalash than to WHG. And South Indian is equal close to it as is WHG!

Yep, I'm surprise that EHG is so far away from all of them, even from WHG.

expected or unexpected, WHG shows it's closest relation to CHG. Than followed by Kalash (everything that shows close relation to CHG seems to show also to Kalash), South Indian and Anatolian farmer. EHG again not very close and beside general West Eurasian affinities nothing to show close relationship.
Makes me wonder if some people were right with their theory that "ANE" is not a real component and just hiding some WHG like ancestry in South_Central Asians. So "South Indian" and "Kalash" might indeed hide some WHG like ancestry in it. We know that Anatolian farmers are halfway "WHG-UHG" like no suprise that they show affinities.

Kalash as expected very close to CHG, followed than by South Indian. My explanation for that is, Kalash represents slightly drifted version of CHG which settled in South_Central Asia, and a part of that new Kalash component moved into the Indian subcontinent and merged with the local H&G creating the "South Indian" component. And again opposite to what you would expect geographically, Kalash is closer to WHG as to EHG.

I can't believe Kalash is closer to WHG than to EHG! I guess because it is close to CHG, which is related to WHG.

The South Indian component seems to be a hybrid of West and East Eurasian DNA (as we knew for long). It must be something "Proto Dravidian" + Something Proto Southeast Asian. This is why they score as second closest "East Asian". And again despite geographic closeness, even South Indian scores better with WHG than with EHG.

That's another surprise. There might be a lot of ENF and CHG material in S Indians from all the migrations. If ancient S Indian was tested, we would have had way different results perhaps.

 

[LeBrok]

Bunch of stuff

#

Tomenable

I think this ties in with thinking about PIE and IE as two stages when I think it was more likely three stages, that is PIE was a collection of components which may have come from multiple sources but all somehow ended up together somewhere north of the Black/Caspian Seas - stage one. Then over time these components developed into the full IE package which included lots of tattooed dudes on chariots with bronze weapons - stage two (or stage three imo).

I think there was an interim stage two between PIE and the full IE package which had most of the elements but critically *not* the full military component.

My personal labels for the three stages are PIE, wagon IE (or copper IE) and horse IE (or bronze IE).

So to me wagon IE passing through eastern Europe (and the middle east and central Asia etc) as traders/artisans without conquering anybody (but becoming very influential) seems perfectly plausible.

Stage 1) PIE
Stage 2) wagon/copper IE spreading all over as traders/artisans
Stage 3) horse/bronze IE spreading all over as conquerors

I like it.

 

[Angela]

Someone in Anthrogenica posted these averages of Eurasia K11 CHG-NAF (GedrosiaDNA in GEDmatch): https://drive.google.com/file/d/0B-ObXiVfL-RzOEpDQ2U3MTJzNTA/view Compared to that autosomal map, they are a bit different.

He didn't even include northern Italy as a sample? No wonder I wind up in the Balkans. France isn't there either, so I can't compare the CHG for northern Italy and France on the PUNT calculator with the CHG admixture map done by the other blogger.

However, at least here in this PUNT calculator the CHG figures for mainland Greece and Sicily are identical, which correlates with every other analysis which has been done which measures "West Asian". Given that mainland Greece to one degree or another was influenced by the Slavic migrations, which must have diluted some of these "components" to some degree, the CHG must once have been even higher. I guess the Romans dumped a lot of Near Eastern, and/or, what was it, oh yes, Parthian slaves in Greece, too. (Sarcasm Alert)

Also, once again, Albanians and Tuscans are very similar in terms of the admixture percentages. I guess some very West Asian like ancient Etruscans also colonized Albania. (Another Sarcasm Alert)

I also don't understand the EHG and WHG figures in that calculator. The Yamnaya, Corded Ware etc. are, correct me if I'm wrong, still being modeled with EHG and CHG in about equal proportions. I see the CHG alright, but lower than you would expect if northern Europeans can be modeled as 50% Yamnaya, and the EHG numbers are lower yet. Also, why are the WHG numbers so high all of a sudden? This doesn't correlate with any other analysis, does it? Everyone was using an ancient genome for WHG, none of this ghost population stuff, and they never showed any figures like this. I don't see how the addition of Bichon should change things, given how homogeneous they were as a group. Might the number be inflated because it is showing the additional WHG picked up in Europe by the farmers? Still, I don't think that would account for all of it. Is the calculator failing to discriminate properly between EHG and WHG?

Also, while it's interesting seeing the percentages for these "components", it's not very helpful in terms of figuring out what happened with the Bronze Age migrations into Europe. The newcomers weren't mixing with Anatolian farmers. They were mixing with MN people.

Just generally in terms of the CHG, it's still not clear to me that all of it came with "Yamnaya like" groups. I'll speak to it more directly in the thread about the Greek Neolithic, but prior work has suggested that in Greece the big change autosomally was from the early Neolithic to the Middle Neolithic, and there was no big change in the Bronze Age. It's possible that's true for other parts of southern Europe too.

 

[Alan]

One thing I realized about this calculator again, it must be flawed if the source for CHG(Satsurbila) is only ~80% CHG and 15% EHG by this calculator. This is not possible if the CHG samples are older than the EHG samples and what seems to have happened here. The CHG ancestry in EHG is been shown as EHG ancestry in CHG.

Therefore the real CHG frequency might be slightly higher for all populations.


EDIT: Yep obviously this calculator is biased towards EHG, that gets visible when looking at the Yamna frequency. This can't even be Caucaso_Gedrosia. Yamna is shown as 83% EHG and 3% CHG. A good chunk of CHG is gettting eaten up by EHG. No way that 12.000 BC Satsurbila has 20% of EHG ancestry. This calculator shows even me as 11% East African. Sorry but the scores on this calculator are simply out of place. Reminds me once again to never give too much trust in most amateur calculators.

 

[Angela]

One thing I realized about this calculator again, it must be flawed if the source for CHG(Satsurbila) is only ~80% CHG and 15% EHG by this calculator. This is not possible if the CHG samples are older than the EHG samples and what seems to have happened here. The CHG ancestry in EHG is been shown as EHG ancestry in CHG.

Therefore the real CHG frequency might be slightly higher for all populations.


EDIT: Yep obviously this calculator is biased towards EHG, that gets visible when looking at the Yamna frequency. This can't even be Caucaso_Gedrosia. Yamna is shown as 83% EHG and 3% CHG. A good chunk of CHG is gettting eaten up by EHG. No way that 12.000 BC Satsurbila has 20% of EHG ancestry. This calculator shows even me as 11% East African. Sorry but the scores on this calculator are simply out of place. Reminds me once again to never give too much trust in most amateur calculators.

I don't know who made the maps, but isn't this the calculator made by Kurd on Anthrogenica? I also think it's off, as I said in post #336.

"I also don't understand the EHG and WHG figures in that calculator. The Yamnaya, Corded Ware etc. are, correct me if I'm wrong, still being modeled with EHG and CHG in about equal proportions. I see the CHG alright, but lower than you would expect if northern Europeans can be modeled as 50% Yamnaya, and the EHG numbers are lower yet. Also, why are the WHG numbers so high all of a sudden? This doesn't correlate with any other analysis, does it? Everyone was using an ancient genome for WHG, none of this ghost population stuff, and they never showed any figures like this. I don't see how the addition of Bichon should change things, given how homogeneous they were as a group. Might the number be inflated because it is showing the additional WHG picked up in Europe by the farmers? Still, I don't think that would account for all of it. Is the calculator failing to discriminate properly between EHG and WHG?"

 

[Alan]

Yep it is probably from the user "Kurd" generally his calculations make sense. He even calculated that yamna is 50/50 CHG/EHG. But he seems to have some.errors in his calculator yet.

I think the main reason for that is, that his calculator eats up shared ancestry which probably came from chg to ehg but is shown here as ehg ancestry in chg. This might be one of the main reasons.

As a side effect the EHG scores of virtually everyone are too high.

 

[MOESAN]

He didn't even include northern Italy as a sample? No wonder I wind up in the Balkans. France isn't there either, so I can't compare the CHG for northern Italy and France on the PUNT calculator with the CHG admixture map done by the other blogger.

However, at least here in this PUNT calculator the CHG figures for mainland Greece and Sicily are identical, which correlates with every other analysis which has been done which measures "West Asian". Given that mainland Greece to one degree or another was influenced by the Slavic migrations, which must have diluted some of these "components" to some degree, the CHG must once have been even higher. I guess the Romans dumped a lot of Near Eastern, and/or, what was it, oh yes, Parthian slaves in Greece, too. (Sarcasm Alert)

Also, once again, Albanians and Tuscans are very similar in terms of the admixture percentages. I guess some very West Asian like ancient Etruscans also colonized Albania. (Another Sarcasm Alert)

I also don't understand the EHG and WHG figures in that calculator. The Yamnaya, Corded Ware etc. are, correct me if I'm wrong, still being modeled with EHG and CHG in about equal proportions. I see the CHG alright, but lower than you would expect if northern Europeans can be modeled as 50% Yamnaya, and the EHG numbers are lower yet. Also, why are the WHG numbers so high all of a sudden? This doesn't correlate with any other analysis, does it? Everyone was using an ancient genome for WHG, none of this ghost population stuff, and they never showed any figures like this. I don't see how the addition of Bichon should change things, given how homogeneous they were as a group. Might the number be inflated because it is showing the additional WHG picked up in Europe by the farmers? Still, I don't think that would account for all of it. Is the calculator failing to discriminate properly between EHG and WHG?

Also, while it's interesting seeing the percentages for these "components", it's not very helpful in terms of figuring out what happened with the Bronze Age migrations into Europe. The newcomers weren't mixing with Anatolian farmers. They were mixing with MN people.

Just generally in terms of the CHG, it's still not clear to me that all of it came with "Yamnaya like" groups. I'll speak to it more directly in the thread about the Greek Neolithic, but prior work has suggested that in Greece the big change autosomally was from the early Neolithic to the Middle Neolithic, and there was no big change in the Bronze Age. It's possible that's true for other parts of southern Europe too.

I don't know if it can help but I have since a long time the impression (confirmed by the little aDNAwe have) that Copper and even Bronze Age in Europe ought in some part to a very southern population without too much Yamnayalike imput. We have Montenegro people and Copper of Hungary and Spain. Could these people be akin to the last wave of Neolithic? either someones from Neolithic, acculturated to metallurgy skills or reinforced their auDNAb by new waves from S-East with beginning metallurgy, new waves of same human stock? largely EEF plus some evolved CHGlike imput from East? Cyprus and Egea people of the time?