PDA

View Full Version : Yamna_K6



Fire Haired14
13-03-15, 23:29
This is just something Davidski is trying out.

http://eurogenes.blogspot.com/2015/03/yamnaya-related-ancestry-proportions-in.html

Eurogenes K15 and ANE K8 spreadsheets for comparison and reference.

https://docs.google.com/spreadsheets/d/19c_bZjUV_RouKyGyLHmMDw57WwAVabXFJOaso_gcuRE/edit#gid=1872836177

https://docs.google.com/spreadsheets/d/1kd9Q9vFrL1Cra9ayqMYVFKXrUdnThmQJVMtjczLhoTs/edit#gid=74932529

This test was an unsupervised run. If I'm interpreting Davidski correctly that means when he tried to create ancestor clusters with modern(excluding northeast Euros, Caucasus, and south Asia because of drift) and ancient genomes a pure Yamna cluster appears which most of the Yamna samples almost score 100% in.

The rest of west Eurasian's ancestry is mostly thrown into two clusters; Middle eastern and pre-Yamna European(prob. very EEF-like). Those two should turn out very similar to each other(ENF+WHG, no ANE?).

Looks like there's certainly room for a lot of Yamna-like ancestry in Central, south, and west Asia. Kazakhs look like an even mix of Yamna(Ancient Indo Iranians?) and something very east Asian-like(Turks?).

skaheen15
14-03-15, 00:32
Interesting.
So, at this point, there can be little doubt that Yamnaya went a long way toward creating modern Europe's population.

Alan
14-03-15, 04:53
By god this guy will try anything in his power to turn his beloved Northeast Europeans as "Yamna" like as possible. Haak paper says Russians/Lezgians closely followed by Norwegians have second most Yamna DNA after Mordovians.

But in his work Polish are as Yamna as Norwegians and Lithuanians even more Yamna than Russians. Lezgians who are according to the actual study as close to Yamna as Russians but here appear less Yamna like than 3/4 of any of the European populations. Suprise!

And why exclude Northeast Europe, Caucasus and South Asia for his ancestral components? Doesn't it turn out how he wants? Going by this Calculator, contrary to Haak paper. Armenians are less than 15% Yamna.

I quote

But I'd say this looks like a very reasonable attempt, with more or less comparable results to those published by Haak et al.

Yup just with the "small" difference that different from the Haak paper suddenly the Northeast Euros turn out more Yamna like than Norwegians. And Lithuanians turn out as much Yamna as Mordovians!! While in the original paper that wasn't the case. This can only be explained that this "calculator" is making populations with more "European genes" appear more Yamna. Explains also why Spaniards here are more Yamna than Turks while in the Haak paper it was the opposite.
https://drive.google.com/file/d/0B9o3EYTdM8lQd19HNEdENTdtZTg/view


The agenda shines through his work like the sun in the hot summer days. If you can't bring the PIEs to the Baltics, than try to bring the Baltics to PIEs.

Fire Haired14
14-03-15, 05:24
Alan, Polish weren't tested in Haak 2015. His numbers are very similar to Haak 2015's, they don't have to be identical to be correct.

You're basing your statement that Lezgins are as close to Yamna as Russians are on one PCA from Haak 2015, that's not nearly enough. I don't think Davidski forced an agenda into this test.

Alan
14-03-15, 05:33
Alan, Polish weren't tested in Haak 2015. His numbers are very similar to Haak 2015's, they don't have to be identical to be correct.

Just compare the Yamna ancestry in Lithuanians vs Norwegians and Yamna ancestry in Lithuanians vs Norwegians at the Haak paper. Lithuanians in Eurog. K6 have suddenly significantly more Yamna than Norwegians. He has created a calculator which once again turns all towards Northeast Europeans.

According to this new "calculator" Armenians are less than 15% Yamna! while according to Haak paper they are ~25% Yamna like. That can only be explained that the number of Yamna ancestry in more "West Asian" admixed populations is pushed down while the number in more "EHG" admixed populations pushed up (suprise that is basically Northeast Europeans).


You're basing your statement that Lezgins are as close to Yamna as Russians are on one PCA from Haak 2015, that's not nearly enough. I don't think Davidski forced an agenda into this test.

I am not basing my statement on the PCA map, but the fst distances which are in accordance to the Haak graph.


page 26
http://biorxiv.org/content/biorxiv/early/2015/02/10/013433.full.pdf

I also think I know why he used the K6 calculator instead of K8 or K9. It would turn yamna more Central_West Asian like. Thats what even the people under his comment section are speculating.

Alan
14-03-15, 06:22
Another thing, didn't he give the word out and make everyone believe that Yamna is 40-45% WHG, 30-35% ANE and 25% ENF, while at the same time his own calculator spit out 35-40/30-35/25/5 (ANE/WHG/ENF/SE)?
In K8 North Indians appear with less than 3% WHG, suddenly in this K6 calculator they show up with extra WHG (additional to that what might have come from Yamna).

Fire Haired14
14-03-15, 06:39
Another thing, didn't he give the word out and make everyone believe that Yamna is 40-45% WHG, 30-35% ANE and 25% ENF, while at the same time his own calculator spit out 35/35/25/5 (ANE/WHG/ENF/SE)?

I don't feel like I can trust his word or work anymore.

No, that's what others said. It's basically true because the minor we don't know exactly what the minor exotic scores for Yamna mean.

Alan
14-03-15, 06:54
No, that's what others said. It's basically true because the minor we don't know exactly what the minor exotic scores for Yamna mean.

What did others said? That Yamna is 40-45% WHG and 30% ANE, while non Yamna sample even reaches 40% (mostly 35%) and some of the samples reach as high as 39% ANE.
So it's not true. You can't post "assumptions" as "confirmed" facts.

So if you open a thread with the title "Yamna 25 ENF , 30-35% ANE and 40-45%. Than it deceives people.

I actually thought he said that because you presented it as if this are the ultimate scores.


I feel exactly like the guy who asked this

@David, not to be rude but why are you doing something the researchers have already done?

Exactly why would you do the work a scientist already has done? Maybe you didn't like the results?

I simply have lost allot of trust in his work. He puts afford in it, I admit, but there is too often a too strong agenda behind his work.

Sile
14-03-15, 06:59
What did others said? That Yamna is 40-45% and 30% ANE, while non Yamna sample even reaches 40% (mostly 35%) and some of the samples reach as high as 39% ANE. So it's not true but more of a bluff. You can't post "assumptions" as "confirmed" facts.

I actually thought he said that because you presented it as if this are the ultimate scores.


I feel exactly like the guy who asked this


Exactly why would you do the work a scientist already has done? Maybe you didn't like the results?

I simply have lost trust in his work. He puts afford in it, I admit, but his work is too often too biased.

so he is doctoring his work

it seems the Haak paper has destroyed his pre-promises ( a year before ) that yamnya would be R1a ...........as you know seven R1b where found

Sile
14-03-15, 07:01
Alan, Polish weren't tested in Haak 2015. His numbers are very similar to Haak 2015's, they don't have to be identical to be correct.

You're basing your statement that Lezgins are as close to Yamna as Russians are on one PCA from Haak 2015, that's not nearly enough. I don't think Davidski forced an agenda into this test.

haak paper is about the ancients............there was no polish in the ancients.

Alan
14-03-15, 07:05
so he is doctoring his work

it seems the Haak paper has destroyed his pre-promises ( a year before ) that yamnya would be R1a ...........as you know seven R1b where found
++

At least someone seems to see through the mask. And now ask yourself again why suddenly Northeast Europeans are scoring more Yamna ancestry than anyone else in Europe. And West and South_Central Asians on average even less than the least Yamna (Spain) like European population.

David can do great work if it goes in accordence with what he expects and wishes. If he appears to be mistaken on something he does everything in his power to straighten things up as suitable as possible for his own theories and believes.

Fire Haired14
14-03-15, 07:19
Oh my God Alan just stop it.

Angela
14-03-15, 14:27
Where the results conflict with the Haak et al formal stats they're obviously wrong. I don't know how there can even be a debate about it. Even within the results themselves there are glaring inconsistencies. There's been an attempt to get rid of that 50% "Near Eastern" and any "Central Asian" in the Yamnaya Indo-Europeans and to move their center of gravity to certain areas of Europe ever since the results came out. It just took a while.

Anyone who knows anything at all about these programs and how they work knows that through the choice of K or the selection of references or the samples used you can manipulate the results. I don't know how it could have taken some of you so long to reach these conclusions.

If there are any actual academics involved in any of this, they should really do some research and re-examine their associations, or they are going to get burnt good and proper.

Oh, and for the record, I have no dog in the fight; the less Yamnaya my people and I turn out to have the better I'll like it.

giuseppe rossi
14-03-15, 15:47
People are getting mad at Polako! Hahahahaha

This software can't tell apart Yamna ancestry from indigenous Mesolitich stuff from La Brana swarthoids, so it's completely useless.

Aberdeen
14-03-15, 16:18
so he is doctoring his work

it seems the Haak paper has destroyed his pre-promises ( a year before ) that yamnya would be R1a ...........as you know seven R1b where found

Seven R1b were found in one location. I'm not saying he doesn't appear to have an agenda, but it seems quite certain to me that R1a was part of the IE expansion. I don't take that admixture stuff too clearly no matter who's doing it. I don't have to be a geneticist to know that it sometimes produces strange results and I think that just illustrates the old saying about any computer calculations: garbage in = garbage out.

Alan
14-03-15, 16:53
Seven R1b were found in one location. I'm not saying he doesn't appear to have an agenda, but it seems quite certain to me that R1a was part of the IE expansion. I don't take that admixture stuff too clearly no matter who's doing it. I don't have to be a geneticist to know that it sometimes produces strange results and I think that just illustrates the old saying about any computer calculations: garbage in = garbage out.

Surely R1a was part of the Yamna expansion. Without doubt for me because Reich said both existed there. Remember he has 50+ samples and only published 6 of them. Means to me some of the other might be R1a too and he is just holding it back for another study to be published.


Anyone who knows anything at all about these programs and how they work knows that through the choice of K or the selection of references or the samples used you can manipulate the results. I don't know how it could have taken some of you so long to reach these conclusions.


Gedrosia for example appears only in higher Ks. Now you can imagine yourself why the usage of higher Ks could be contraproductive for him. And the selection of reference populations is so true.

As I said he has actually great knowledge, but it is useless if you 24/7 use this knowledge to manipulate data and of course People who lose out in these manipulations will be upset most.

Alan
14-03-15, 17:13
And also great that in the Comment section people have realized that the Steppe population was replaced by "Polish"(Slavic) like people.

Goes in accordence with my theory that the original Pontic populations were almost completely replaced by the Slavic expansion(after beeing replaced by the Turkic earlier).

This is why we nowadays have this gab between, Eastern Europe, the Urals, Central Asia and North Caucasus. It's Altaic admixture in the East and Slavic admixture in the West.

Angela
14-03-15, 18:21
Just took a closer look at the scores for the ancient samples. While Oetzi is virtually 100% "Pre-Yamnaya", Stuttgart is only 75% "Pre-Yamnaya"? If she isn't "Pre-Yamnaya" then what is she? All those EEF samples are by definition "Pre-Yamnaya"! Then I thought that perhaps this cluster is supposed to represent Middle Neolithic people just prior to the arrival of the Yamnaya people, but only Oetzi is 100% Pre-Yamnaya, while the other MN sample suddenly has additional "Middle Eastern".

The whole thing makes so little sense that I'm done giving this "experiment" head room. If someone wants to know the best estimate so far for Yamnaya Indo-European ancestry in modern Europeans, the figures are given in Haak et al, although I think it will change as more ancient samples become available. You're not going to find the numbers in this run.

To be honest, the whole discussion of this topic is becoming a bit nausea inducing, and I say this as someone who has been following it as an intellectual puzzle for years. At the end of the day, what does it matter who has what percent of Yamnaya ancestry? What does any of this have to do with real life or the real issues confronting all of us?

Fire Haired14
14-03-15, 19:37
Where the results conflict with the Haak et al formal stats they're obviously wrong. I don't know how there can even be a debate about it. Even within the results themselves there are glaring inconsistencies. There's been an attempt to get rid of that 50% "Near Eastern" and any "Central Asian" in the Yamnaya Indo-Europeans and to move their center of gravity to certain areas of Europe ever since the results came out. It just took a while.

Everyone who consistently visits Eurogenes blog should know that Yamna-related, pre-Yamna, and Middle Easyerm all carry ENF. Davidski is not trying to make Yamna appear non-ENF just like 23andme isn't trying to make people appear non-ENF by giving percentages in German-French, etc. BTW, the Central Asian component from his K9 is pretty much pure ANE.

This test isn't meant to measure differnt forms of ancestry in Yamna it's meant to measure Yamna-related ancestry in modern people.


Anyone who knows anything at all about these programs and how they work knows that through the choice of K or the selection of references or the samples used you can manipulate the results. I don't know how it could have taken some of you so long to reach these conclusions.

If there are any actual academics involved in any of this, they should really do some research and re-examine their associations, or they are going to get burnt good and proper.

Oh, and for the record, I have no dog in the fight; the less Yamnaya my people and I turn out to have the better I'll like it.

You're saying no one should use Ks?

Fire Haired14
14-03-15, 19:40
Just took a closer look at the scores for the ancient samples. While Oetzi is virtually 100% "Pre-Yamnaya", Stuttgart is only 75% "Pre-Yamnaya"? If she isn't "Pre-Yamnaya" then what is she? All those EEF samples are by definition "Pre-Yamnaya"! Then I thought that perhaps this cluster is supposed to represent Middle Neolithic people just prior to the arrival of the Yamnaya people, but only Oetzi is 100% Pre-Yamnaya, while the other MN sample suddenly has additional "Middle Eastern".

The whole thing makes so little sense that I'm done giving this "experiment" head room. If someone wants to know the best estimate so far for Yamnaya Indo-European ancestry in modern Europeans, the figures are given in Haak et al, although I think it will change as more ancient samples become available. You're not going to find the numbers in this run.

To be honest, the whole discussion of this topic is becoming a bit nausea inducing, and I say this as someone who has been following it as an intellectual puzzle for years. At the end of the day, what does it matter who has what percent of Yamnaya ancestry? What does any of this have to do with real life or the real issues confronting all of us?

Pre-Yamna, Yamna, etc. all the components in this test are simply components just like "West-Med" in K15. Davidski gave them those names because they best represent Yamna, pre-Yamna, etc. ancestry. When he added ancient genomes in his database they created components modern ones couldn't, because they better represent the ancestors of modern people than other modern people.

No one is(or at least should) taking this as a competition to see who's the most Yamna-descended. That was never the point. Yamna-like people passed on a lot of genes so when researching modern origins they're very much in subject.

Fire Haired14
14-03-15, 19:46
And also great that in the Comment section people have realized that the Steppe population was replaced by "Polish"(Slavic) like people.

Goes in accordence with my theory that the original Pontic populations were almost completely replaced by the Slavic expansion(after beeing replaced by the Turkic earlier).

This is why we nowadays have this gab between, Eastern Europe, the Urals and North Caucasus.

Yeah, that does appear to be the case. Central Asians look like a mix of Yamna and Siberians, Ural people look like a mix of Yamna, Siberians, and Poles. Obviously Yamna was there before the other two.

Greying Wanderer
14-03-15, 22:21
Seems to me there is obviously a flaw in the three way model: EEF, WHG, ANE.

1) There's at least two sets of farmers, a more maritime Levant set and another more overland set and the "near eastern" component may be a composite of the two rather than a single component.

2) There seems to be multiple sources of ANE.

So whether his alternative is ultimately right or wrong Davidski's tinkering with the model is likely to produce something useful.

Sile
14-03-15, 23:59
So whether his alternative is ultimately right or wrong Davidski's tinkering with the model is likely to produce something useful.

sure can, toilet paper is useful too

Greying Wanderer
15-03-15, 00:25
sure can, toilet paper is useful too

If there are more than three critical components then any model using just the three components is going to be flawed.

Alan
15-03-15, 01:23
Just took a closer look at the scores for the ancient samples. While Oetzi is virtually 100% "Pre-Yamnaya", Stuttgart is only 75% "Pre-Yamnaya"? If she isn't "Pre-Yamnaya" then what is she? All those EEF samples are by definition "Pre-Yamnaya"! Then I thought that perhaps this cluster is supposed to represent Middle Neolithic people just prior to the arrival of the Yamnaya people, but only Oetzi is 100% Pre-Yamnaya, while the other MN sample suddenly has additional "Middle Eastern".

The whole thing makes so little sense that I'm done giving this "experiment" head room. If someone wants to know the best estimate so far for Yamnaya Indo-European ancestry in modern Europeans, the figures are given in Haak et al, although I think it will change as more ancient samples become available. You're not going to find the numbers in this run.

To be honest, the whole discussion of this topic is becoming a bit nausea inducing, and I say this as someone who has been following it as an intellectual puzzle for years. At the end of the day, what does it matter who has what percent of Yamnaya ancestry? What does any of this have to do with real life or the real issues confronting all of us?

When a pre Yamna Stuttgart turns out 25% Yamna and the general data is not consistent with the actual study than you know something is wrong here.
As I said K6 is far too low to even be able to differentiate between farmer and H&G let alone the various farmer and H&G types. K6 only differentiates between "European, Near Eastern, South Asian, Amerindian, East Asian and African." And this is also based on the reference populations you use for it. I assume this might be the reason why some of the EEF genes turn out as Yamna because it can't differentiate between EEF and "West Asian" and probably can't differentiate between pre and post Yamna WHG like ancestry.

This comes not suprisingly accommodating for Northeast Europeans who lack Gedrosia.
For example take a look at the Balochi with highest Gedrosia scores. They appear to have only 5%!! Yamna ancestry while anyone surrounding them has suddenly 40%. It almost seems like something "contra West Asian/Gedrosia" is going on in this calculator. And suddenly Northeast Europeans have the top Yamna scores instead of Mordovians, Norwegians, Russians, Lezgians (only Udmurts top them), while strictly taken we could argue that Northeast Europeans have "least" Yamna ancestry because they lack Gedrosia/West Asian which is crucial component among Yamna. This might be the reason why he needed to eradicate this component by using lower Ks

In order to make Northeast Europeans as "Yamna" like as possible you need to use lower Ks where the components are not specified enough. Even the Haak paper uses at least K16
page 22 => http://biorxiv.org/content/biorxiv/early/2015/02/10/013433.full.pdf

So how comes someone could thing he can get correct results from a K6 run? Look it can happen that sometimes results get inconsistent I understand that, but I can't believe this is coincidence if it is always Northeast Europeans profiting from this on the cost of Western Asians.

I am no expert on this field so I can't tell exactly what is wrong, but comparing it with Haak paper, there is definitely something wrong.

Alan
15-03-15, 01:31
If there are more than three critical components then any model using just the three components is going to be flawed.

If the results are that significantly different from the actual Haak paper, than this calculator is of not much use.

giuseppe rossi
15-03-15, 12:09
I am honest. The Yamna component from this run is identical to the North Euro component of Dodecad k12b.

I bet that Loschbour would score 80% Yamna on this run.

Which leads to a simple conclusion: peer reviewed studies are still much superior to amateur genome bloggers.

John Doe
15-03-15, 14:16
I think he sent me my results:
Yamnaya_related 0.148505
WHG_extra 0.007126
ENA 0.003979
Middle_Eastern 0.415615
Pre-Yamnaya 0.411474
Sub-Saharan 0.0133

Apparently Pre-Yamnaya is supposed to be ENF with some Mesolithic admixture, and the Middle east is how it was when ENF didn't get to Europe yet. Yamnaya component apparently reaches the highest frequency among the Udmurt of northern Russia.

Aberdeen
15-03-15, 14:38
If Corded Ware, northern Yamnaya and Slavic are all R1a Russian hunter gatherer mixed with some source for an early Middle Eastern farmer component but that Middle Eastern component got mixed in with the R1a from different sources (Neolithic European farmer, Caucasian and Cucutei-Trypellian), it might be difficult to tell one from the other in modern populations, I think. And we don't yet know how different R1a from the Yamnaya horizon looks autosomally from R1b from the Yamnaya horizon. I think that leaves lots of room for confusing results that might fit some people's agendas, even without deliberate manipulation of data.

John Doe
15-03-15, 14:43
If Corded Ware, northern Yamnaya and Slavic are all R1a Russian hunter gatherer mixed with some source for an early Middle Eastern farmer component but that Middle Eastern component got mixed in with the R1a from different sources (Neolithic European farmer, Caucasian and Cucutei-Trypellian), it might be difficult to tell one from the other in modern populations, I think. And we don't yet know how different R1a from the Yamnaya horizon looks autosomally from R1b from the Yamnaya horizon. I think that leaves lots of room for confusing results that might fit some people's agendas, even without deliberate manipulation of data.

Good point, studies, especially those that aren't peer reviewed, should never be taken as gospel.

Angela
15-03-15, 16:31
Another thing, didn't he give the word out and make everyone believe that Yamna is 40-45% WHG, 30-35% ANE and 25% ENF, while at the same time his own calculator spit out 35-40/30-35/25/5 (ANE/WHG/ENF/SE)?
In K8 North Indians appear with less than 3% WHG, suddenly in this K6 calculator they show up with extra WHG (additional to that what might have come from Yamna).

Is it a coincidence that Patterson stated that the fact that some populations speaking Indo-European languages have virtually no WHG raises some questions as to whether all the Indo-European languages spread from the Pontic-Caspian steppe? :smile: Perhaps a pre-emptive strike?

Just so it's clear, I'm still an agnostic about all of this.

Angela
15-03-15, 16:45
Fire-Haired:You're saying no one should use Ks?

I said no such thing. You know better than that, Fire-Haired. I meant, as I'm sure you understand, that depending on the nature of each run, you get clarity with X number of "K", and you can tell the optimum "K" level by a statistical analysis of the data.


Pre-Yamna, Yamna, etc. all the components in this test are simply components just like "West-Med" in K15. Davidski gave them those names because they best represent Yamna, pre-Yamna, etc. ancestry. When he added ancient genomes in his database they created components modern ones couldn't, because they better represent the ancestors of modern people than other modern people.

No one is(or at least should) taking this as a competition to see who's the most Yamna-descended. That was never the point. Yamna-like people passed on a lot of genes so when researching modern origins they're very much in subject.

No, they don't best represent these components, because the data is internally inconsistent.

As to your latter statement, I will pay you the compliment of believing you are just still naive, and have also not been reading blog posts by the "usual suspects" for the last ten years. Everything wasn't always so PC, you know. It's true that many of those blogs either crashed or have been deliberately "cleansed", but a lot of people took an awful lot of screenshots. Someday they'll surface and the ****will hit the fan, and it won't just be about wanting certain populations to have a large Yamnaya Indo-European component. When that happens, academics and professional people with a reputation to lose are going to pay a price for their associations if they can be linked to some of these anonymous "internet names".


Greying Wanderer: Seems to me there is obviously a flaw in the three way model: EEF, WHG, ANE.

1) There's at least two sets of farmers, a more maritime Levant set and another more overland set and the "near eastern" component may be a composite of the two rather than a single component.

There is indeed something wrong with the EEF, WHG, ANE model, which is why it's clear the Reich Lab has quietly dropped it. The Mal'ta sample is either too old, or too poor quality, or just not very informative. It's clear, for example, the EHG can't really be modeled as an actual mixture of WHG and ANE according to formal stats.

As to point number one, all of the EEF are very homogenous, which is another clue that this K6 is not correct or helpful. That's clear from every academic paper that has dealt with the subject. Theirs was a maritime expansion originating somewhere around northern Syria/southeast Anatolia some of whose people followed the littoral of the northern Mediterranean, and some of whom went inland after reaching Greece. That information is nicely backed up by the findings of Paschou et al.

The "Near Eastern" component, in so far as I can tell from the data so far is probably vast majority the same element, which probably traveled north from the center of gravity to the Caucasus, Iran, Central Asia and India. I've posted numerous maps of the movement of the Neolithic to those areas. In addition, there's some sort of older "Central Asian/Sindh" element, for lack of a better description, which is a minority component.

Alan
15-03-15, 17:04
Is it a coincidence that Patterson stated that the fact that some populations speaking Indo-European languages have virtually no WHG raises some questions as to whether all the Indo-European languages spread from the Pontic-Caspian steppe? Perhaps a pre-emptive strike? :smile:

Just so it's clear, I'm still an agnostic about all of this.

Therefore it is hard for me to believe Yamna is some kind of PIE Urheimat. PIEs seem more like an ethno_cultural complex.

Alan
15-03-15, 17:12
I am honest. The Yamna component from this run is identical to the North Euro component of Dodecad k12b.

I bet that Loschbour would score 80% Yamna on this run.

Which leads to a simple conclusion: peer reviewed studies are still much superior to amateur genome bloggers.

Now that you say it. Balochi have most Gedrosia and least North European compared to their neighbors in east. And they score only 5% Yamna while all the eastern neighbors with higher North European component score suddenly 25-40%. It seems like in this calculator "North European"(mostly WHG) is crucial.

That gets also obvious if you take a look at the Yamna scores in West Asia. It seems to go hand in hand with the North Euro component. West Asian have the least North Euro and therefore least Yamna, Europeans the most North Euro and therefore Yamna and Central Asians something in between.
So Yamna = pred Northeast Euro??

Nevermind just a thought.

Whatever it is, there are obvious flaws.

Greying Wanderer
15-03-15, 17:51
There is indeed something wrong with the EEF, WHG, ANE model, which is why it's clear the Reich Lab has quietly dropped it. The Mal'ta sample is either too old, or too poor quality, or just not very informative. It's clear, for example, the EHG can't really be modeled as an actual mixture of WHG and ANE according to formal stats.

As to point number one, all of the EEF are very homogenous, which is another clue that this K6 is not correct or helpful. That's clear from every academic paper that has dealt with the subject. Theirs was a maritime expansion originating somewhere around northern Syria/southeast Anatolia some of whose people followed the littoral of the northern Mediterranean, and some of whom went inland after reaching Greece. That information is nicely backed up by the findings of Paschou et al.

The "Near Eastern" component, in so far as I can tell from the data so far is probably vast majority the same element, which probably traveled north from the center of gravity to the Caucasus, Iran, Central Asia and India. I've posted numerous maps of the movement of the Neolithic to those areas. In addition, there's some sort of older "Central Asian/Sindh" element, for lack of a better description, which is a minority component.

Maybe. I think the earlier more maritime Levant expansion mostly merged with a second more overland one that originated further east but there may be a few places they didn't merge - maybe islands like Sardinia - and finding the split may be helpful like finding one of the corner pieces of a jigsaw puzzle.

Aberdeen
15-03-15, 17:53
Therefore it is hard for me to believe Yamna is some kind of PIE Urheimat. PIEs seem more like an ethno_cultural complex.

For the umpteenth time, I'd like to point out that Yamnaya could have been the ethno-cultural complex that produced the Bronze Age Indo-European expansion, as indicated by the archeological evidence, even if the Proto-IE language may have developed elsewhere among one of the probably two major components that probably combined to create Yamnaya.

Alan
15-03-15, 17:55
As to your latter statement, I will pay you the compliment of believing you are just still naive, and have also not been reading blog posts by the "usual suspects" for the last ten years. Everything wasn't always so PC, you know. It's true that many of those blogs either crashed or have been deliberately "cleansed", but a lot of people took an awful lot of screenshots. Someday they'll surface and the ****will hit the fan, and it won't just be about wanting certain populations to have a large Yamnaya Indo-European component. When that happens, academics and professional people with a reputation to lose are going to pay a price for their associations if they can be linked to some of these anonymous "internet names".




Indeed it seems like he doesn't know the history of some of these bloggers. Only this can explain why he doesn't see any bias in it. The blogger might have changed the tone and became more "open" to other theories. But no real live person can change his sentiments within a few years by 100%. I remember times when the suspect was claiming Scythians came straight out of the Baltics.

Of course he wouldn't throw out these kind of theories. Anyone makes mistakes so I don't really hold on the "mistakes of the past". But it made me cautious to believe anything what they say.

I am long enough around to know the background of allot of these bloggers.


In addition, there's some sort of older "Central Asian/Sindh" element, for lack of a better description, which is a minority component.

Kalash specific DNA probably. David calls it "South Eurasian".

Greying Wanderer
15-03-15, 18:33
If a population X expanded from somewhere in Central Asia (for example) into Iran, India, Near East and Europe and merged with whoever was there already they'd create a variety of X + something else populations (possibly with the y dna coming mostly from the X population).

If the same X population moved onto but were partially repulsed from the steppe (cos horses) that might also lead to the resulting steppe population being an X + something else population also (although possibly with the y dna coming mostly from the non X population).

If so then if the steppe population later expanded out into the surrounding regions you'd have an X + something else population expanding over a variety of X + something else populations.

In which case it might be difficult to separate the components autosomally but the proportions of X to non-X y haplogroups might give a clue to the proportions.

Sile
15-03-15, 19:58
I think he sent me my results:
Yamnaya_related 0.148505
WHG_extra 0.007126
ENA 0.003979
Middle_Eastern 0.415615
Pre-Yamnaya 0.411474
Sub-Saharan 0.0133

Apparently Pre-Yamnaya is supposed to be ENF with some Mesolithic admixture, and the Middle east is how it was when ENF didn't get to Europe yet. Yamnaya component apparently reaches the highest frequency among the Udmurt of northern Russia.

is he correct!

pre-yamnya + yamnya = 54.9%

and you are E-M35 haplogroup

he said there was no E in yamnya

Fire Haired14
15-03-15, 20:24
Angela, you clearly don't like Davidski for some reason which is why you're biasedly against Yamna K6. Ancient genetics don't have to give people headaches because of how complicated they're. Some things in life are simple. I don't really see a reason to argue with you about this.

The Ancient genomes from haak 2015 are such good proxies of modern peoples' ancestors that components were created that nearly perfectly represent Middle Neolithic Euros, Yamna, and WHG-survival. The ancient individuals don't have to score 100% in a component for the test to be accurate.

Fire Haired14
15-03-15, 20:25
is he correct!

pre-yamnya + yamnya = 54.9%

and you are E-M35 haplogroup

he said there was no E in yamnya

There are some African-Americans with British Y DNA but with over 70% West African ancestry. Does this mean British Y DNA originated in west Africa? No.

Fire Haired14
15-03-15, 20:29
I think he sent me my results:
Yamnaya_related 0.148505
WHG_extra 0.007126
ENA 0.003979
Middle_Eastern 0.415615
Pre-Yamnaya 0.411474
Sub-Saharan 0.0133

Apparently Pre-Yamnaya is supposed to be ENF with some Mesolithic admixture, and the Middle east is how it was when ENF didn't get to Europe yet. Yamnaya component apparently reaches the highest frequency among the Udmurt of northern Russia.

The most interesting component in this test to me is the Middle Eastern one. It isn't random that even though Jews and Sardinians have similar amounts of ENF-WHG that Sardinians score ~20% in Middle Eastern and Jews score ~40%. This component is doing a pretty good job singling out recent Middle eastern ancestry vs Neolithic Middle eastern ancestry.

Fire Haired14
15-03-15, 20:35
Here are my Yamna K6 results. Yamna scored ~33-38% ANE in K8 and I scored ~12-13%, so just over 1/3 Yamna-like(assuming most of my ANE is of Steppe-origin), consistent with my 36% Yamna-related score.

Yamnaya_related 36.5151
WHG_extra 2.436
ENA 1.7989
Middle_Eastern 11.3951
Pre-Yamnaya 46.883
Sub-Saharan 0.9719

Angela
15-03-15, 21:02
Angela, you clearly don't like Davidski for some reason which is why you're biasedly against Yamna K6. Ancient genetics don't have to give people headaches because of how complicated they're. Some things in life are simple. I don't really see a reason to argue with you about this.

The Ancient genomes from haak 2015 are such good proxies of modern peoples' ancestors that components were created that nearly perfectly represent Middle Neolithic Euros, Yamna, and WHG-survival. The ancient individuals don't have to score 100% in a component for the test to be accurate.

You're very quick to ascribe motives to person X but not to person Y. Why is that I wonder?

I'm not so petty, Fire-Haired. I don't know any of these bloggers from a hole in the wall...haven't exchanged a single personal message with any of them, nor do I wish to do so. I know of them only from their own posts. What I dislike in general in people is racism and lack of integrity.

You'll hopefully learn some cynicism sooner rather than later, for your own good. Let's leave it at that, shall we?

Ed. Well, I dislike a lot of things about human beings, but it would take too long to go over all of them, and this isn't the place to list them.:smile:

Alan
15-03-15, 23:49
You're very quick to ascribe motives to person X but not to person Y. Why is that I wonder?


t's not hard to tell why. If you know his sentiments you would realize that what David and Genetiker claim go in accordence what he wants to believe. People will easier trust people they want to believe, than people who actually tell how it is. That is human.

MOESAN
16-03-15, 00:50
Angela, I found very interesting to know the true proportions of Yamnaya we could all have -for History it has some weight - the problem is how to know that, because in past we had surely common ancestors with Yamnaya without Yamanya or before Yamnaya arrival - the question remaining : agendas and spirit blindness -
all that saud, yes these poolings questions even well made can abuse us about real past events - but I like put my brain at work for the pleasure and without drama -

Fire Haired14
16-03-15, 00:52
not post....