An attempt at deep West Eurasian phylogeny

[TanTin]

f3-outgroup with Bacho Kiro and Loschbour included  and some modern populations.

f3=f3("v52.2_HO_public",p1,p2,p2)

[Kale]

Please be patient and more tolerant all of you guys.
Kale is the one who started this discussion. We have it since AG forum, then it was in the previos version before we moved here.  So let Kale to decide who should be moderated and who may contribute.  G25 allows us to do some visualization. I can do similar visualization by using other tools.  So I have some idea what archi is trying to show us, but the way he is doing it is not much satisfactory. So I understand PopGenist82   too.

Yes, patience is important, but like you said, it's not what Archi is doing, it's completely fine to post things you think are correct, even if they turn out not to be, it's his attitude, which I think everyone has exhausted their patience for.

IMO the affinity lies between Dzudzuana & Bacho MUP, via a 'West Eurasian' group which is parallel to but not from Kostenki/ Sunghir.
In other words, the recently hypothesised 'proto-Gravettian migration' is not from Eastern Europe, but southeastern Europe. 

Bacho MUP then contributes to Vestonice Gravettians, as do the Kostenki/Sunghir group.

add. something like this

That's a rather interesting graph. Kostenki/Sunghir have no special relation other than lacking the IUP admix.
I'd be interested to see how Muierii fits in there, if it doesn't break that could prove a promising option.

[AimSmall]

DNA archi & PopGenist82:

Both of you please chill out.  We won't be swinging the ban hammer as freely as AG would, but don't push it here.

I've seen no trolling, only bickering.  If you don't like each other's posts, ignore them.

Let's get back to the research and enjoy why we come together here.

[PopGenist82]

DNA archi & PopGenist82:

Both of you please chill out.  We won't be swinging the ban hammer as freely as AG would, but don't push it here.

I've seen no trolling, only bickering.  If you don't like each other's posts, ignore them.

Let's get back to the research and enjoy why we come together here.

Ok, but its not really bickering, I’d call it the educated suffering the insolence of the malignantly ignorant.

[ph2ter]

DNA archi & PopGenist82:

Both of you please chill out.  We won't be swinging the ban hammer as freely as AG would, but don't push it here.

I've seen no trolling, only bickering.  If you don't like each other's posts, ignore them.

Let's get back to the research and enjoy why we come together here.

As we can see, this will not end in peace.
We have been protected at AG from such Internet personalities.
Such kind of negative behaviour we cannot call just bickering.
It will destroy any meaningful discussion and many users will leave this forum because of that.

We have such type of forums already. We don't need this forum to be the copy of the Apricity or Eupedia.

[Rober_tce]

As AimSmall has said, we should moderate the tone of conversation. This thread is being monitored.

[TanTin]

That's a rather interesting graph. Kostenki/Sunghir have no special relation other than lacking the IUP admix.
I'd be interested to see how Muierii fits in there, if it doesn't break that could prove a promising option.

No idea how he did it, but Kostenki has been always near Sunghir and Muierii  (PM) is usually near BK-1653  - late Bacho Kiro.
So far these are the very first early Eurasians. PM - Muierii  and BK-1653 precede WHG (Loschbour) . Now we have plenty of WHG and their ancestors.



As I remember, I generated this image from G25 data.
I have also my own PCA data, so I already generated many instances of very similar configurations. Depending on the individuals included the configuration may change slightly from one dataset to another. Anyone may reproduce it . For the best visualization you just need to include the principal component for WHG and use the 3D projection.

[Norfern-Ostrobothnian]

I think it's caused by a lack of a shared edge between Belgium UP GoyetQ116-1 and lack of admixture into BK1653 from BK IUP. Also I think NEO283 needs a West Eurasian clade that split off before all other West Eurasians

[Kale]

Another observation when running graphs.
If allowing ZlatyKun to be admixed, it can be a mix of ~35-40% Basal Eurasian proper (sharing a non-zero drift branch with the basal throughout the Near East), and the remaining is generic Eurasian, Ust-Ishim-like. At that point, instead of having a pure basal edge into Kostenki-cluster, they can have actual ZlatyKun mixture (with minor shared drift, 1-2 units).

That would raise an interesting proposition for tracing the migration path of y-hg IJ, as it removes one of the potential connections between hg-I bearing populations and hg-J bearing populations (a post 42kbp basal introgression from Near East into Kostenki-cluster). Either J would enter into Near East via Muierii > AHG/Georgia_UP, which is suspect by virtue of lack of J in Anatolia, or the other possibility, J is the Western lineage of ANE. J2 takes Central Asian route into Iran_N, J1 takes Northwest route to EHG > CHG.

[TanTin]

Oase-1 may be contaminated, but why was Salkhit left out? Both may be relevant to the origins of ANE.

Another observation when running graphs.
If allowing ZlatyKun to be admixed, it can be a mix of ~35-40% Basal Eurasian proper (sharing a non-zero drift branch with the basal throughout the Near East), and the remaining is generic Eurasian, Ust-Ishim-like. At that point, instead of having a pure basal edge into Kostenki-cluster, they can have actual ZlatyKun mixture (with minor shared drift, 1-2 units).

That would raise an interesting proposition for tracing the migration path of y-hg IJ, as it removes one of the potential connections between hg-I bearing populations and hg-J bearing populations (a post 42kbp basal introgression from Near East into Kostenki-cluster). Either J would enter into Near East via Muierii > AHG/Georgia_UP, which is suspect by virtue of lack of J in Anatolia, or the other possibility, J is the Western lineage of ANE. J2 takes Central Asian route into Iran_N, J1 takes Northwest route to EHG > CHG.

For the moment  ZlatyKun is the most  Basal Eurasian proper. So far, there are 2 or 3 branches proven with a number of samples:
East Asia /Tianyuan/
West Europe / Sunghir-Kostenki - WHG/
- Another branch is going to form for Central Europe and Anatholia,  represented by Dzudzuana for the moment. 

- EHG could be considered as another main branch connected to the  Basal Eurasian.

We have also CHG, however I think they are not the best representative for this branch. There is a branch going to India-Pakistan and this branch is huge one, however we don't have many ancient samples . CHG looks to me some mixture with this SouthAsia branch.  I think CHG are a subbranch from the India-Pakistan main branch.  

We can show 3 trees:  by Y , mtDNA and autosomal .   All the trees must show us the same or very similar configuration.  However the first 2 trees are pure trees, no internal mixtures when the 3-rd autosomal tree is a mixture between the first 2.

[Woz]

Y'all are trying to trace the phylogeny of Eurasians without any Middle Eastern UP samples (except for the relatively unhelpful Dzudzuana). Seems like an excercise in futility.

Either J would enter into Near East via Muierii > AHG/Georgia_UP, which is suspect by virtue of lack of J in Anatolia, or the other possibility, J is the Western lineage of ANE. J2 takes Central Asian route into Iran_N, J1 takes Northwest route to EHG > CHG.

What makes you think that Anatolia was NOT occupied in the UP by the J's? Or I's? Anatolia seems to have been largely depopulated during the LGM. Pinarbasi might be a purely Mesolithic population with no deeper roots in Anatolia, much like WHG's in most of Europe.

[PopGenist82]

Y'all are trying to trace the phylogeny of Eurasians without any Middle Eastern UP samples (except for the relatively unhelpful Dzudzuana). Seems like an excercise in futility.

The utility of qpGraph is its ability to infer deep histories even with unsampled phases. The fact that you find Dzudzuana (not to mention Kotias UP) “uninformative” is quite odd. 

What makes you think that Anatolia was NOT occupied in the UP by the J's? Or I's? Anatolia seems to have been largely depopulated during the LGM. Pinarbasi might be a purely Mesolithic population with no deeper roots in Anatolia,

Anatolia was also uninhabited before the LGM, with no Gravettian sites, , so your “full of I-J” theory doesn’t work.
Pinarbasi ~ 15000 calBP is properly called Epipaleolithic, or post-LGM 

much like WHG's in most of Europe.

Obviously not true, given that WHG draw much of their ancestry from Vestonice, and there is a clear link of the 2 via Yhg I, mtDNA U5

[Kale]

Y'all are trying to trace the phylogeny of Eurasians without any Middle Eastern UP samples (except for the relatively unhelpful Dzudzuana). Seems like an excercise in futility.

Either J would enter into Near East via Muierii > AHG/Georgia_UP, which is suspect by virtue of lack of J in Anatolia, or the other possibility, J is the Western lineage of ANE. J2 takes Central Asian route into Iran_N, J1 takes Northwest route to EHG > CHG.

What makes you think that Anatolia was NOT occupied in the UP by the J's? Or I's? Anatolia seems to have been largely depopulated during the LGM. Pinarbasi might be a purely Mesolithic population with no deeper roots in Anatolia, much like WHG's in most of Europe.

Good thing futile is a subjective term. 
Your point about Anatolia seems a bit redundant... if pre-LGM Anatolians were a dead end, they could have been martians for all it matters, and had any lineage you'd like. 
Pinarbasi and near relatives, who play an important role in the graph on the other hand, are probably not likely to have ever had much y-hg J, based upon Neolithic Anatolians (and European descendents), whom the sample count has become astronomical, yet the J count is relatively low and I'm guessing sitted well within deeper Mesopotamian/Iranian variation. Of course the time between the split of IJ and the first branching of J is 15,000+ years so that's a lot of time and consequently a lot of space for it to have potentially been, but Anatolia, or more precisely populations genetically resembling Pinarbasi/Georgia_UP, would not be the top of the list.

[Jerome]

Y'all are trying to trace the phylogeny of Eurasians without any Middle Eastern UP samples (except for the relatively unhelpful Dzudzuana). Seems like an excercise in futility.

Either J would enter into Near East via Muierii > AHG/Georgia_UP, which is suspect by virtue of lack of J in Anatolia, or the other possibility, J is the Western lineage of ANE. J2 takes Central Asian route into Iran_N, J1 takes Northwest route to EHG > CHG.

What makes you think that Anatolia was NOT occupied in the UP by the J's? Or I's? Anatolia seems to have been largely depopulated during the LGM. Pinarbasi might be a purely Mesolithic population with no deeper roots in Anatolia, much like WHG's in most of Europe.

You are correct.
There's the Kairan B site in Anatolia,occupied from 40k bc to LGM,I think that would be a good place to search for CWE/WHG.

It's clear that WHG isnt just post LGM 'post LGM bottlenecked drifted gravettian'.
Rather it's gravettians that are a mix of something CWE/WHG and aurignacian kind of Ancestry.
(Fu et al modelling)

The F4 runs by norfern ostrobothnia on kotias UP have already shown that,that Kotias Up can't just be gravettian/muierii and have something different,more related to WHG than to gravettians.
And dzudzuana is 5000 years before LGM and 10,000 years before WHG samples in europe and it already has WHG kind of Ancestry,much before.


I think pre LGM UP samples from Aegean/Bulgaria/Balkans will reveal WHGs without the need to model them as bottlenecked gravettians.

WHG/CWE kind of Ancestry arrived much earlier,around 30,000 BC,at this see you start seeing I in vestonice/krems and southern Balkans/aegean from this time will reveal WHGs.


As for I in Anatolia,Ftdna has shown that the C1a2 in pinarbasi is downstream of vestonice C1a2,so it could be possible that pinarbasi lost their 'I' to gravettian seafarers.

[Jerome]

I feel people are drawing conclusions too early without any samples from UP middle East.
According to them there were only 1 kind of west eurasian Ancestry, aurignacian-kostenki,which transformed into ANE,WHG,Gravettians and even reached middle East to form dzudzuana.

Compare east eurasia,where we know that many kinds of east Eurasian groups like australo-melanisan,Tibetan HG, AASI,east asia branch(tianyuan,jomon) existed.
It would be naive to think that West Eurasia was different and only 1 one kind of west eurasian group existed.

Many people also think that middle east was only populated by basal eurasians and West eurasian Ancestry in middle East came from europe, which is tremendously wrong.
If middle East was just populated by basal eurasians,how do they explain Haplogroups G-M201,H-P69 in middle east?
These are clearly west eurasian Haplogroups and formed around 47,000 BC when GHIJK started to differentiate and west and east eurasians started to split from each other.
These haplogroups aren't found in any UP european sample and so they can't be from there either.

Bottomline, people should wait for ancient samples before drawing conclusions.
Everybody would think that Taforalt had European UP Ancestry from aurignacians,if we didn't have the pinarbasi, natufian and dzudzuana samples.
Similar case is happening again, people are just force-fitting different kinds of Ancestry without any proximal samples.