Basal Eurasian discussion

[Jerome]

Can you model Pinarbasi with the same model? I want to see if they have the same Onge related ancestry since they can be modeled with 30% Zarzian ancestry in your previous model

That model was using Iron gates+WHG to model Pinarbasi so Iran_N may simply be getting inflated to compensate for some eastern Ancestry.
A better model would for Pinarbasi would be Iron gates/WHG+mbuti/mota (for basal proxy)+Tepe_AbdulHosein,this would show how much real IranN/Zarzian is in Pinarbasi,if there is any.

There is some evidence of Zarzian lithics in the anti-taurus region,so Pinarbasi's ancestors could have simply picked some Up when they migrated from North Levant.

Keep in Mind epipaleolithic Anatolia is derived from Levant,not directly from Dzudzuana.

Almarri et al 2021 put the split between Sardinians(ANF) and Levantines at 16K BC.
The first anatolian epipaleolithic industries appear around 14,500 BC(Karain B,Okuzini) from all the studies I have read on epipaleolithic Anatolia and all of them show strong connections in lithics and foraging methods to Kebarans/Pre-Natufian Levant, especially the ucakzigil kebarans of North Levant.
Pinarbasi industry itself dates to 14,000 BC.

The original people of western Anatolia may well have been WHG types considering that we find C1a2 and I2a in Anatolia all though post LGM Anatolia seems to be very sparsely populated with not much evidence of human inhabitation until around 14,000 BC when Kebaran connection starts showing up.
Just 1000 years later Southern Levantines get mushabian/IBM influx and become different from North Levantines and anatolians.

I think Western Aegean/Greece and southern balkans are likely where WHG lived in and expanded from After LGM.
The early industries of these regions are different and more connected to epigravettian.
Then around the Holocene/12-10,000 BC we start seeing the Theopetra Caves and other similar sites,which show more connection to Anatolia.
This also fits in well with the anatolian ancestry in iron gates.

[old europe]

Can you model Pinarbasi with the same model? I want to see if they have the same Onge related ancestry since they can be modeled with 30% Zarzian ancestry in your previous model

That model was using Iron gates+WHG to model Pinarbasi so Iran_N may simply be getting inflated to compensate for some eastern Ancestry.
A better model would for Pinarbasi would be Iron gates/WHG+mbuti/mota (for basal proxy)+Tepe_AbdulHosein,this would show how much real IranN/Zarzian is in Pinarbasi,if there is any.

There is some evidence of Zarzian lithics in the anti-taurus region,so Pinarbasi's ancestors could have simply picked some Up when they migrated from North Levant.

Keep in Mind epipaleolithic Anatolia is derived from Levant,not directly from Dzudzuana.

Almarri et al 2021 put the split between Sardinians(ANF) and Levantines at 16K BC.
The first anatolian epipaleolithic industries appear around 14,500 BC(Karain B,Okuzini) from all the studies I have read on epipaleolithic Anatolia and all of them show strong connections in lithics and foraging methods to Kebarans/Pre-Natufian Levant, especially the ucakzigil kebarans of North Levant.
Pinarbasi industry itself dates to 14,000 BC.

The original people of western Anatolia may well have been WHG types considering that we find C1a2 and I2a in Anatolia all though post LGM Anatolia seems to be very sparsely populated with not much evidence of human inhabitation until around 14,000 BC when Kebaran connection starts showing up.
Just 1000 years later Southern Levantines get mushabian/IBM influx and become different from North Levantines and anatolians.

I think Western Aegean/Greece and southern balkans are likely where WHG lived in and expanded from After LGM.
The early industries of these regions are different and more connected to epigravettian.
Then around the Holocene/12-10,000 BC we start seeing the Theopetra Caves and other similar sites,which show more connection to Anatolia.
This also fits in well with the anatolian ancestry in iron gates.

Good post
some questions

1) Do you think the WHG like dna in the western aegean region ultimately is rooted in the Muierii/ BK 1653 genetic cluster?
2) The Mushabian/ IBM influx in the Levant is associated with the sprerad of afro asiatic languages?

[Jerome]

Thanks!
The results are interesting,onge is still present,in greater quantities than before but I think this is a case of Wezmeh in particular who seems to be pretty different compared to other iran_n samples like the GanjDareh_N.
Could you model that sample similarly?

Also,now mota/mbuti/deep ancestry takes a major hit,and it seems it's not needed anymore In that large amounts,this does cast a doubt on the basalness of iranN,if only 6% mota/deep Ancestry is needed.

Zlaty Kun is a bit more of an early crown Eurasian so it may eating up a bit of deep ancestry, I wonder what the results will be if bacho Kiro IUP is used instead of Zlaty Kun.
Ust ishim could work too but he has a bit more eastern affinity.

A minor request,could you try this model with BachoKiro_IUP instead of Zlaty Kun,and MA1 instead of AG3?(since MA1 is more eastern than AG3 and the ANE in iranN could be more AG3 and using AG3 may be inflating the Onge).
Also,GanjDareh instead of Wezmeh as the target so the results could be compared a bit to lazaridis' 2018 modelling with Dzudzuana.

Iran_N's ANE is more related to AG3 than to MA1, and I would argue that AG3/MA1 are about equally East/West, if there is any Eastern leaning, it would be from AG3 due to a trifle of UKY/Kolyma stuff.

Iran_GanjDareh_N
BachoKiro_IUP          0.232202 0.128585  1.80582
Kostenki14              0.275729 0.0478508  5.76226
AG3                    0.191370 0.0177896 10.7574
Andaman_100BP.SG        0.198286 0.0799078  2.48144
Mota.SG 0.102413 0.0158679  6.45410
Tail: 0.07
right = c('SouthAfrica_2000BP.SG', 'Ust_Ishim.DG', 'Papuan.DG', 'China_UP', 'RUS_Primorsky_DevilsCave_N.SG', 'GoyetQ116_1', 'Muierii1', 'BachoKiro_BK1653', 'Yana_UP.SG', 'MA1.SG', 'Peru_RioUncallane_1800BP.SG')

It is a haaardddd fail if ZlatyKun is put in the right.

Hmmm,does the model fail if Andaman is put in right pops?(Standard error for andaman seems to be very high at 8% and Z score is medium at 2.5,so wonder how things will change if it goes in right pops?

[Norfern-Ostrobothnian]

I personally feel like qpGraph is going to be more helpful in this situation than qpAdm due to the lack of proximate sources.

[Jerome]

Double Post

[Jerome]

I personally feel like qpGraph is going to be more helpful in this situation than qpAdm due to the lack of proximate sources.

Kale had once ran F4 comparing the east Eurasian ancestry stream,and it seems to have a preference of Jomon>Tianyuan and Jomon>Songe plain(Amur19K-East Asian)


Mbuti.DG Iran_GanjDareh_N Onge.DG Japan_Jomon 0.0016129 4.3404
Mbuti~ Iran_GanjDareh Tianyuan Jomon 2.10E-03 5.44861
Mbuti~ Iran_GanjDareh SongnenPlain_19KBP Jomon 7.67E-04 2.00342


Mbuti.DG Turkmenistan_TepeAnau_CA Japan_Jomon Tianyuan_AR33K -2.95E-03 -7.57412 828828
Mbuti.DG Turkmenistan_TepeAnau_CA Japan_Jomon Onge.DG -1.88E-03 -4.84385 846766
 
Mbuti.DG Turkmenistan_Geoksyur_CA SongnenPlain_19KBP Onge.DG -8.36E-04 -2.22005 702949

SongenPlain19K is preferred over onge and Jomon over SP19K and Jomon over Tianyuan..

The pull seems to be towards an east Eurasian branch which split after tianyuan but before/around jomon had split,rather than towards something like Onge/Late East Asian..

Kale also modelled MA1/AG3 as Muierri+Onge/Andaman+Minor tianyuan/AR33K ,so this makes me wonder if the East Eurasian in ANE could be better represented by something like Jomon instead of Onge?

MA1.SG
Muierii1 0.657092 0.0226929 28.9559
Andaman_100BP.SG 0.309727 0.0544981 5.68326
AR33K 0.0331809 0.0482626 0.687508
Tail: 0.55
right = c('Congo_Mbuti.DG', 'ZlatyKun.SG', 'Ust_Ishim.DG', 'BachoKiro_IUP', 'Papuan.DG', 'Tianyuan', 'Kostenki14', 'Sunghir.SG', 'Gravettian', 'GoyetQ116_1', 'Jomon.SG', 'RUS_Primorsky_Boisman_MN', 'Taiwan_Hanben_IA'

Yana_UP.SG
Muierii1 0.632288 0.0188614 33.5228
Andaman_100BP.SG 0.218551 0.0424194 5.15215
AR33K 0.149161 0.0376843 3.95818
Tail: 0.30

AG3
Muierii1 0.589039 0.0240579 24.4843
Andaman_100BP.SG 0.244668 0.0669798 3.65286
AR33K 0.0727044 0.0492340 1.47671
RUS_Yakutia_KolymaRiver_Meso.SG 0.0935885 0.0403508 2.31937
Tail: 0.21

The affinity in iranN seems to be pretty strange and towards an east Eurasian branch which split after tianyuan but before/around jomon had split,rather than towards something like Onge/East Asian..

I wonder if this type was the kulbalkian of central Asia,this could be a case where when ANE(Ydna R2 population) migrated from southern Siberia/altai towards caspian-aral sea towards Iran,it could have picked this kulbulakian admixture in the way.

Lazaridis modelling showed that the ratio of ANE/ENA(Onge) was 22/10 so something like 70/30 ANE:ENA,this could mean that the ANE population that came had 30% of central Asian/kulbulakian?/Basal East Asian Ancestry (split after tianyuan but split before east Asians diversified)..

Although we don't have such a population to model with,but we have populations with similar ratios of ANE:ENA(Although the ENA in these following sampels is Devil's Gate like)

Altai_6500BP->70% ANE/30% DevilsGate
Tarim_EMBA2->80% ANE/20% Devil's Gate 

If Altai_6500BP is added in leftpops to match the ratio of admixture of such an hypothetical population,maybe we could use this model to see if IranN still receives extra ENA/Onge/Jomon over what can be provided by Altai_6500BP or tarim?

2.
Another case would be Iran_N having some Kolyma/Ust-Kyakhta admixture? that came admixed with the ANE.
Afterall something like Amur_19K is preffered over Onge.

Mbuti.DG Turkmenistan_Geoksyur_CA SongnenPlain_19KBP Onge.DG -8.36E-04 -2.22005 702949

But Jomon over amur_19k
 Mbuti~ Iran_GanjDareh SongnenPlain_19KBP Jomon 7.67E-04 2.00342


I think 1st case is more likely(para-east Asian like people in central Asia)...

[Kale]

My more recent qpgraph modeling resolves such East-Asian specific (including Jomon) attractedness by having some Yana ancestry into them via something like Salkhit.

[Jerome]

My more recent qpgraph modeling resolves such East-Asian specific (including Jomon) attractedness by having some Yana ancestry into them via something like Salkhit.

True,but aurignacians and Gravettians don't show this effect,right?
The closeness to jomon over others.

Mbuti.DG Kostenki14 Onge.DG Jomon -0.000045 -0.105 104415
It will be even lesser if Ar19k was used.


Could you try a model with jomon in left pops for chg/iranN to see if there's still an onge contribution?


Also, do you think it's possible that this eastern ancestry came with ANE who absorbed it on their way southwards through altai/Central Asia?
Amur19K seems to be preffered over onge which does hint towards this.

[Jerome]

Can you model Pinarbasi with the same model? I want to see if they have the same Onge related ancestry since they can be modeled with 30% Zarzian ancestry in your previous model

That model was using Iron gates+WHG to model Pinarbasi so Iran_N may simply be getting inflated to compensate for some eastern Ancestry.
A better model would for Pinarbasi would be Iron gates/WHG+mbuti/mota (for basal proxy)+Tepe_AbdulHosein,this would show how much real IranN/Zarzian is in Pinarbasi,if there is any.

There is some evidence of Zarzian lithics in the anti-taurus region,so Pinarbasi's ancestors could have simply picked some Up when they migrated from North Levant.

Keep in Mind epipaleolithic Anatolia is derived from Levant,not directly from Dzudzuana.

Almarri et al 2021 put the split between Sardinians(ANF) and Levantines at 16K BC.
The first anatolian epipaleolithic industries appear around 14,500 BC(Karain B,Okuzini) from all the studies I have read on epipaleolithic Anatolia and all of them show strong connections in lithics and foraging methods to Kebarans/Pre-Natufian Levant, especially the ucakzigil kebarans of North Levant.
Pinarbasi industry itself dates to 14,000 BC.

The original people of western Anatolia may well have been WHG types considering that we find C1a2 and I2a in Anatolia all though post LGM Anatolia seems to be very sparsely populated with not much evidence of human inhabitation until around 14,000 BC when Kebaran connection starts showing up.
Just 1000 years later Southern Levantines get mushabian/IBM influx and become different from North Levantines and anatolians.

I think Western Aegean/Greece and southern balkans are likely where WHG lived in and expanded from After LGM.
The early industries of these regions are different and more connected to epigravettian.
Then around the Holocene/12-10,000 BC we start seeing the Theopetra Caves and other similar sites,which show more connection to Anatolia.
This also fits in well with the anatolian ancestry in iron gates.

Good post
some questions

1) Do you think the WHG like dna in the western aegean region ultimately is rooted in the Muierii/ BK 1653 genetic cluster?
2) The Mushabian/ IBM influx in the Levant is associated with the sprerad of afro asiatic languages?

1)It's a bit difficult and early to say,WHG does show some eastern affinity to Pinarbasi and Dzuduana over other aurignacians and Gravettians.
I would say ultimately WHG does have some of this eastern component that came pretty early to Europe(33-28K BC) and definitely not in the LGM as some studies have said.
(Another thing,Kamm et al 2018 put the timing of mixing of west Eurasian and Basal euraisna ancestries in Anatolians/Levantines at 32,000 BC.)
The rest of the ancestry of WHG is Aurignacian+Vestonice like definitely..

We need samples from 28K-22k BC southern balkans,Aegean and western coast of Anatolia to say anything definite.
Otherwise many things are just speculation.



2)The split between Anatolians and Levantines dates to 16-15,000 BC.
I would say this is a bit early for proto-Afro Asiatic which seems to date to 12,000 BC.
Anatolians probably separated at a pretty early stage in which proto-afroasiatic didn't fully flesh out.
As for who brought it,I would say this again is a very stretched question,since even if Mushabians brought some haplogroups and also some autosomal ancestry this dosent necessarily mean the langauge arrived with them too.

We see this case with Etruscans,Basques and even the Semitic langauge family in which Levant recived 50% Iran_chalcolithic admixture but the langauge didn't change(apart from some minor substrate from an unknown agglutinative langauge).
Same case could have happened.

[Desdonas]

IIRC in AG before, Agamemnon mentioned that Afroasiatic is more related to Niger-Congo, and the situation fits the phylogeny of E1b1. In my knowledge, Afroasiatic is not closely related to potential ANF languages like Basque.

[Jerome]

IIRC in AG before, Agamemnon mentioned that Afroasiatic is more related to Niger-Congo, and the situation fits the phylogeny of E1b1. In my knowledge, Afroasiatic is not closely related to potential ANF languages like Basque.

Strange theory,but E1b1a and E1b1b split quite early(Pre-LGM) so in terms of uniparentals I see no such connection,on top of that the there isn't any archaeological connection holding natufians together with Central/West Sahara/Sahel.

His theory may be based on the supposed taforalt-connection(taforalt being in West Africa) with Natufians but actually that's a very dubious link.
It's much more likely that the recent southern ancestry in natufians is actually from Cyrenicia or Nile valley,and it's actually very possible that Taforalt itself migrated to NW Africa from that region seperately since there's a strong hiatus and discontinuity between the earlier Aterian of NW Africa and the later iberomaurusians both in their industries and morphology.

The iberomaurusian industry which appears around 22K BC(Start of LGM) rather shows strong connection to the northern Nile valley and cyrenician coast which is quite directing on its origin.

The recent takrori genomes from Libya will be interesting since morphologically they show connection to taforalt and genetically too(from some hints of the abstract)..

I think Nazlet Khater must be sequenced soon, alongwith the 55kya Taramsa Hill.
These two are likely the missing links in all of this(ANA&Basal)...

[Jerome]

Btw,does anyone know if there's something similar to Gendstats(or any similar diagnostic tool) for Admixtools2?

I tried installing admixtools1 and also installed all the libraries but my device has too low RAM for admixtools1 and I can't get even qpFstats to even load.
It gets stuck and after 2-3 minutes the console says 'Aborted,Core dumped'
And I don't like installing admixtools2 since it has no gendstats and from what I have read,has a lot of bugs and errors..

@Kale,any advice?
I have read that your device has low RAM/memory too,so how do you run admixtools1 on it?

[Jerome]

IIRC in AG before, Agamemnon mentioned that Afroasiatic is more related to Niger-Congo, and the situation fits the phylogeny of E1b1. In my knowledge, Afroasiatic is not closely related to potential ANF languages like Basque.

Basque is a very late langauge and I won't expect Okuzini people (anatolian epipaleolithic) to have a similar language as of Kebarans/Natufians uptil 8000 BC.
On top of that Barcin/Boncuklu farmers see a strong influx from Zagros/NE Mesopotamia compared to earlier Pinarbasi HG and Southern arc Paper no.2 shows around 25% Mesopotamia_N input into Barcin Farmers over pinarbasi.

I won't expect them to maintain a similar language after diverging 7000 years ago from Levantines,and on top of that this eastern input may have changed things too..

[Kale]

Btw,does anyone know if there's something similar to Gendstats(or any similar diagnostic tool) for Admixtools2?

I tried installing admixtools1 and also installed all the libraries but my device has too low RAM for admixtools1 and I can't get even qpFstats to even load.
It gets stuck and after 2-3 minutes the console says 'Aborted,Core dumped'
And I don't like installing admixtools2 since it has no gendstats and from what I have read,has a lot of bugs and errors..

@Kale,any advice?
I have read that your device has low RAM/memory too,so how do you run admixtools1 on it?

I use admixtools2 these days. To use admixtools1 I have to create smaller datasets (200 individuals works fine, I haven't pushed it beyond that).
As far as I can tell, you can only get Gendstats using f2_from_precomp in qpadm runs, which reallllllyyyy sucks because I have to make abbreviated datasets to generate the f2 stats! Direct from genotype takes less than a minute on admixtools2, I wish I could get gendstats that way.

[Jerome]

Btw,does anyone know if there's something similar to Gendstats(or any similar diagnostic tool) for Admixtools2?

I tried installing admixtools1 and also installed all the libraries but my device has too low RAM for admixtools1 and I can't get even qpFstats to even load.
It gets stuck and after 2-3 minutes the console says 'Aborted,Core dumped'
And I don't like installing admixtools2 since it has no gendstats and from what I have read,has a lot of bugs and errors..

@Kale,any advice?
I have read that your device has low RAM/memory too,so how do you run admixtools1 on it?

I use admixtools2 these days. To use admixtools1 I have to create smaller datasets (200 individuals works fine, I haven't pushed it beyond that).
As far as I can tell, you can only get Gendstats using f2_from_precomp in qpadm runs, which reallllllyyyy sucks because I have to make abbreviated datasets to generate the f2 stats! Direct from genotype takes less than a minute on admixtools2, I wish I could get gendstats that way.

Is f2_from_precomp the admixtools2 parallel of qpFstats?

Also,does f2_from_precomp take more time to run the models than geno file?
Why else would an abbreviated dataset be needed?

What about f2_from_geno(),will it solve the issue here?