An attempt at deep West Eurasian phylogeny

[Kale]

Yes, Ganj Dareh has some ANE and possibly some Anatolian, but I'm not sold on the latter. It's not necessary.  I think, in my graphs, Iran only has 12% ANE, after being the root of West Asia. You can simply create Anatolians and Dzudzuana/Kotias as a geographic and genetic mix of something UP Europe and something very close to Iran. It works, and makes sense.

What does that root of West Asia look like? It is just Basal? Or are there West-Eurasian and/or Onge-like inputs also?

[Chad]

Yes, Ganj Dareh has some ANE and possibly some Anatolian, but I'm not sold on the latter. It's not necessary.  I think, in my graphs, Iran only has 12% ANE, after being the root of West Asia. You can simply create Anatolians and Dzudzuana/Kotias as a geographic and genetic mix of something UP Europe and something very close to Iran. It works, and makes sense.

What does that root of West Asia look like? It is just Basal? Or are there West-Eurasian and/or Onge-like inputs also?

ANE works well as the source of Eurasian admixture. Isn't Kotias UP the same mtdna as Yana?

[Kale]

Yana is U2'3'4'7'8'9*, Kotias_UP is U4'9.
I'm really interested to see that graph if you can dig it up. I've yet to find a way to model Iran_N without some sort of Anatolian/Dzudzuana-like input, curious to see the implications that has for the rest of the graph structure.

[Norfern-Ostrobothnian]

Iran N has West Eurasian that is for sure. It cannot be solely ANE because unlike ANE Iran N prefers WHG over every other Upper Paleolithic European just like Anatolian does. Iran N also prefers Neolithic Anatolians over Kotias UP so it is very late. It also prefers Anatolian Neolithic more than Anatolia Epipaleolithic, but this could be Iran Neolithic ancestry in Neolithic Anatolians too. I just think Iran N is a very late genotype and this is also backed up by the fact the Mesolithic samples from the Caspian are more ANE heavy than Zagrosian Neolithic people. I think the genotype formed during the Zarzian but finalized during the Neolithic.
Chimp.REF Georgia_Kotias_UP Italy_Mesolithic.SG Czech_Vestonice16 -0.00357 0.000681 -5.25 0.000000156
Chimp.REF Russia_MA1_HG.SG Italy_Mesolithic.SG Czech_Vestonice16 -0.0000764 0.000716 -0.107 0.915
Chimp.REF Iran_GanjDareh_N Georgia_Kotias_UP Turkey_N 0.00331 0.000435 7.62 2.59e-14
Chimp.REF Iran_GanjDareh_N Georgia_Kotias_UP Turkey_Epipaleolithic 0.00170 0.000535 3.17 0.00150

Also here's Tutkaulian exhibiting it's near pure ANE behavior, I seriously would consider adding it to your graphs to see what happens.
Chimp.REF Russia_Yana_UP.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000767 0.000790 -0.971 0.332
Chimp.REF Russia_Sunghir3.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000509 0.000849 -0.600 0.548
Chimp.REF China_Tianyuan Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000528 0.000880 -0.600 0.549
Afontova Gora seems to work best as a mix of Tutkaulian and Mal'Ta
Chimp.REF Russia_AfontovaGora3 Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00199 0.00119 1.66 0.0962
Chimp.REF Russia_MA1_HG.SG Russia_AfontovaGora3 Tajikistan_Tutkaul_N -0.00521 0.00106 -4.91 0.000000910
Chimp.REF Tajikistan_Tutkaul_N Russia_AfontovaGora3 Russia_MA1_HG.SG -0.00719 0.00115 -6.27 3.61e-10
EHG and Yamnaya both prefer Tutkaul over Malta, which could be the CHG related ancestry but also could be linked to their ANE origin.
Chimp.REF Russia_Karelia_HG Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00244 0.000863 2.83 0.00467
Chimp.REF Russia_Samara_EBA_Yamnaya Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00361 0.000704 5.13 0.000000290
Iran N and Anatolia Neolithic prefer Tutkaulian over Malta, but Anatolia Epipaleolithic and Kotias UP do not, at least not significantly enough.
Chimp.REF Iran_Wezmeh_N.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00376 0.000846 4.44 0.00000881
Chimp.REF Turkey_N Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00259 0.000658 3.94 0.0000812
Chimp.REF Turkey_Epipaleolithic Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.000716 0.000871 0.822 0.411
Chimp.REF Georgia_Kotias_UP Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00152 0.000949 1.60 0.109

[Chad]

Iran N has West Eurasian that is for sure. It cannot be solely ANE because unlike ANE Iran N prefers WHG over every other Upper Paleolithic European just like Anatolian does. Iran N also prefers Neolithic Anatolians over Kotias UP so it is very late. It also prefers Anatolian Neolithic more than Anatolia Epipaleolithic, but this could be Iran Neolithic ancestry in Neolithic Anatolians too. I just think Iran N is a very late genotype and this is also backed up by the fact the Mesolithic samples from the Caspian are more ANE heavy than Zagrosian Neolithic people. I think the genotype formed during the Zarzian but finalized during the Neolithic.
Chimp.REF Georgia_Kotias_UP Italy_Mesolithic.SG Czech_Vestonice16 -0.00357 0.000681 -5.25 0.000000156
Chimp.REF Russia_MA1_HG.SG Italy_Mesolithic.SG Czech_Vestonice16 -0.0000764 0.000716 -0.107 0.915
Chimp.REF Iran_GanjDareh_N Georgia_Kotias_UP Turkey_N 0.00331 0.000435  7.62 2.59e-14
Chimp.REF Iran_GanjDareh_N Georgia_Kotias_UP Turkey_Epipaleolithic 0.00170 0.000535  3.17 0.00150

Also here's Tutkaulian exhibiting it's near pure ANE behavior, I seriously would consider adding it to your graphs to see what happens.
Chimp.REF Russia_Yana_UP.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000767 0.000790 -0.971 0.332
Chimp.REF Russia_Sunghir3.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000509 0.000849 -0.600 0.548
Chimp.REF China_Tianyuan Russia_MA1_HG.SG Tajikistan_Tutkaul_N -0.000528 0.000880 -0.600 0.549
Afontova Gora seems to work best as a mix of Tutkaulian and Mal'Ta
Chimp.REF Russia_AfontovaGora3 Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00199 0.00119  1.66 0.0962
Chimp.REF Russia_MA1_HG.SG Russia_AfontovaGora3 Tajikistan_Tutkaul_N -0.00521 0.00106 -4.91 0.000000910
Chimp.REF Tajikistan_Tutkaul_N Russia_AfontovaGora3 Russia_MA1_HG.SG -0.00719 0.00115 -6.27 3.61e-10
EHG and Yamnaya both prefer Tutkaul over Malta, which could be the CHG related ancestry but also could be linked to their ANE origin.
Chimp.REF Russia_Karelia_HG Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00244 0.000863  2.83 0.00467
Chimp.REF Russia_Samara_EBA_Yamnaya Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00361 0.000704  5.13 0.000000290
Iran N and Anatolia Neolithic prefer Tutkaulian over Malta, but Anatolia Epipaleolithic and Kotias UP do not, at least not significantly enough.
Chimp.REF Iran_Wezmeh_N.SG Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00376 0.000846  4.44 0.00000881
Chimp.REF Turkey_N Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00259 0.000658  3.94 0.0000812
Chimp.REF Turkey_Epipaleolithic Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.000716 0.000871 0.822 0.411
Chimp.REF Georgia_Kotias_UP Russia_MA1_HG.SG Tajikistan_Tutkaul_N 0.00152 0.000949  1.60 0.109

That's because Iranian related ancestry is in whg via its Anatolian ancestry.

[kolompar]

Kotias UP was U4'9, Yana were U2'3'4'7'8'9 (and MA1 was also U). That alone doesn't mean they are closely related, TMRCA is estimated around the time of OoA here (last few pages): https://www.ncbi.nlm.nih.gov/pmc/article...n/mmc1.pdf It's from a 2009 paper but it looks OK and I haven't found anything that's much better or contradicts it.
But U can be associated with West Eurasia in general, not found in the few early Europeans we have but dominates in later genetic "West Eurasians". The Dzudzuana samples were U6 and N* and SAT29 was also N* by the way.

So based on those estimates and YFull, a look on West Eurasian mtDNA:

A few haplogroups splitting straight from N, N1 splits already around OoA.
N1b TMRCA around end of LGM, found in Natufian and other Levantines and in Anatolian Neolithic at a lower rate.
N1a1a is more frequent in Anatolia, N1a1b found in 8000BC Iraq and also includes I (dates to LGM) which is found from Germany LBK through Russia Steppe Eneolithic and Israel_C to Turkmenistan_C and Shahr-i Sokhta. N1a2 is found in BA Turkmenistan.

X also splits from N but has a later TMRCA with even X1'2'3 (found in Ganj Dareh) probably after LGM and X2 definitely, found in Anatolian Neolithic and... North American Natives, already in Kennewick man 7000BC.

W post-LGM, Anatolian and also in Central Asia and Shahr-i Sokhta.

Others are all downstream from R, which was also likely already differentiated by OoA and has subclades found all over Eurasia. Ust'-Ishim man famously belonged to R*, differing from the root sequence of R by a single mutation.

R0a is post-LGM and seems to be strictly Natufian-related, although according to Wikipedia some believe it evolved in Ice Age oases in South Arabia (). Also supposedly found in Iberomaurusian, don't know how reliable that is https://doi.org/10.1080%2F24701394.2016.1258406

HV also comes from R0, found in Natufian, Iran Meso, Iraq PPN, EEF... Has dozens of recent subclades on YFull, but the most important is H: very recent TMRCA (15300 ybp), shared between 8000BC Iraq, Kotias, Iron Gates, Anatolian Neolithic, and is the most potent female haplogroup at 40-50% of modern Europeans.

R2'JT is another major Middle Eastern haplogroup, with deep Paleolithic splits. R2 is in Iran Neolithic, as is J1d (also in two Iraqi Neolithics), J1c looks Anatolian, J1b steppe-related(?), J2 in Natufian_contam and Anatolian. J TMRCA around 30k BC.
T is post-LGM but shared between Jordan PPNB, Iran, Anatolia, and Vologda Veretye Meso.

(There's also ancient Egyptians everywhere on YFull, where do they come from?)

And finally, U forms probably at the time of OoA...

U1 is found in Iraq PPN, Iran Hajji Firuz Neo, Shahr-i Sokhta and Iran related populations but not in Anatolian//EEF.

U5 is only known from a few Gravettians and the two post-LGM WHG lines.

U6 in Muierii and Dzudzuana (is that the publicly available one, has anyone checked?), but all moderns come from Iberomaurusian. Dates kind of consistent with an entry during the LGM (low sea levels) along with M1.

U2'3'4'7'8'9 and even U2 break up a few hundred years after U.
Kostenki is U2*, Sunghir and Gravettians are on their own branch U2f with no descendants. Others are U2a, U2b and a branch with U2c'd and U2e. U2e is EHG, d Middle Eastern, but the others are Indian, possibly not even from Iran Neo but Paleolithic.

U3 is Anatolian.
U7 is found in Iran and Azerbaijan Neo and also in Steppe Maikop.

U4'9 is interesting, U9 seems Middle Eastern but U4 is the most typical EHG line. They split before Kotias so she isn't the ancestor, but U4 has a recent TMRCA so could be from the Caucasus.

U8 also splits around OoA. Only U8c is found among Sunghir and Gravettians, although U8a appears in Magdalenians. U8b includes K, one of the largest Anatolian lineages. K2b in Pinarbasi, K3 in Satsurblia, K2a in Iran Meso and EEF, K1a is shared between Anatolian Neo, Israel and Jordan PPNB and Iron Gates, K1b is Anatolian and Iran Neolithic. K1 and K2 each has TMRCA around the LGM or after it. U8* is found in Anatolian Neo according to YFull.

Arguably even U2 and U8 have more diversity in the Middle East than Europe, and U as a whole definitely does. If those U2'3'4'7'8'9 samples are closely related it could be just 4-5 lines out of dozens that came to Europe in the Paleolithic. So while U could be tied to a certain ancestry, I'm not sure if it's European.

[TanTin]

Recently I've been doing successful PCA projections with the main focus on ancient / basal/ and WHG groups.
And I have a very good results from Onge / Jarawa group, which has been projected near the very basal individuals, including Tianyan.
Kale, are you able to include Onge and Jarawa in your diagram ?

[TanTin]

You may notice the colors on my diagram,  this is a test for a specific snp marker.  In this case it is for CHR 1 and you may see the snp name.


When I was checking for these specific groups I noticed a pattern with snp deletion.
For example: we have a major allele C and minor allele T. Almost all Onge and Jar are with the minor allele T.   But in many cases we may find that in some groups there is no major or minor allele, so that meand there was snp deletion,  or such snip marker is missing in some groups.  
There is a certain pattern for these deleted snips and in some groups we may find a number of such deleted snips.  I don't know if this could be related to Neanderthal deserts , may be someone could explain that better.
So as a consequence of such deleted snips you can not use the F3/F4 statistics to count for these deleted snips. However these "missing snips" are also forming clusters..  I can give you some examples of such clusters formed by missing snips, where all other groups would have the markers, exept for the clusters where it is missing.

[Kale]

I've often wondered something similar to what you are saying...
Take a parent populations with a number of low frequency alleles. It has two child populations, each of which head their separate ways, having no further contact.
Each child experiences it's own bottleneck event (which are more likely to remove minor alleles?), now the children share more alleles with each other than the parent despite no contact?

[Horatio McCallister]

Recently I've been doing successful PCA projections with the main focus on ancient / basal/ and WHG groups.
And I have a very good results from Onge / Jarawa group, which has been projected near the very basal individuals, including Tianyan.
Kale, are you able to include Onge and Jarawa in your diagram ?

Could you re-run this with ancient MENA samples included? Ibermaurusian, Natufians, Kotias, etc (I assume the Iran listed is already Iran_N?).

[TanTin]

Could you re-run this with ancient MENA samples included? Ibermaurusian, Natufians, Kotias, etc (I assume the Iran listed is already Iran_N?).

Here I give you another diagram:  I calculate the number of alleles matching to a specific individual.
I have a list of 60 snps.
The individual, that I used for reference is: "MOK23.SG"  - Mokrin 
As you may notice there are some zones with low or 0 number of snips.  Jar/ Onge are such zone, but surprisingly the vikings and some WHG are also from that zero zone.

[kolompar]

Tried my simple two-way split kind of Basal+East=West model in qpAdm, with Kale's pops. Looks like Ust'-Ishim might work best as the source for Kostenki which makes sense geographically.
right = c("Mbuti.DG", "China_UP", "Onge_Jarawa", "Czechia_Bohemia_UP_HG", "Bulgaria_BachoKiro_LatePleistocene", "Romania_Oase_UP_enhanced", "Iran_N", "Ethiopia_4500BP", "Papuan.DG")

Code:

$weights
# A tibble: 2 x 5
  target            left                  weight    se    z
  <chr>            <chr>                  <dbl>  <dbl> <dbl>
1 Russia_Kostenki14 Turkey_Epipaleolithic  0.554 0.0533 10.4
2 Russia_Kostenki14 Russia_Ust_Ishim_HG.DG  0.446 0.0533  8.38

$rankdrop
# A tibble: 2 x 7
  f4rank  dof  chisq        p dofdiff chisqdiff  p_nested
  <int> <int>  <dbl>    <dbl>  <int>    <dbl>    <dbl>
1      1    7  3.38 8.48e- 1      9      265.  7.19e-52
2      0    16 268.  9.18e-48      NA      NA  NA

I guess you can get a lot of things to pass with such limited right pops, but qpGraph works with this much too.
So the same thing in qpGraph, dotfile:

Show Content

Spoiler
digraph G {
label = "";
labelloc=t;
labeljust=l;
size = "7.5,10";
R [shape = point];
Rr_Rrlr [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b [shape = point];
Rr_Rrlr_Rrlrj [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjo [shape = point];
admix [shape = point];
R  ->  MbutiDG  [ label = "29" ];
R  ->  Rr_Rrlr  [ label = "29" ];
Rr_Rrlr  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  [ label = "0" ];
Rr_Rrlr  ->  Czechia_Bohemia_UP_HG  [ label = "150" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  [ label = "2" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic  [ label = "14" ];
Rr_Rrlr_Rrlrj  ->  Onge_Jarawa  [ label = "28" ];
Rr_Rrlr_Rrlrj  ->  China_UP  [ label = "20" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  ->  Rr_Rrlr_Rrlrj  [ label = "4" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjo  [ label = "4" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  [ label = "1" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Bulgaria_BachoKiro_LatePleistocene  [ label = "40" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic  ->  Turkey_Epipaleolithic  [ label = "142" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic  ->  admix  [ style=dotted, label = "53%" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjo  ->  Russia_Ust_Ishim_HGDG  [ label = "1" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjo  ->  admix  [ style=dotted, label = "47%" ];
admix  ->  Russia_Kostenki14  [ label = "147" ];
}

Worst f4 |Z|=2.14, find_graphs doesn't find anything better, possibly the best graph with these pops. Gravettians also work as Kostenki-like, so might be just one post-CI Middle Eastern wave until WHG.

Adding Iran and MA1 in case it needs it, optimizing with find_graphs. Only needed Onge_Jarawa, Pinarbasi is BachoKiro+proto-Iran, also makes sense geographically, I would expect bidirectional geneflow around Caucasus/Anatolia. Worst residual 2.82, dot:

Show Content

Spoiler
digraph G {
label = "";
labelloc=t;
labeljust=l;
size = "7.5,10";
R [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_bl [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj [shape = point];
admix [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjw [shape = point];
admix_Russia_Kostenki14 [shape = point];
admixb [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwz [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_re [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_u [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh [shape = point];
admixu [shape = point];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwzv [shape = point];
admixc [shape = point];
R  ->  MbutiDG  [ label = "28" ];
R  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_bl  [ label = "28" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_bl  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r  [ label = "2" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_bl  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  [ label = "21" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  ->  Russia_Ust_Ishim_HGDG  [ label = "2" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  ->  admix  [ style=dotted, label = "75%" ];
admix  ->  admix_Russia_Kostenki14  [ label = "5" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj  [ label = "2" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjw  [ label = "4" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjw  ->  China_UP  [ label = "20" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjw  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwz  [ label = "8" ];
admix_Russia_Kostenki14  ->  Russia_Kostenki14  [ label = "141" ];
admix_Russia_Kostenki14  ->  admixb  [ style=dotted, label = "85%" ];
admixb  ->  Russia_MA1_HGSG  [ label = "144" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwz  ->  Onge_Jarawa  [ label = "20" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwz  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwzv  [ label = "72" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r  ->  Czechia_Bohemia_UP_HG  [ label = "150" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_re  [ label = "3" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  [ label = "55" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_r_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  ->  admixc  [ style=dotted, label = "83%" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_re  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b  [ label = "1" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_re  ->  Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_u  [ label = "1" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_u  ->  Bulgaria_BachoKiro_LatePleistocene  [ label = "40" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_u  ->  admixu  [ style=dotted, label = "57%" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  ->  admix  [ style=dotted, label = "25%" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrj_b_Turkey_Epipaleolithic_rh  ->  admixu  [ style=dotted, label = "43%" ];
admixu  ->  Turkey_Epipaleolithic  [ label = "143" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwzv  ->  admixb  [ style=dotted, label = "15%" ];
Rr_Rrlr_Rr_Rrlr_Rr_Rrlr_Rrlrjwzv  ->  admixc  [ style=dotted, label = "17%" ];
admixc  ->  Iran_N  [ label = "7" ];
}

[TanTin]

Here is a new PCA with focus on Europe and Denisova/Neanderthals.


This is for a snip on 1st CHR. "snp_1_1596979"
Among humans it is found rarely, only few Papuans have it + and one viking + 1 in Spain. 

And here is another example for another snip:



We may create hundreds and thousands of such pictures and examples .  In general this is just a small visualization of what we have in the raw data used for PCA and for F3/F4 statistics.

[TanTin]

Here is another very good example for the Denisova ancestry in Papuans.
On CHR 2.  Snip: snp_2_10311419

Here is the list:
1          HGDP00548        C                      Papuan
2          HGDP00551        C      Ignore_Papuan_discovery
3          HGDP00663        C                      Nasioi
4      HGDP00551_WGA        C      Ignore_Papuan_discovery
5        Denisova3.DG        C                  Denisova.DG
6  Denisova3_snpAD.DG        C                  Denisova.DG
7              TAP003        C      French_Polynesia_400BP
8      S_Papuan-9.DG        C  Ignore_Papuan_discovery.DG
9      HGDP00551.SDG        C Ignore_Papuan(discovery).SDG
10            JP14.SG        C        Ireland_MN_contam.SG
11      HGDP00663.SDG        C                  Nasioi.SDG
12      S_Papuan-7.DG        C                    Papuan.DG
13      HGDP00548.SDG        C                  Papuan.SDG
14      FUT001_noUDG        C              Vanuatu_1100BP
15            FUT002        C              Vanuatu_1200BP
16              I3921        C              Vanuatu_1300BP
17        I3921_noUDG        C        Vanuatu_1300BP_noUDG
18        I4096_noUDG        C        Vanuatu_1300BP_noUDG
19              I4106        C                Vanuatu_150BP
20        I4105_noUDG        C          Vanuatu_150BP_noUDG
21            I10967        C    Vanuatu_200BP_oPolynesian
22              I4451        C              Vanuatu_2300BP
23            TAN002        C              Vanuatu_2500BP
24            MAL004        C              Vanuatu_2600BP
25            I10968        C                Vanuatu_400BP
26              I5259        C                Vanuatu_500BP

You may notice that I take screenshots from 3D.  Below I will rotate the picture: 

[TanTin]

And the last example: on CHR 21 .  rs116130419

Show Content

Spoiler
              famInd my_color                            X
1          HGDP00452        A                  Ignore_Biaka
2          HGDP00464        A                        Biaka
3          HGDP00540        A                        Papuan
4          HGDP00543        A                        Papuan
5          HGDP00546        A                        Papuan
6          HGDP00550        A                        Papuan
7          HGDP00555        A                        Papuan
8          HG01885.SG        A                        ACB.SG
9          HG01914.SG        A                        ACB.SG
10        HG02332.SG        A                        ACB.SG
11        HG02479.SG        A                        ACB.SG
12        HG02481.SG        A                        ACB.SG
13        HG02508.SG        A                        ACB.SG
14        HG02545.SG        A                        ACB.SG
15        HG02549.SG        A                        ACB.SG
16        NA19909.SG        A                        ASW.SG
17        NA19916.SG        A                        ASW.SG
18        NA20291.SG        A                        ASW.SG
19      HGDP00464.SDG        A                    Biaka.SDG
20        NE56_noUDG        A  China_AmurRiver_LPaleolithic
21        HG01142.SG        A                        CLM.SG
22        HG01360.SG        A                        CLM.SG
23            VLI050        A            Czech_EBA_Unetice
24      Denisova3.DG        A                  Denisova.DG
25 Denisova3_snpAD.DG        A                  Denisova.DG
26        HG02923.SG        A                        ESN.SG
27        HG03126.SG        A                        ESN.SG
28        HG03511.SG        A                        ESN.SG
29  Rochedane_Fu2016        A              France_Rochedane
30        HG02574.SG        A                        GWD.SG
31        HG02757.SG        A                        GWD.SG
32        HG02798.SG        A                        GWD.SG
33        HG02860.SG        A                        GWD.SG
34        HG03024.SG        A                        GWD.SG
35        HG03539.SG        A                        GWD.SG
36            I25518        A            Hungary_IA_LaTene
37      HGDP00452.SDG        A              Ignore_Biaka.SDG
38        NA19390.SG        A                        LWK.SG
39        NA19403.SG        A                        LWK.SG
40        NA19428.SG        A                        LWK.SG
41        HG03082.SG        A                        MSL.SG
42        HG03388.SG        A                        MSL.SG
43        HG03557.SG        A                        MSL.SG
44        HG03558.SG        A                        MSL.SG
45        HG03567.SG        A                        MSL.SG
46        HG03577.SG        A                        MSL.SG
47      S_Papuan-2.DG        A                    Papuan.DG
48      S_Papuan-4.DG        A                    Papuan.DG
49    B_Papuan-15.DG        A                    Papuan.DG
50      S_Papuan-1.DG        A                    Papuan.DG
51    S_Papuan-11.DG        A                    Papuan.DG
52      HGDP00540.SDG        A                    Papuan.SDG
53      HGDP00543.SDG        A                    Papuan.SDG
54      HGDP00546.SDG        A                    Papuan.SDG
55      HGDP00550.SDG        A                    Papuan.SDG
56      HGDP00555.SDG        A                    Papuan.SDG
57        HG02089.SG        A                        PEL.SG
58    DA253_noUDG.SG        A Russia_Shamanka_Eneolithic.SG
59    VK219_noUDG.SG        A              Russia_Viking.SG
60          mfo001.SG        A        South_Africa_400BP.SG
61    VK517_noUDG.SG        A Sweden_Viking_1d.rel.VK527.SG
62              I4106        A                Vanuatu_150BP
63        I4425_noUDG        A                Vanuatu_150BP
64        I4106_noUDG        A          Vanuatu_150BP_noUDG
65            I10968        A                Vanuatu_400BP
66              I5259        A                Vanuatu_500BP
67        NA18853.SG        A                      YRI_o.SG
68        NA19200.SG        A                      YRI_o.SG
69        NA19113.SG        A                        YRI.SG
70        NA19171.SG        A                        YRI.SG
71              A1822        A    Hungary_DanubeTisza_EAvar
72  VK517.SG_3.2Mall        A Sweden_Viking_1d.rel.VK527.SG