Ust'ishim

[Codaman]

Lake Mungo Man(LM3) from Australia is dated to 40kya. There is a gap of several thousands of years between it and the next oldest AMH from SE and South Asia, and it forms a physical type of it's own characterized by relatively high gracility and tall stature at 196 cm which may suggest some length of isolated development.
The AMH from Tam Pa Ling Cave in Laos should be even older than Mungo Man but the dating on them has a pretty wide range, 46k to 60k with the lower and middle looking more reasonable for a variety of reasons. If that range holds true they should exceed or at least match Ust'Ishim in age. In my opinion, SE Asia was settled very rapidly after OoA probably much like Native Americans found their way to Chile not that long after they had made it past Canada. I'm not aware of any compelling factors which suggest OoA happened before 55kya, although genetic separation could have started much earlier within Africa.

[kolompar]

Would SE-Asia have been the best evironment for OOA? Their ancestors spent the preceeding (up to) 20,000 years in NE-Africa/Levant, which I'd assume would be more savannah-like at the time?
If anything that would make the steppe at least a more familiar environment (If ~50 degrees latitude was steppe during the IUP).
Here's a map of IUP sites. Clusters around Baikal and Altai. Also shows how much of a Northern outlier Ust-Ishim is.
https://en.wikipedia.org/wiki/File:Initi..._sites.jpg

Even if they felt more at home in arid regions, the tropics should have a much higher carrying capacity. But Bab-el-Mandeb is very close to the Ethiopian highlands and from there it's just another small leap to India.
I can't see humans spending a substantial amount of time in a less favorable environment already out of Africa but before spreading to the rest of Eurasia, that just raises two more questions, why did they stop there and then why could they continue? I think as soon as they got out of Africa they just spread rapidly everywhere, that's what the DNA shows for me, with a date of around 46k BC from Y-haplogroups.
I'm not sure how reliable the map is, I think the idea of IUP only came after ancient DNA and wasn't a thing in archeology before. It definitely has a sampling bias and maybe also in assigning finds to it.

[Desdonas]

Lake Mungo Man(LM3) from Australia is dated to 40kya. There is a gap of several thousands of years between it and the next oldest AMH from SE and South Asia, and it forms a physical type of it's own characterized by relatively high gracility and tall stature at 196 cm which may suggest some length of isolated development.
The AMH from Tam Pa Ling Cave in Laos should be even older than Mungo Man but the dating on them has a pretty wide range, 46k to 60k with the lower and middle looking more reasonable for a variety of reasons. If that range holds true they should exceed or at least match Ust'Ishim in age. In my opinion, SE Asia was settled very rapidly after OoA probably much like Native Americans found their way to Chile not that long after they had made it past Canada. I'm not aware of any compelling factors which suggest OoA happened before 55kya, although genetic separation could have started much earlier within Africa.

Yes, there is no evidence to suggest that the OoA population expanded to North Eurasia and East Eurasia before 55kya. But the genome before 40kya in SW Asia is still blank, so we cannot rule out the possibility that the OoA population first experienced a long-time bottleneck in Arabia before the Neanderthal admixture. IIRC on Razib's blog, he suggests that 55kya is the chronology of the shared Neanderthal admixture. Also considering the situation of Basal Eurasians, I infer that an Eurasian meta population entered SW Asia during 65-60kya.

https://mdpi-res.com/d_attachment/quater...1699256332

This Siberian paper indicates that AMH arrived in northeastern Siberia at 50kya, and one figure included shows that the earliest archaeological site in the region was at 49kya. At last, 23mofang now adjusts the age of F to around 50500 years.

[Desdonas]

It is rather counterintuitive longitudinally, perhaps latitude is the rationalizing factor?
Stage 1) A population exists in Kazakhstan ancestral to Ust-Ishim and East-Eurasians
Stage 2) Ust-Ishim's ancestors migrate North (he's fairly far up, 57 degrees)
Stage 3) The population remaining at lower latitudes shortly thereafter expands East (as East-Eurasians proper) and West (BachoKiro/Oase)

Drift ~= time / population size. We only have a few hundred years to play with if a clean branching will account for the shared y-line of Ust-Ishim and Oase, but a small population size is conceivable at the initial peopling of Eurasia.

I would associate East Eurasian with the India/Southeast Asia region, based on modern distribution and haplogroups. I don't know how habitable or even crossable Kazakhstan or Western China and Mongolia was. Maybe somewhere along the Himalaya foothills is possible for either kind of ancestry Ust'-Ishim has, based on his location, but I don't know what route Bacho Kiro could have taken while seemingly avoiding Ust'-like people.
But Southeast Asia definitely would have been the most suitable environment for the OoA population, just compare the range of non-human primates for example. https://commons.wikimedia.org/wiki/File:...ensity.png
So it could have been a simple numerical advantage that helped them conquer a lot of Eurasia after a few hundred years. And if we link East Eurasian to the Asian tropics then latitude could be a factor in that Ust' might have ancestry from a more cold adapted population.
So what was that population? Zlaty Kun like? They aren't really related in terms of f-stats but I remember they shared some IBD and would kind of fit my idea geopgraphically.
Why aren't West Eurasians (Europeans and Anatolia/Iran) more related to Ust'? Even the Europeans who are closest in f3 are not significant in f4(Mbuti, x, Ust', East). Is it extra East levelling it out, something like Bacho Kiro for Europe and maybe the LT population or later AASI for Iran?

Considering the possible existence of Denisovans in Southeast Asia (especially D2), it is unlikely that this is the homeland of Crown Eurasians/Ust Ishim. Here, I agree more with the figures in Vallini's paper regarding the repeated expansion of Crown Eurasians from the Iranian plateau. Oceanians, AASI and Hoabinhian passed through South Asia, Ust'Ishim and Bacho Kiro passed through Central Asia, while there's also an arrow passing through the Tarim Basin (possibly Tianyuan).

Regarding the IUP, this paper proposes that, though the technology and typology of the IUP/Emiran of the Southern Levant have clear similarities with the IUP of Central Eurasia, the appearance of IUP industries in the Levant and Altai is nearly synchronous. The possibility of directed migration or transfer of technological traditions from the Levant to Siberia and Central Asia in this chronological interval requires further discussion, especially regarding the absence of IUP assemblages in the Middle East.

https://www.sciencedirect.com/science/ar...842300132X

[Desdonas]

In fact, while adjusting the ages, 23mofang has also added SNPs at various levels, including GHIJK, HIJK, IJK, K, and K2. For example, at HIJK level, it is only defined by PF3494 in yfull. While 23mofang currently also has BY50278, BY31109, BY26103, BY43077, BY23717, MF659167, and MF663664. (8 SNPs in total)

[Horatio McCallister]

In fact, while adjusting the ages, 23mofang has also added SNPs at various levels, including GHIJK, HIJK, IJK, K, and K2. For example, at HIJK level, it is only defined by PF3494 in yfull. While 23mofang currently also has BY50278, BY31109, BY26103, BY43077, BY23717, MF659167, and MF663664. (8 SNPs in total)

Maybe a silly question, but where does 23mofang get its non-Chinese samples? I can believe for clades like O, C, N, and probably D they have better resolution than y-full, but I'm wondering how they calibrate non-East Asian clades since I imagine they have very few non-Chinese customers.

[Desdonas]

In fact, while adjusting the ages, 23mofang has also added SNPs at various levels, including GHIJK, HIJK, IJK, K, and K2. For example, at HIJK level, it is only defined by PF3494 in yfull. While 23mofang currently also has BY50278, BY31109, BY26103, BY43077, BY23717, MF659167, and MF663664. (8 SNPs in total)

Maybe a silly question, but where does 23mofang get its non-Chinese samples? I can believe for clades like O, C, N, and probably D they have better resolution than y-full, but I'm wondering how they calibrate non-East Asian clades since I imagine they have very few non-Chinese customers.

In fact, due to historical migrations, yhg-J, R, etc. are widely present in China. In addition, F2-Y27277 also has a considerable number of test samples in China. Therefore, although it is not as good as other companies in research on various Western Eurasian clades, there may still be enough capacity to adjust "macro" levels such as F, GHIJK, and K.

[Ebizur]

In fact, while adjusting the ages, 23mofang has also added SNPs at various levels, including GHIJK, HIJK, IJK, K, and K2. For example, at HIJK level, it is only defined by PF3494 in yfull. While 23mofang currently also has BY50278, BY31109, BY26103, BY43077, BY23717, MF659167, and MF663664. (8 SNPs in total)

Maybe a silly question, but where does 23mofang get its non-Chinese samples? I can believe for clades like O, C, N, and probably D they have better resolution than y-full, but I'm wondering how they calibrate non-East Asian clades since I imagine they have very few non-Chinese customers.

In fact, due to historical migrations, yhg-J, R, etc. are widely present in China. In addition, F2-Y27277 also has a considerable number of test samples in China. Therefore, although it is not as good as other companies in research on various Western Eurasian clades, there may still be enough capacity to adjust "macro" levels such as F, GHIJK, and K.

To be precise, Western Eurasian haplogroups are currently represented at the following proportions among males in China according to 23mofang:

Haplogroup E 0.14% (also includes a few individuals of recent African paternal lineage, but most members of haplogroup E in China belong to typically Western Eurasian subclades of E-M35.1)
Haplogroup G 0.18% (of whom about 4/5 belong to G2-P287 and about 1/3 belong to G2a2b2a1a1-PF3345)
Haplogroup H 0.06%
Haplogroup I 0.05% (of whom about 58% belong to I2-M438 and about 42% belong to I1-M253)
Haplogroup J 0.72% (0.59% J2-M172, 0.13% J1-L255)
Haplogroup L 0.12%
Haplogroup T 0.07%
Haplogroup Q(xQ1a1a-M120) ≈0.31% (Q1b-M346 0.18%, Q2-L275 0.09%, Q1a2a-L712 0.04%)
Haplogroup R 1.68% (1.15% R1a1a1-M417, 0.16% R1b1a1b-M269, 0.11% R2a-M124, 0.09% R1b1a1a1-M478, 0.09% R1b2-PH155, 0.03% R1b1a2-V1636, 0.03% R2b-FGC50368 [former R* as found among Burusho, Kalash, etc.]; there are also a few members of other subgroups, including R1a(xR1a1a1-M417), R1b1b-V88, and R-FTE1 > R-MF441059)

Haplogroup Q1a1a-M120 by itself accounts for the Y-DNA of about 2.48% of all males in present-day China according to 23mofang. It is especially common, accounting for more than 4% of all males, among Han Chinese in northern parts of China proper, such as Shaanxi and Shanxi. However, Q1a1a-M120 is not regularly found outside East/Southeast Asia.

[CLTVTE]

In “The deep population history of northern East Asia from the Late Pleistocene to the Holocene”, the Q-M120-affiliated sample, such as AR9.2K_outlier, is not closer to AfontovaGora, than AR19K was. It means that there was no additional Western Eurasian ancestry for the Q-M120-related population in Northeast China during the last 19000 years. The AR19K sample clustered the closest to the Japanese in “The deep population history of northern East Asia from the Late Pleistocene to the Holocene”, which is consistent with the prevalence of the Paleolithic Western Eurasian-related ancestry in the territory of Northeast Asia in the first place. AR9.2K_outlier could not be modeled using Early Neolithic Shandong specimens in the same article. The Q-M120-related population migrated to the territory of the Jiangsu Province in the Neolithic, interacting with bearers of haplogroups from other migrations, such as mtDNA F4a2, mtDNA D4g2a1, mtDNA D4g2b1, and later distributed to other locations from there. In Northeast China, Tungusic specimens proved to be more closely affiliated to the most ancient Q-M120-related specimens.

As for bearers of numerous Western Eurasian haplogroups, their late arrival to China did not influence the formation of language families in China.

For example, when distinct Indo-European languages separated from their common ancestor or ancestors 5000-8000 years ago (according to different estimates), we can see perfect reconstructions of their common lexicon, and, most importantly, the basic lexicon items from different Indo-European branches represent perfect semantic matches. No similar qualitative connection to Western Eurasian languages can be observed for any language family in East Asia in the broad sense, where even connections of Japonic and Koreanic, putative East Asian members of the Altaic language family, to other Altaic languages, are contested by the most scrupulous linguists.

[alchemist223]

The thing with Ust'-Ishim's haplogroup is it's between K2a and K2b, and that split looks clean and late enough to suggest East Eurasians were already formed by then and it was the actual split between East Asians and Oceanians.

But they form a clade against Ust'-Ishim:

Code:

pop1    pop2    pop3    pop4    est    se    z    p    n
Mbuti.DG    Russia_Ust_Ishim.DG    China_Tianyuan    Papuan.DG    -0.000785811    0.000659358    -1.191781159    0.233347065    884492
Mbuti.DG    Russia_Ust_Ishim.DG    China_Tianyuan    ONG.SG    0.000254893    0.000590266    0.431827632    0.665866695    878414
Mbuti.DG    China_Tianyuan    Russia_Ust_Ishim.DG    ONG.SG    0.004607883    0.00070986    6.49125768    8.51E-11    878414
Mbuti.DG    China_Tianyuan    Russia_Ust_Ishim.DG    Papuan.DG    0.003761995    0.000730911    5.146997851    2.65E-07    884492
Mbuti.DG    ONG.SG    Russia_Ust_Ishim.DG    China_Tianyuan    0.00435299    0.000612967    7.10151002    1.23E-12    878414
Mbuti.DG    Papuan.DG    Russia_Ust_Ishim.DG    China_Tianyuan    0.004547806    0.000658013    6.911422681    4.80E-12    884492

And even kind of holds with the probably older Bacho Kiro:

Code:

Mbuti.DG    Bulgaria_BachoKiro_LatePleistocene    Russia_Ust_Ishim.DG    China_Tianyuan    0.002020002    0.000696419    2.900552838    0.00372505    827041
Mbuti.DG    Bulgaria_BachoKiro_LatePleistocene    Russia_Ust_Ishim.DG    ONG.SG    0.001028649    0.000560273    1.835979814    0.066360642    985969
Mbuti.DG    Bulgaria_BachoKiro_LatePleistocene    Russia_Ust_Ishim.DG    Papuan.DG    -0.000608792    0.000609033    -0.999604501    0.317501944    992072

I can't see how these populations could have spent time together after the split with Ust'-Ishim. So he must have an admixture that pulls him away, and it can't be archaic that affects these stats as Papuan have more than him while Onge seem to have less.

My impression is that the Ust'-Ishim man's haplogroup is K2a, no? I believe he is positive for SNPs CTS11667 and M2308.

[kolompar]

Considering the possible existence of Denisovans in Southeast Asia (especially D2), it is unlikely that this is the homeland of Crown Eurasians/Ust Ishim. Here, I agree more with the figures in Vallini's paper regarding the repeated expansion of Crown Eurasians from the Iranian plateau. Oceanians, AASI and Hoabinhian passed through South Asia, Ust'Ishim and Bacho Kiro passed through Central Asia, while there's also an arrow passing through the Tarim Basin (possibly Tianyuan).

Regarding the IUP, this paper proposes that, though the technology and typology of the IUP/Emiran of the Southern Levant have clear similarities with the IUP of Central Eurasia, the appearance of IUP industries in the Levant and Altai is nearly synchronous. The possibility of directed migration or transfer of technological traditions from the Levant to Siberia and Central Asia in this chronological interval requires further discussion, especially regarding the absence of IUP assemblages in the Middle East.

https://www.sciencedirect.com/science/ar...842300132X

Denisovan would help a lot but I can't find much resources for it and I don't think it can be easily and reliably calculated from just f-stats.
This is the paper that introduced D0/1/2 ancestry but it only gives estimates for Oceanians and larger regions with many different groups lumped together so I don't know how widespread it actually is. Andamanese were said to have no Denisovan, and they have overall low levels of archaic. Other Asian pops like Jomon, Hoabinhian and Longlin also have similarly low archaic.
https://www.cell.com/cell/fulltext/S0092...19)30218-1

This is the Salkhit paper. They claim no Denisovan in Ust'-Ishim and any early European, it was before Bacho Kiro but Oase should be included. So yes, it makes it unlikely they come from far in the East, unless the Denisovan admixture is later. There are also a few good figures in the supplements on the relations of the earliest ancients. For moderns I think they only check the Denisovan segments shared with Salkhit and Tianyuan.
https://www.science.org/doi/10.1126/science.abc1166

My impression is that the Ust'-Ishim man's haplogroup is K2a, no? I believe he is positive for SNPs CTS11667 and M2308.

My point was just it's a clearly East Eurasian haplogroup, so I guess that stands even more as K2a should be more specifically East Asian.

[Desdonas]

Considering the possible existence of Denisovans in Southeast Asia (especially D2), it is unlikely that this is the homeland of Crown Eurasians/Ust Ishim. Here, I agree more with the figures in Vallini's paper regarding the repeated expansion of Crown Eurasians from the Iranian plateau. Oceanians, AASI and Hoabinhian passed through South Asia, Ust'Ishim and Bacho Kiro passed through Central Asia, while there's also an arrow passing through the Tarim Basin (possibly Tianyuan).

Regarding the IUP, this paper proposes that, though the technology and typology of the IUP/Emiran of the Southern Levant have clear similarities with the IUP of Central Eurasia, the appearance of IUP industries in the Levant and Altai is nearly synchronous. The possibility of directed migration or transfer of technological traditions from the Levant to Siberia and Central Asia in this chronological interval requires further discussion, especially regarding the absence of IUP assemblages in the Middle East.

https://www.sciencedirect.com/science/ar...842300132X

Denisovan would help a lot but I can't find much resources for it and I don't think it can be easily and reliably calculated from just f-stats.
This is the paper that introduced D0/1/2 ancestry but it only gives estimates for Oceanians and larger regions with many different groups lumped together so I don't know how widespread it actually is. Andamanese were said to have no Denisovan, and they have overall low levels of archaic. Other Asian pops like Jomon, Hoabinhian and Longlin also have similarly low archaic.
https://www.cell.com/cell/fulltext/S0092...19)30218-1

This is the Salkhit paper. They claim no Denisovan in Ust'-Ishim and any early European, it was before Bacho Kiro but Oase should be included. So yes, it makes it unlikely they come from far in the East, unless the Denisovan admixture is later. There are also a few good figures in the supplements on the relations of the earliest ancients. For moderns I think they only check the Denisovan segments shared with Salkhit and Tianyuan.
https://www.science.org/doi/10.1126/science.abc1166

My impression is that the Ust'-Ishim man's haplogroup is K2a, no? I believe he is positive for SNPs CTS11667 and M2308.

My point was just it's a clearly East Eurasian haplogroup, so I guess that stands even more as K2a should be more specifically East Asian.

OK. But K2a is only defined by two mutations, perhaps 100-200 years. Also, K-Y28299 on Yfull is obviously an AASI lineage, and the MF106925 on FT or 23mofang is found in Vietnam and Thailand, so it could be Hoabinhian derived. If East Asians are "Tianyuan+Onge" with extra drift, their autosomal lineage may form between 40 and 30kya, which is much later than K2a* and Ust-Ishim.

In other words, Ust-Ishim and Oase are both outgroup of the Tianyuan-Onge meta branch, while East Asians are nested in this branch.

[Ebizur]

OK. But K2a is only defined by two mutations, perhaps 100-200 years. Also, K-Y28299 on Yfull is obviously an AASI lineage, and the MF106925 on FT or 23mofang is found in Vietnam and Thailand, so it could be Hoabinhian derived. If East Asians are "Tianyuan+Onge" with extra drift, their autosomal lineage may form between 40 and 30kya, which is much later than K2a* and Ust-Ishim.

In other words, Ust-Ishim and Oase are both outgroup of the Tianyuan-Onge meta branch, while East Asians are nested in this branch.

The phylogeny of Y-DNA haplogroup K2a according to FTDNA is currently as follows:

K2a-M2308 (TMRCA 44,325 [95% CI 52,190 - 37,623] ybp)
*K-V7320 Russia x1 (The Ust-Ishim specimen himself?)
*K-F549 (TMRCA 43,850 [95% CI 51,640 - 37,213] ybp)
**K-Y28394 (TMRCA 43,592 [95% CI 51,341 - 36,991] ybp)
***K-Y28299 (TMRCA 23,291 [95% CI 27,864 - 19,443] ybp) Pakistan (Azad Kashmir) x2, India x6, Pakistan x1, Unknown origin x1
***K-F14963 (TMRCA unknown) Indonesia (Sumatra) x2, Indonesia (Sulawesi) x1, Malaysia x1, Singapore x1, Thailand x1, Unknown origin x1
**K-FTC181 (TMRCA unknown) Thailand x1
**K-MF106925 (TMRCA unknown) Thailand x1, Vietnam x1
**NO-M214 (TMRCA 37,589 [95% CI 42,647 - 32,846] ybp)

China-based 23mofang and TheYtree currently have a more nested structure, with K-MF106925 and NO-M214 sharing the M2311 SNP in contradistinction to K-Y28355=K-Y28394.

As for the ethnolinguistic identities of the members from India, Thailand, and Vietnam, TheYtree provides the following information:

K2-M526 > K-M2308 > K-M2335 (KAL117 Kalueang from Thailand & BST125 SouthernThai_TK)

K2-M526 > K-M2308 > K-M2335 > K-MY1504 > K-Y28300 (ERR445319 Malayalam from India) > K-Y28343 > K-Y28301 (HG03742 from India & ERR445238 Telugu from India; I suspect that these latter two samples may represent the very same individual)

K2-M526 > K-M2308 > K-M2335 > K-M2311 > K-Y86941 (TMRCA 38360 ybp) > K-CTS2230 (TN313 ERR3268500 Htin [TN3] from Thailand)
K2-M526 > K-M2308 > K-M2335 > K-M2311 > K-Y86941 (TMRCA 38360 ybp) > K-F21014 (Mang304 Mang from Mường Tè district, Lai Châu province, Vietnam)

So, the members of K-Y86941 from Thailand and Vietnam both belong to Austroasiatic-speaking minority ethnic groups, but they do not appear to share a recent common ancestor. A different subclade of K2a(xNO-M214) appears to be present among some Tai-speaking ethnic groups in Thailand. Overall, the diversity seems to be concentrated in Southeast Asia.

[CLTVTE]

K2-M526 > K-M2308 > K-M2335 (KAL117 Kalueang from Thailand & BST125 SouthernThai_TK)
K2-M526 > K-M2308 > K-M2335 > K-MY1504 > K-Y28300 (ERR445319 Malayalam from India) > K-Y28343 > K-Y28301 (HG03742 from India & ERR445238 Telugu from India; I suspect that these latter two samples may represent the very same individual)
K2-M526 > K-M2308 > K-M2335 > K-M2311 > K-Y86941 (TMRCA 38360 ybp) > K-CTS2230 (TN313 ERR3268500 Htin [TN3] from Thailand)
K2-M526 > K-M2308 > K-M2335 > K-M2311 > K-Y86941 (TMRCA 38360 ybp) > K-F21014 (Mang304 Mang from Mường Tè district, Lai Châu province, Vietnam)

In China, they do not think that possessing K-M2335, K-MY1504 > K-Y28300, K-M2311 > K-Y86941, K-M2311 > K-Y86941 should define the ethnolinguistic identity of Dravidian, Austroasiatic and Tai-Kadai speakers.

It is thought in China that yDNA O1b1-M95 lineage may be much more ancient, than is usually estimated, and yDNA O1b1-M95 is crucial for the migration of Austroasiatic-speaking populations to India, but it was also observed in China that yDNA O1b1-M95 is the most diversified in the Tai-Kadai Daic-speaking populations. It is consistent with the finding of the basal yDNA O1b-M268 in the Tai-Kadai-related ancient DNA in China in "Human population history at the crossroads of East and Southeast Asia since 11,000 years ago". It corresponds quite well to the Korean-Tai-Kadai-Austronesian linguistic grouping in Jager, 2015 (“Support for linguistic macrofamilies from weighted sequence alignment”). Indeed, starting from AR33K ca. 33000 year ago, the ancient East Asian ancestry started to appear in Northeast China in “Ancient genomes reveal the complex genetic history of Prehistoric Eurasian modern humans”, and yDNA O1b-M268>O1b2-P49-related population should have started to separate in this direction from the homeland of ancient Southern East Asians in the eastern part of Southern China. In the beginning, mtDNA B2-related population should have been diversified, since mtDNA B4b* lineages were claimed to be found in Mexico. If any part of the heritage, left in China, had been mixed with populations of China in the Palaeolithic, e.g. mtDNA D4b1d bearers on the territory of the modern Jiangsu Province, and therefore its ownership became controversial, Chinese citizens can “take” such Paleolithic remains for themselves.

In terms of female lineages, most common East Asian lineages of Sinitic, Tai-Kadai and Austroasiatic populations belong to the same mtDNA branches in “Ancient Mitogenomes Reveal the Origins and Genetic Structure of the Neolithic Shimao Population in Northern China”.

A few modern female lineages in China were inferred to have an affinity to the Ust-Ishim, being remains of populations, distantly related to the Ust’-Ishim individual. Since the Ust-Ishim is quite ancient (his yDNA K2a* was even estimated to separate from other yDNA K2a lineages more than 52000 years ago), such individuals as ancestors of ca. 45400-year-old branch’s yDNA K2a-M2335* KAL117 Kalueang from Thailand and yDNA K2a-M2335* BST125 from Thailand (https://www.theytree.com/tree/K-M2335) had to arrive to Southeast Asia later and distribute the ancestry related to the mentioned “Ust-Ishim-affiliated” female lineages, because it was shown in “The population history of northeastern Siberia since the Pleistocene” that individuals, having an affinity to the Ust-Ishim, live in modern India, Southeast Asia, Papua-New Guinea and even in Australia. The Tai-Kadai speakers, Austroasiatic speakers and even Proto-Austronesians, settling in Taiwan, could either contact Ust-Ishim-affiliated female lineages directly or they could contact such yDNA K2a-M2335* individuals, interacting with Ust-Ishim-affiliated female lineages before. In the linguistic dimension, such contacts could have led to the formation of “Ust-Ishim-affiliated” Tai-Kadai-Austroasiatic linguistic macrocluster of Jager, 2017 (“From words to features to trees: Computing a world tree of languages from word lists”), and other linguistic macroclusters, joined this “Ust-Ishim-affilieated” Tai-Kadai-Austroasiatic linguistic macrocluster either because they were related to Tai-Kadai-Austroasiatic-Austronesian languages (in case of languages of the Sino-Tibetan-Hmong-Mien macrocluster) or because representatives of other clusters contacted either northern Ust-Ishim-related branches (for Eurasian languages) or contacted yDNA K2a-M2335* individuals, who had interacted with Ust-Ishim-affiliated female lineages before and distributed to Southeast Asia (for languages of Papuan and Australian clusters). The interaction of bearers of ancestry, related to mtDNA M “Ust’-Ishim-affiliated” lineages with mtDNA P-related ancestors of Australians on the one hand and the interaction of bearers of ancestry, related to mtDNA M “Ust’-Ishim-affiliated” lineages with ancestors of mtDNA M-related Papuans on the other hand can explain the appearance of the Australian-Papuan-Eurasian macrocluster in Jager, 2017, because the Eurasian members had female ancestors of Ust-Ishim related lineages.

The individuals of yDNA K2a-MY1504 or K2a-Y28299 (https://www.theytree.com/tree/K-MY1504 or https://www.yfull.com/tree/K-Y28299/) had to be quite widely distributed in the western part of Southern China, Southeast Asia and India in the past, reaching Indonesia and adjacent territories at some point. Their rarity in Southeast Asia today makes it clear that they did not contribute much to language families of East Asian descent, whose members settled in Southeast Asia.

The individuals of yDNA K2a-Y86941 (https://www.theytree.com/tree/K-Y86941) also once had mtDNA lineages, characteristic of their population, which are today found in Tai-Kadai, Austroasiatic, Indian and Indochinese populations, but not in Austronesians.

The individuals, carrying mtDNA M21a (for example, the ancestors of the Batek), had to be once distributed from Malaysia to Thailand and the Philippines, from Malaysia to Thailand and to Myanmar. They are considered in China to be the shared substratum for some Tai-Kadai, Austroasiatic, Austronesian languages in the areas, where such populations lived. Their mtDNA M21a has a relative, such as mtDNA M21b, which was likely to be accompanied by differing y chromosome haplogroups (possibly yDNA D-M174) and could have once reached the Japanese Archipelago in the Palaeolithic, being widely distributed in Southeast Asia today. Additionally, mtDNA M21a shares mutation C16256T with mtDNA M2a1a1b1, which is distributed in the Nihali population today, whose language clustered as an outgroup to all Western Eurasian languages in Jager, 2017.

Nahali
The c. 24 % substrate found in the Nahali (Nihali) language of Central India has many words for plants (and animals) that cannot be linked to IA, Drav. or Munda. A few prominent ones (including some imported plants) include the following words. 146 āndij ‘root like sweet potato’, baḍágo ‘guava’ (cf. 9125 bádara— n. ‘fruit of the jujube tree?), baru 'mulberry' (cf. various trees CDIAL 5872), buṭu 'kind of grass', bhed(a)rā 'potato' (!), bōy 'grass, fodder' (cf. DEDR 4535, grasses), ḍotako 'edible root', dhāwrā 'gum tree', dhongāri 'type of grass', gugudo 'edible root', hardo 'tumeric', jiryāngā 'tomato'(!), jodu/jūḍ 'bamboo', jhāpon 'mushroom', khila 'parched rice', khude 'gourd', lubā 'incense', malkā 'pea (pod)', māyko 'mahua tree' (cf. DEDR 4772?), óhan 'mortar (with pestle)', oró 'millet, jawar' (cf. § 3.2.3 ōḍī 'wild rice'?), phellyā 'groundnut', phendrā 'vine', raymonyā 'wild thorny bush', riṭhā 'soap nut', sokorā 'bread', siḍu 'mahua wine', sundu 'pod for beans', chāgā 'variety of thorny grass', chepiyā 'variety of grass', chunco 'a vegetable', chunḍu 'bean' (cf. CDIAL 4856?), tamāko'o 'tomato'(!), tāmku 'tobacco'(!), tāndur 'rice, cooked rice', ṭó 'ear (of corn)'. Nahali gele 'maize' is from Korku, gohũ 'wheat' from IA, and many other common plants are loanwords from Munda, Drav. or IA.

[ChrisR]

In G25, Ust'-Ishim appears on the West-East Eurasian cline where it is leaning towards the Southern East Asian/Siberian cluster:
(X = PC2, Y = PC11, Z = PC25)

I would like to ask how to create such a 3D distribution with certain G25 coordinate axes?