okarinaofsteiner's East Eurasian GEDmatch megathread

[MNOPSC1b]

For better visualization and understanding, I marked the cardinal directions on this PCA chart

[Tsakhur]

I think you misread the graph (I assume you're talking about this graph)



On this graph the horizontal line is the north-south axis, whereas the vertical line is the east-west axis. The Wa is slightly northern-shifted compared to Thai and Cambodian, which is understandable since they live in Southwest China and likely have more interactions with Sino-Tibetan speakers. However in terms of the east-west axis Wa, Thai, and Cambodian are very similar and the difference is negligible. Thai is slightly more western-shifted than the other two, though not by much.

And from this graph, I fail to see that there's a large Sinitic influence in Thailand. The position of the Thai on the graph (close to Cambodian and Wa) does not suggest a strong Sinitic influence. Although this could be due to sampling, perhaps the Thai samples on the graph are drawn from native Thais rather than from Chinese Thais.

Yes that's the figure I'm referring to. Correct, I believe I misuse the word "west-east" here. It's "north-south" cline which I am inferring in regards to relative position of the Wa, who are slightly more northern-shifted as you have highlighted, compared to Thais and Khmers.

My bad. You are right. I was initially preoccupied with the fact that many Thai G25 samples that I have observed possessed a lot of Sinitic ancestry so didn't notice the position of the Thai when making that statement. But yes, the Thai samples here are very likely from the local Siamese of Central and Eastern regions of Thailand who lacked Han/have very little admixture rather than Chinese/Chinese admixed ones who seem to be exclude from this data. What I wonder though is if the slight northern-shifted of the Thai compared to the Khmers come from the minor Dai/Tai-Kadai-like affinity that the former might possessed.

[observer_t]

Yes that's the figure I'm referring to. Correct, I believe I misuse the word "west-east" here. It's "north-south" cline which I am inferring in regards to relative position of the Wa, who are slightly more northern-shifted as you have highlighted, compared to Thais and Khmers.

My bad. You are right. I was initially preoccupied with the fact that many Thai G25 samples that I have observed possessed a lot of Sinitic ancestry so didn't notice the position of the Thai when making that statement. But yes, the Thai samples here are very likely from the local Siamese of Central and Eastern regions of Thailand who lacked Han/have very little admixture rather than Chinese/Chinese admixed ones who seem to be exclude from this data. What I wonder though is if the slight northern-shifted of the Thai compared to the Khmers come from the minor Dai/Tai-Kadai-like affinity that the former might possessed.

probab.PNG

When modeled with the same components, Thai and Cambodian interestingly have the same % of Upper Yellow River/Qijia. The primary difference here does seem to be that Thai have a greater % Dai compared to Cambodian (who have a greater % LAO_LN_BA). So although this is probably not the actual proportion of Dai input in Thais/Cambodians it does seem their slight northern shift is probably due to Dai, rather than direct Chinese or Tibeto-burman admix

[Tsakhur]

Attachment 54667

That Baotou individual has a remarkably low Slab Grave/Ulaanzukh with a high Oroqen. I think generally when a population has affinities to groups that live in the inland Manchuria/eastern Mongolia (ie. Hulunbuir) region, they tend to use one of CHN_Amur_EN, Nanai, or Oroqen. Populations that are more northern geographically use CHN_Amur_EN and/or Nanai, while populations that are more southern geographically use Oroqen (see my earlier PCA). This is probably why Buryats, Khakass, Tuvinians, and Altaians have CHN_Amur_EN while Mongolia-Mongols and Kazakhs have Oroqen instead

Baotou is also not too distant from Ulanqab, where Yumin and then Miaozigou were located. Yumin seems to be mostly similar to Slab Grave, but then Miaozigou is mostly Upper Yellow River/Tibeto-burman like (with some minor western Manchuria-like input in the form of Amur_EN). It would seem that populations in this region regained Slab Grave at some point (or alternatively Slab Grave-associated and Miaozigou-like populations lived separately), but then received a heavy flow of Oroqen-like input later

Wow the Baotou individual also score high Yellow_River. Your deductions are plausible in my opinion regarding the geography of that particular population influencing what ancestral lineages they are more likely to descended from.

Makes me wonder if the Barga, Uzemchin, Dariganga and other Mongolics from Dornod, Sukhbaatar and easternmost part of Dornogovi will be scoring high Amur_EN, Nanai or Oroqen along with substantial Yellow River compared to the Khalkha majority to the west of them. Unfortunately, there is a lack of autosomal data from that regions of the Mongolian Plateau. Won't be surprised if the Mongols from those two/three easternmost provinces will be genetically resembled the Daur of Inner Mongolia and Heilongjiang, having very little to almost zero Western Eurasian affinity and elevated Oroqen/Nanai and Yellow River components.

Off topic, I do realize phenotypical discussion isn't really allow here, but referring to your previous post of how you noticed some Western Mongols and Altaians have a more "Northern-Central Chinese" look which likely correlated with their significant Ulaanzukh/Slab Grave blood, I think you have a point here.

 

Good examples are Tsakhiagiin Elbegdorj and Yumjaagiin Tsedenbal, the former leaders of Mongolia also have those phenotypes you are referring to and they are both Oirats (Zakhchin and Dorbet subtribes) from Khovd and Uvs in Western Mongolia.

Anyway back to the main subject, there are some additional Inner_Mongolian samples I simulated from their gedmatch kits. The first one seem to be from Hailar in Hulunbuir and is Daur-like or can be modelled as a mix between mostly Buryat+Xibo/Japanese/Northern Han while the second one is probably also from Northern parts of Inner Mongolia and is mostly Buryat or Khamnegan-like or a mix between mostly Oroqen+ Paleosiberian (Yukaghir) /Samoyedic (Selkup)/Yeniseian (Ket) populations according to HarappaWorld. The rest of the three Inner Mongolians have a more generic Mongola/Mongol_Hulunbuir or Xibo-like profile with two being more Western-shifted than average:

Btw I simulated the Inner Mongolian1 and Inner Mongolian2 from both HarappaWorld and Eurogenes K13 and there seems to be huge differences in the proportions of the Slab Grave/Ulaanzukh and Tungusic ancestries that they score depending on the two calculators. Thus, I will include the sims of the two individuals from both of the two calcs.

I would suggested adding Krasnoyarsk_BA in your run. Buryats seems to have Nganasan-related affinity which will help improve the fits. The Buryat-like/Khamnegan-like Inner_Mongolian_2 sample also requires Nganasan/Krasnoyarsk ancestry to get good fits.

Inner_Mongolian1-HW-sim_scaled,0.0376,-0.4028,0.0615,-0.0406,-0.0395,-0.0266,0.0163,0.0185,0.0012,0.0099,-0.0421,-0.0029,-0.002,0.0008,0.0031,-0.0025,0.0006,-0.0002,0.0009,0.01,-0.009,-0.0091,-0.0158,-0.0021,-0.0007
Inner_Mongolian1-K13-sim_scaled,0.0353,-0.4007,0.0597,-0.0385,-0.0409,-0.0262,0.013,0.0187,0.0012,0.0118,-0.0328,-0.0018,-0.0022,0.0038,-0.0004,-0.0016,-0.0048,0.0051,0.0063,0.0079,-0.0012,-0.0146,-0.0044,-0.0031,0.0012

Inner_Mongolian2-HW-sim_scaled,0.0411,-0.3916,0.0816,-0.0302,-0.0688,-0.0418,0.0168,0.0234,0.0071,0.0123,-0.0163,0.0005,-0.005,0.0036,-0.0004,-0.007,-0.0009,-0.0002,0.0055,0.0107,-0.0144,-0.0085,-0.0174,0.0025,-0.0017
Inner_Mongolian2_K13-sim_scaled,0.0385,-0.395,0.0827,-0.0251,-0.0707,-0.0418,0.0181,0.0237,0.0058,0.0127,-0.0128,0.0005,-0.0037,-0.0019,-0.0041,-0.0064,-0.0016,0.007,0.0116,0.0104,-0.0019,-0.0164,-0.007,-0.0005,0.0075

Inner_Mongolian3-K13_sim_scaled,0.0351,-0.4019,0.0385,-0.0473,-0.0017,-0.0082,0.01,0.0092,-0.0026,0.0036,-0.0562,-0.0014,0.0056,-0.0044,-0.0009,-0.0008,-0.0017,-0.0004,0.0012,0.0045,-0.0014,-0.0041,-0.0014,0.0018,0.0006
Inner_Mongolian4-K13-sim_scaled,0.0353,-0.4001,0.0429,-0.044,-0.0085,-0.012,0.0101,0.0116,-0.0026,0.005,-0.0538,-0.0014,0.0043,-0.0054,-0.001,0.0004,0,-0.0004,0.0021,0.0039,-0.0045,-0.0057,-0.0033,0.0008,0.0022
Inner_Mongolian5-K13-sim_scaled,0.0328,-0.4242,0.0441,-0.0467,-0.0045,-0.0115,0.0087,0.0064,-0.0022,0.0059,-0.0546,-0.0027,0.0066,-0.0044,-0.0028,-0.0022,-0.0023,0.0009,0.0042,-0.0004,0.0031,-0.0068,-0.0018,0.0031,-0.0017

Furthermore, here are Golden_Horde Period Mongol Warrior and two Daur gedmatch kits that I simulated from HarappaWorld:

DA28_Golden_Horde-K13-sim_scaled,0.0392,-0.42,0.1047,-0.0237,-0.0916,-0.0545,0.0233,0.0286,0.0134,0.0138,0.0114,0.0039,-0.0028,-0.0017,-0.0104,-0.0142,-0.0056,0.0109,0.0175,0.0116,0.0135,-0.019,0.0022,-0.001,0.0087
Daur1-HW-sim_scaled,0.031,-0.4334,0.0528,-0.0518,-0.0179,-0.0161,0.0135,0.0181,0.0015,0.0123,-0.053,-0.0009,0.0021,0.0063,-0.0031,-0.0046,-0.0049,0.0005,0.0009,0.007,-0.0013,-0.01,-0.0122,0.0004,0.0026
Daur2-HW-sim_scaled,0.0326,-0.4194,0.0483,-0.0476,-0.015,-0.017,0.0162,0.0153,-0.0002,0.0094,-0.0496,-0.0034,0.0042,0.0017,-0.0002,0.0015,-0.0008,0.001,0.0025,0.0017,0.0011,-0.0054,-0.0099,0.0031,-0.0015

And another Mongol_Baotou individual who is majority Ulaanzukh/Slab Grave with lower Tungusic and Yellow River affinity:

Mongol_Baotou_(Mid-Mongolia_Profile):55381809296824,0.03187,-0.390979,0.059208,-0.04845,-0.020004,-0.015339,0.009165,0.008769,0.003681,0.01385,-0.054887,-0.006744,0.002379,0.007844,-0.002986,-0.002254,-0.006128,-0.004054,0.005028,-0.001,-0.005615,0.00643,-0.012325,-0.006025,0.002754

Here are my modelling of them:

[utu]

For better visualization and understanding, I marked the cardinal directions on this PCA chart

Yes, the explanation is quite insightful. This is the original meaning of this picture. The genetic effects given by the inland population of the Austroasiatic-like ancient Mekong to the indigenous peoples of Yungui and Lingnan are obviously not small, and the results indicated by each data are also accurate. but, the indigenous peoples of ancient Lingnan (Guangdong and Guangxi Hainan) are extremely similar to the primitive Austronesian ancient Fujian This is also the main source of Zhuang and Hlai people today, The essence of their cultural customs belongs to the ancient marine ethnic group Baiyue. For example, the cultural relics of bronze drums and Bronze Bucket displayed in the museum depict a group of groups wearing feather headgear boating on the river. Some other patterns may prove ship be at sea, because under the water at the bottom of the ship are painted with many animals from the sea. In addition, there are relevant archaeological evidence, As a famous seaport in ancient and modern times, Guangzhou has unearthed the remains of a shipyard in the Qin Dynasty, so the evidence has become self-evident. The modern Tai-kradai family was formed after the two ancient ancestors merged here. However, in their culture, the culture related to the Proto-Austronesians is more obvious, while the cultural style of the Austroasiatic-like has been weakened to a deeper extent
The site of an ancient shipyard

Bronze Drum Pattern

[East-Asia]

Those Dusun and Igorot should have quite a lot of "Cordilleran" autosomal component which is named after some Phillipine natives. This component was estimated to separate from Tibetans more than 32000 years ago, and "Cordilleran component" migrated to the Phillipines prior to Austronesians whose ancestors were more autosomally Chinese-like.

[MethCat]

Thanks to this post, I was able to make this-
https://pastebin.com/agApNDUr

Which combined with this for LAO_Hoabinhian, this for Jomon, and this for S_AASI_Sim... resulted in me learning how to properly use G25 Vahaduo

 

With Tujia as a source pop

Without Tujia (CHN_Upper_Yellow_River_LN, Dai, Igorot, and Nanai only)

Updated version that includes Wa and Nivkh (fit for Japanese is slighly less bad than with only Nanai)


UYR_LN = Himalayan (Tibeto-Burman)
Dai = Tai-Kradai
Igorot = Austronesian
Wa = Austroasiatic
Nanai = Tungusic
Nivkh = NEA "Coastal Ghost"-like
Yakut = Siberian

Iranian_Zoroastrian = Sogdian-like West Eurasian component
Lithuanian_ZR = more generic West Eurasian component

Jomon = "Basal" NEA HG? for non-JP/KR populations

LAO_Hoabinhian, Papuan, and AASI_Sim should be self-explanatory.

Is there any specific reason why Wa is chosen as a representative of AA related ancestry as opposed to the Mon for example? For Thais, Mon should represent their real ancestry far more accurately than Wa. Mlabri perhaps?

Genuinely curious as you seem to know what you're talking about.

[okarinaofsteiner]

 

With Tujia as a source pop

Without Tujia (CHN_Upper_Yellow_River_LN, Dai, Igorot, and Nanai only)

Updated version that includes Wa and Nivkh (fit for Japanese is slighly less bad than with only Nanai)


UYR_LN = Himalayan (Tibeto-Burman)
Dai = Tai-Kradai
Igorot = Austronesian
Wa = Austroasiatic
Nanai = Tungusic
Nivkh = NEA "Coastal Ghost"-like
Yakut = Siberian

Iranian_Zoroastrian = Sogdian-like West Eurasian component
Lithuanian_ZR = more generic West Eurasian component

Jomon = "Basal" NEA HG? for non-JP/KR populations

LAO_Hoabinhian, Papuan, and AASI_Sim should be self-explanatory.

Is there any specific reason why Wa is chosen as a representative of AA related ancestry as opposed to the Mon for example? For Thais, Mon should represent their real ancestry far more accurately than Wa. Mlabri perhaps?

Genuinely curious as you seem to know what you're talking about.

Mon are known to be significantly Indian-mixed. Although the Thai model has a surprisingly good fit with only Dai, Igorot, and Wa as the East Eurasian components, compared to modeling Han Chinese subgroups with Upper Yellow River LN (Qijia) lol.

I added Mlabri to the "Everything Bagel" model I posted to see if I could get less-bad fits for the Indonesian reference populations. It seems like Mlabri is a better fit for the Austroasiatic ancestry in more geographically southern SEA groups with more Hoabinhian ancestry, while Wa is a better for for the AA-like component in more "northern" Mainland SEA groups that also have less Hoabinhian ancestry.

"Everything Bagel" version with both UYR_LN (Qijia) and YR_LN (Longshan), and also both Mlabri and Wa to get better fits for the Indonesian target pops.

https://i.imgur.com/wqFc2LI.png

UYR_LN = Generic Trans-Himalayan (Tibeto-Burman)
YR_LN = Sinitic-specific Trans-Himalayan
Dai = Tai-Kradai
Igorot = Austronesian
Wa = Austroasiatic 1 ("northern?")
Mlabri = Austroasiatic 2 ("southern?")
Nanai = Tungusic
Nivkh = NEA "Coastal Ghost"-like
Amur_EN = Proto-Northern EA(?)
Yakut = Siberian

Jomon = "Basal" NEA HG? for non-JP/KR populations
Mixtec = Amerindian-like, ANE-mixed Proto-EA component

Iranian_Ganj_Dareh_N = Dravidian-like West Eurasian component
Yamnaya_RUS_Samara = Indo-European-like West Eurasian component
Jordanian = Arab-like West Eurasian component
French_Paris = more generic West Eurasian component

LAO_Hoabinhian, Leang_Panninge, Papuan, and AASI_Sim should be self-explanatory. Yoruba is a generic non-Eurasian outgroup.

[CodenameJungle]

"Everything Bagel" version with both UYR_LN (Qijia) and YR_LN (Longshan), and also both Mlabri and Wa to get better fits for the Indonesian target pops.




UYR_LN = Generic Trans-Himalayan (Tibeto-Burman)
YR_LN = Sinitic-specific Trans-Himalayan
Dai = Tai-Kradai
Igorot = Austronesian
Wa = Austroasiatic 1 ("northern?")
Mlabri = Austroasiatic 2 ("southern?")
Nanai = Tungusic
Nivkh = NEA "Coastal Ghost"-like
Amur_EN = Proto-Northern EA(?)
Yakut = Siberian

Jomon = "Basal" NEA HG? for non-JP/KR populations
Mixtec = Amerindian-like, ANE-mixed Proto-EA component

Iranian_Ganj_Dareh_N = Dravidian-like West Eurasian component
Yamnaya_RUS_Samara = Indo-European-like West Eurasian component
Jordanian = Arab-like West Eurasian component
French_Paris = more generic West Eurasian component

LAO_Hoabinhian, Leang_Panninge, Papuan, and AASI_Sim should be self-explanatory. Yoruba is a generic non-Eurasian outgroup.

Very surprised that Tibetan_Gannan did not pick up any AASI/SAHG related admixture but did pick up Yamnaya, Ganj Dareh and Jordanian even for a total of 9% West Eurasian. Thought they would defo score some SAHG whats up with that

[okarinaofsteiner]

Very surprised that Tibetan_Gannan did not pick up any AASI/SAHG related admixture but did pick up Yamnaya, Ganj Dareh and Jordanian even for a total of 9% West Eurasian. Thought they would defo score some SAHG whats up with that

Tibetan_Gannan is the only Tibetan G25 subpopulation that doesn't score any AASI. Even the Qiang populations and Naxi/Yi score some.