okarinaofsteiner's East Eurasian GEDmatch megathread

[Tsakhur]

I don't want to detract from your work

but it seems like some of the samples from Jeong et al. (2020) dataset cannot be confidently attributed to the ethnic grouping many try to assume they are to be, due to nomadism.
For example Khitan period samples even if the were from burial sites during the Khitan Liao dynastic rule and influence over the Mongolian Plateau, how can one decide they are actually of Khitan origin as opposed to others assigned rulership under the Khitans

These are the individual samples utilized to create the Dodecad K12b averages for Mongolics: https://www.theapricity.com/forum/sh...sky-et-al-data

Here are the most Eastern-shifted ones from the study: there are probably those with lower Western Eurasian if there are more sampling.

Target: Khamnegan1_Zabaykalsky_region_Duldurginsky_Akshins ky_Kyrinsky_Aga_districts_Russia:BUR-238
Distance: 206.0766% / 2.06076561
34.6 Nganasan
31.4 Nivh
21.6 Korean
6.6 Avar
5.8 Han_Central

Target: Buryat_Irkutskaya_Ehirit_Irkutsk_region_Bayanday_B okhansky_Olkhonsky_Ekhirit-Bulagatsky_districts_Russia:BUR-706
Distance: 177.2912% / 1.77291241
51.0 Nivh
28.6 Nganasan
12.6 Korean
7.8 Avar


Target: Mongol_Khalkh_Sainnoyon_historical_aimak_Sainnoyon _Mongolia:MON-0344
Distance: 135.6585% / 1.35658527
36.4 Nivh
23.2 Korean
21.8 Nganasan
9.2 Han_Central
7.0 Avar
2.4 Latvian

Target: Mongol_Uuld_Mongolia:MON-558
Distance: 297.4386% / 2.97438618
43.2 Nivh
29.4 Korean
13.8 Nganasan
6.2 Latvian
4.0 Avar
3.4 Han_Central

Target: Mongol_Minorities_Mongolia:MON-0426
Distance: 189.8432% / 1.89843224
33.8 Nivh
23.6 Korean
21.2 Nganasan
10.6 Han_Central
7.2 Latvian
3.6 Avar

Most Eastern-shifted Tuvans and Todzins (a subgroup of Tuvans):

Target: Tuvinian_Tuva_Republic_Barun-Khemchik_Tere-Hol_Tandy_Chaa-Hol_Erzinsky_Russia:TUV-004
Distance: 245.5921% / 2.45592093
43.8 Nivh
39.8 Nganasan
6.6 Avar
5.8 Latvian
4.0 Korean

Target: Todzin_Tuva_Republic_Todzinsky_district_Russia:TUV-199
Distance: 264.7800% / 2.64779953
47.6 Nganasan
42.8 Nivh
9.6 Latvian


Mongol_Minorities group aren't include in the Dodecad K12b averages spreadsheet. But based on the fact that even the most Eastern-shifted individual still have approx 10-11% Western Eurasian, its likely they are from somewhere in Central-Western Mongolia than the Far Eastern parts of the country.

Looks like many Mongolian Plateau groups and surrounded areas such as Buryatia have substantial Turkic ancestry which gives them this Western-related admixture.

[CodenameJungle]

A nearly 100% Japanese 23andme result posted on reddit
99.99 percent japanese .jpg
https://www.reddit.com/r/23andme/com...been_to_japan/

[East-Asia]

The medieval label "Western Eurasian" can include haplogroups that originated in East Asia

"B5b seems to have a wide geographic distribution. B5b5 [KF540694] comes from Hakka, a Sinitic speaking native group from Taiwan [56], while B5b4 [KF849981] is (mostly likely) from Han Chinese [57]"





----------------------------------------------------------------------------------------------------------------------

A nearly 100% Japanese 23andme result posted on reddit
99.99 percent japanese .jpg
https://www.reddit.com/r/23andme/com...been_to_japan/

There is no reason to pretend that Japanese allegedly do not include a component related to Han Chinese.

Ancestry informative SNP panels for discriminating the major East Asian populations: Han Chinese, Japanese and Korean

Cheng-Min Shi (1), Qi Liu (1,3), Shilei Zhao (1,3), Hua Chen (1,2,3)
First published: 26 April 2019 https://doi.org/10.1111/ahg.12320Citations: 8

1CAS Key Laboratory of Genomic and Precision Medicine, Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing, China
2CAS Center for Excellence in Animal Evolution and Genetics, Chinese Academy of Sciences, Kunming, China
3University of Chinese Academy of Sciences, Beijing, China

Ancestry informative markers play an important role in medical genetics and forensic analyses. Several ancestry informative SNP panels have been developed and validated that can differentiate global populations into continental or major regional groups. These global panels have served as good first-tier genetic markers; however, their performance in discriminating populations within regions appears unsatisfactory. To boost ancestry inference for regional populations, second-tier panels with more refined discrimination power among subpopulations within each of the regions need to be developed. In East Asia, Han Chinese, Japanese, and Korean show highly similar externally visible characteristics and are genetically closely related. Reliable ancestry informative genetic markers appear invaluable in discriminating these populations. In the present study, we compiled a genome-wide SNP dataset composing of 317,439 clean SNPs for a total of 1101 unrelated individuals from Han Chinese (817), Koreans (184), and Japanese (100). From this starting dataset, we developed a set of four nested ancestry informative SNP panels including 36, 59, 98, and 142 SNPs, respectively. The results of cross-validation tests indicate that these panels can discriminate the Chinese Han, Japanese, and Korean populations with overall average accuracies ranging from 90% to 99%. In the further performance assessments, these panels also manifested high sensitivity and specificity. In combination with the first-tier global panels, these second-tier panels would contribute to medical genetics and forensic research in East Asia.

[okarinaofsteiner]

Graphics from Razib Khan's substack:



TL;DR of all the conjecture and wordspam on East Eurasian population genetics/history. The "conventional wisdom" among Chinese diaspora former Anthroscape members is that Koreans have some "Southern NE Asian" ancestry, just less than Northern Han who obviously have less than Southern Han. Even Yue-speaking Han seem to have more "Northern NEA" than "Southern NEA" ancestry, if Wuzhuangguoliang and Liangdao are appropriate proxies for "Northern NEA" and "Southern NEA".

---

"Proportions of two genetic lineages that diverged 19,000 years ago in North China (blue) and South China (orange) in Han Chinese regional populations and ethnic minorities. Adapted from Genomic Insights into the Formation of Human Populations in East Asia"

Interesting that Dai, Li, and Ami (Taiwanese aborigine/Gaoshan) are all roughly the same ratio of YellowRiver_LN and Liangdao... Kinh (Vietnamese) seems slightly more "Northern" than Dai (which is to be expected) but also more "Southern" than Zhuang.

---

Genetic distances based on pairwise Fst between the Southern Han and selected populations. The smallest genetic distance, between Northern and Southern Han, is standardized to 1.

Note that all of the farmer-descended "East Asian" populations are within 10 of Southern Han... Uyghur is the next closest at 14.4, then Yakut at 15. The order of the Southeast Asian groups' distances from Southern Han (red asterisk is directly on Jiangxi) suggests that Hoabinhian-like ancestry outweighs differences among "Southern NEA" subbranches (Austronesian, Daic, etc). Southern Han being almost equidistant from Koreans and Dai makes sense in terms how Fujian-like (in terms of northern/southerness) Han score on MDLP K23b ancestry components.

[okarinaofsteiner]

Lhoba (珞巴) results from Arunachal Pradesh

FTDNA: 84% Northern Chinese, 16% Myanmar

GEDmatch MDLP K23b:
52.77% S_EA, 30.36% T_A, 9.51% AN

I am actually from Arunachal Pradesh India. People from my ethnic group reside both in Indian controlled Arunachal Pradesh and in Tibet, China . In China we are categorised as 珞巴, an ethnic minority. After doing some research I found my ancestors probably migrated from the Neolithic Majiayao culture 馬家窯文化 which was an offshoot of the Yangshao culture 仰韶文化. Majiayao culture is where Tibeto burman language and ethnic groups is said to be been formed, which in turn was derived from the earlier Proto-sino Tibetan Yangshao culture.

Admixture analysis on gedmatch show 52.77 % SE Asian, which basically means Tibeto-burman. 30.36 % Tungus Altaic is on the same level found in northern Han. 9.5 % percent Austronesian means unlike Han Chinese who moved to coastal areas and intermarried with Austronesians , my ancestors didn’t intermarry with austronesians too much.

[utu]

Graphics from Razib Khan's substack:



TL;DR of all the conjecture and wordspam on East Eurasian population genetics/history. The "conventional wisdom" among Chinese diaspora former Anthroscape members is that Koreans have some "Southern NE Asian" ancestry, just less than Northern Han who obviously have less than Southern Han. Even Yue-speaking Han seem to have more "Northern NEA" than "Southern NEA" ancestry, if Wuzhuangguoliang and Liangdao are appropriate proxies for "Northern NEA" and "Southern NEA".

---

Interesting that Dai, Li, and Ami (Taiwanese aborigine/Gaoshan) are all roughly the same ratio of YellowRiver_LN and Liangdao... Kinh (Vietnamese) seems slightly more "Northern" than Dai (which is to be expected) but also more "Southern" than Zhuang.

---

Note that all of the "East Asian" populations are within 10 of Southern Han... Uyghur is the next closest at 14.4. The order of the Southeast Asian groups' distances from Southern Han (red asterisk is directly on Jiangxi) suggests that Hoabinhian-like ancestry outweighs differences among "Southern NEA" subbranches (Austronesian, Daic, etc). Southern Han being almost equidistant from Koreans and Dai makes sense in terms how Fujian-like (in terms of northern/southerness) Han score on MDLP K23b ancestry components.

You don't understand that the benchmark he uses here is Guangzhou, a big city in the Pearl River Delta. If you choose the sample from the rural area of western Guangdong in Guangxi, the situation will be different. You will find that there is no difference between Yue langguage Han and Tai-kadai-Zhuang No difference So what needs to be clarified is. These northern elements are the gene exchange from the north to the south in the Stone Age, but it ended in the last two thousand years of history. The ancestors of the modern south are mainly indigenous people. The southern people are not northerners, they are still authentic southerners For example, in this picture, Tai-katai are very close to the Austronesian language

[okarinaofsteiner]

You don't understand that the benchmark he uses here is Guangzhou, a big city in the Pearl River Delta. If you choose the sample from the rural area of western Guangdong in Guangxi, the situation will be different. You will find that there is no difference between Yue langguage Han and Tai-kadai-Zhuang No difference So what needs to be clarified is. These northern elements are the gene exchange from the north to the south in the Stone Age, but it ended in the last two thousand years of history. The father of the modern south is mainly indigenous people. The southern people are not northerners, they are still authentic southerners

For example, in this picture, Tai-katai are very close to the Austronesian language

 

Except for Li/Hlai, the Daic groups all have more of the northern component than any of the Austronesian groups.

After seeing how Chinese adoptees from westernmost Guangdong score on GEDmatch I don't think western Guangdong Han are uniformly Zhuang-like (they exhibit the same Hakka-like vs Zhuang-like split I saw among Cantonese-name/HK samples in my original dataset), even if all Guangxi Han are Zhuang-shifted relative to other speakers of SW Mandarin/Xiang/Yue from neighboring provinces. Interesting that TW_Hakka is slightly more northern than Han_Guangdong but slightly more southern than TW_Minnan; the differences among them and Han_Guangdong are pretty small but still noticeable.

I'm guessing Razib used the G25 reference populations, which I don't know very much about. It would make sense for Han_Guangdong to be from Guangzhou and Han_Fujian to be from Fuzhou, that would explain why the latter is relatively northern-shifted.

[utu]

Except for Li/Hlai, the Daic groups all have more of the northern component than any of the Austronesian groups.

After seeing how Chinese adoptees from westernmost Guangdong score on GEDmatch I don't think western Guangdong Han are uniformly Zhuang-like (they exhibit the same Hakka-like vs Zhuang-like split I saw among Cantonese-name/HK samples in my original dataset), even if all Guangxi Han are Zhuang-shifted relative to other speakers of SW Mandarin/Xiang/Yue from neighboring provinces. Interesting that TW_Hakka is slightly more northern than Han_Guangdong but slightly more southern than TW_Minnan; the differences among them and Han_Guangdong are pretty small but still noticeable.

I'm guessing Razib used the G25 reference populations, which I don't know very much about. It would make sense for Han_Guangdong to be from Guangzhou and Han_Fujian to be from Fuzhou, that would explain why the latter is relatively northern-shifted.

I read a data paper. The content of AN or AA in TK nationality is several times more than you think, and ST is the least, Are you right?, Maybe Guangdong has been washed away by the foreign population. Anyway, the appearance of Guangdong and Han is too different. It is a unique beauty of the south, which is completely different from the northerners,heehee I guess the individual sample selection in western Guangdong is caused by foreign origin It's probably not pure local bloodCarefully observe the details of this PC picture. Han people in the south, especially in Guangdong and Guangxi, and HM. TK. AN are clustered.Of course, the nearest AN family is TK.If Hlai is added above, the gene effect related to Austronesian will be stronger and more prominent

[Tsakhur]

Graphics from Razib Khan's substack:



TL;DR of all the conjecture and wordspam on East Eurasian population genetics/history. The "conventional wisdom" among Chinese diaspora former Anthroscape members is that Koreans have some "Southern NE Asian" ancestry, just less than Northern Han who obviously have less than Southern Han. Even Yue-speaking Han seem to have more "Northern NEA" than "Southern NEA" ancestry, if Wuzhuangguoliang and Liangdao are appropriate proxies for "Northern NEA" and "Southern NEA".

---

Interesting that Dai, Li, and Ami (Taiwanese aborigine/Gaoshan) are all roughly the same ratio of YellowRiver_LN and Liangdao... Kinh (Vietnamese) seems slightly more "Northern" than Dai (which is to be expected) but also more "Southern" than Zhuang.

---

Note that all of the "East Asian" populations are within 10 of Southern Han... Uyghur is the next closest at 14.4. The order of the Southeast Asian groups' distances from Southern Han (red asterisk is directly on Jiangxi) suggests that Hoabinhian-like ancestry outweighs differences among "Southern NEA" subbranches (Austronesian, Daic, etc). Southern Han being almost equidistant from Koreans and Dai makes sense in terms how Fujian-like (in terms of northern/southerness) Han score on MDLP K23b ancestry components.

What is the source behind the Manchurian label: Oroqen, Nanai/Hezhen? No way, its Manchu as they are literally identical to Northern Han Chinese nowadays.

The Mongol reference are probably the Mongola or some other Inner Mongolian groups. Outer Mongolians should be more distant genetically.

[okarinaofsteiner]

What is the source behind the Manchurian label: Oroqen, Nanai/Hezhen? No way, its Manchu as they are literally identical to Northern Han Chinese nowadays.

The Mongol reference are probably the Mongola or some other Inner Mongolian groups. Outer Mongolians should be more distant genetically.

As someone who has partial Manchu ancestry that shows up in my DNA Land and 23andMe v6.0 results (scored enough Siberian to not be classified as trace), I can assure you the bolded part isn't true.

Reference population is probably Oroqen. I think Mongol could be from either region, although there are more ethnic Mongols in the PRC than in Mongolia itself.

Are you right?, Maybe Guangdong has been washed away by the foreign population. Anyway, the appearance of Guangdong and Han is too different. It is a unique beauty of the south, which is completely different from the northerners,heehee I guess the individual sample selection in western Guangdong is caused by foreign origin It's probably not pure local bloodCarefully observe the details of this PC picture. Han people in the south, especially in Guangdong and Guangxi, and HM. TK. AN are clustered.Of course, the nearest AN family is TK.If Hlai is added above, the gene effect related to Austronesian will be stronger and more prominent

The 81 Guangdong adoptees (in green) were almost all from Maoming (4), Yangjiang (21), and Zhanjiang (53). There's one super Northern (almost Korean-like but not quite) outlier from Foshan, one Sichuan/Jiangxi-like outlier from Zhuhai, and one Yunnan/Hubei-like outlier from Zhanjiang. The rest are all within the expected range for Guangdong, with a couple Zhanjiang outliers that are extra SEA-shifted or score Southern Vietnamese-levels of the Hoabinhian component.