okarinaofsteiner's East Eurasian GEDmatch megathread

[alienation]

also i am wondering if there is also a northwest vs southeast kind of “divide” in Shandong that can also be picked up among prefectural level communities in the northeast. Maybe the signals that show Dalian to be different from Shenyang to be somewhere along these lines as well, rather than direct immigration from individual provinces

[Tsakhur]

No way Han_Central is from Henan or anywhere else on the North China Plain, if Tujia is modeled as 100% Han_Central with an okay distance. The 2-way models of Miaozu and She also suggest Han_Central is closest to a relatively northern-shifted Southern Han subgroup. I'd guess something like Han_Hubei or possibly northernmost Hunan?

From what you've shown, Han_Central peaks in Tujia, a Tibeto-Burman group that lives close to Miao. My guess is Han_Central=Han_Hubei. Hubei is home to Han, Tujia and Miao groups at the same time. I've seen Han_North somewhere else which was the same as Han_Shanxi. It's probably still the same case here. She and Miao are Lower and Middle Yangtzean, respectively, they have little to do with Central Plains/North China Plains if we consider Shandong Longshan Late Neolithic as the ultimate "Central Plains" proxy for now:

Distance to: CHN_Yellow_River_LN
0.01972074 Han_Shandong
0.02095711 Han_Henan
0.02373198 Han_Shanxi
0.02805565 Han_Jiangsu
0.03422704 Han_Shanghai
0.03553201 Tibetan_Xinlong
0.03621028 Han_Zhejiang
0.03675140 Qiang_Danba
0.03729584 Korean
0.04343545 Naxi
0.04643241 Han_Sichuan
0.04772277 Yi
0.04772448 Yugur
0.04797378 Han_Hubei
0.05547853 Tujia
0.05689642 Tibetan_Yunnan
0.05704829 Tibetan_Yajiang
0.05863696 Han_Chongqing
0.06013805 Han_Fujian
0.06358011 Qiang_Daofu
0.06370796 Japanese
0.06447369 Tu
0.06622914 Miao
0.06675737 She
0.06918140 Bonan

I would regard Tujia/Han_Central as a Tibeto-Burman shifted Miao signal, which explains very well the low Yi for simulated Han_North, even when the top matches for Han_North are Yi and Naxi aside from Manchu_Liaoning. It's surprising how well "Han_Central" captures the Tibeto-Burman shift in Northern Han as a Tibeto-Burman + Middle Yangtzean proxy.

That's make more sense. My bad, you are very likely correct. That might explain why Han_North, Manchu Liaoning and Dungan/Hui needs the Korean element besides Han_Central in order to get great model fits as Korean is a proxy for Northeast Asian ancestry in this run.

Makes me wonder if ancient Sinitics such as Yellow_River_LBIA be model as a mix of Korean+Han_Central (since it already contain Tibeto-Burman affinities) in this Dodecad K12b (I know its not correct to modern pops to model something that is more ancient or ancestral than them)?

[Max_H]

Interesting that Manchu is modeled as more Han_Central and more Nivh than Han_North. Do we know what the regional origin of Han_Central and Han_North is? I also noticed none of the non-Han samples contained both Han_Central and Yizu.

Han_Central appears as Anhui/Zhejiang/Hubei mix, Han_North Dongbei or maybe Shandong/Hebei, especially considering the high Korean.

Target: Han_Shandong
Distance: 1.6894% / 0.01689353
51.4 Han_Zhejiang
48.6 Korean

Target: Han_Shandong
Distance: 2.0940% / 0.02093977
59.8 Korean
40.2 Han_Hubei

[alienation]

there seems to be somewhat different lvls or extents to which populations in East Asia have more inland components versus more ‘purely East Asian’ components. i think Japanese and Northern Hans tend to score slightly more inland compared to Koreans who are relatively “pure East Asian” like samples from the Amur River, for semantic purposes. the Yellow River MN, LN and Shimao and Miaozigou samples tend to be further “”pure EA”” relative to contemporary Hans, who could receive further inland pre-Tibetan like components from toward the foothills of the Himalayas rather than from a West Asian or western steppe source, it may even be a very slight ANE source from central asia or the Mongolian Plateau from ancient times

[alienation]

also I would urge posts including mine to be on the cautious side for using modern ethnic labels or names of ethncities to describe ancestries. it would be wrong to call a certain ancestry a name given to one or more ethnicities, when ethnicities are not to be equated with ancestries which could give false impressions with false lines of reasoning and associate ethncities to be identical based on their ancestral compositions as an inherent categorical problem

[alienation]

just because it appears as Kikuyu or Huaorani doesn’t mean it is Kikuyu... also we can’t jumble Ulchi and Nivkh in the same ethnic category just because they have similar or shared ancestries...

[okarinaofsteiner]

Han_Central appears as Anhui/Zhejiang/Hubei mix, Han_North Dongbei or maybe Shandong/Hebei, especially considering the high Korean.

Target: Han_Shandong
Distance: 1.6894% / 0.01689353
51.4 Han_Zhejiang
48.6 Korean

Target: Han_Shandong
Distance: 2.0940% / 0.02093977
59.8 Korean
40.2 Han_Hubei

Why Zhejiang or even Anhui? The Tujia model makes me think the reference population would be from further inland from the East China Sea.

there seems to be somewhat different lvls or extents to which populations in East Asia have more inland components versus more ‘purely East Asian’ components. i think Japanese and Northern Hans tend to score slightly more inland compared to Koreans who are relatively “pure East Asian” like samples from the Amur River, for semantic purposes. the Yellow River MN, LN and Shimao and Miaozigou samples tend to be further “”pure EA”” relative to contemporary Hans, who could receive further inland pre-Tibetan like components from toward the foothills of the Himalayas rather than from a West Asian or western steppe source, it may even be a very slight ANE source from central asia or the Mongolian Plateau from ancient times

Why would Japanese have more continental components than Koreans? Or are you simply referring to how Jomon-like ancestry could be modeled as something Central Asian-like, and appears as such in a global PCA?

[okarinaofsteiner]

23andMe v6.0 1/4 Aeta Filipino American results!

82.6% Filipino and Austronesian (surprisingly high), 17.2% Melanesian (surprisingly low until you realize OP's father is 1/4 Aeta lol)

My grandfather is where the aeta comes from. He has a parent that is Agta from Isabela. My grandfather was a pastor, but before then he ran away from home and met my grandmother. So I don’t know too much about his parents, and neither does my dad. We know about his father, who had many children with my great-grandma and some that weren’t hers, too �� my great-grandma, we don’t know much about.

My grandmother is Igorot from Baguio. We have a lot of family in Pagasinan as well. My dad came to the US at 1 years old. Supposedly we have a family tree and documents in my grandfathers briefcase, im not too sure but I’ll see if I can get ahold of it soon!

1/4 Aeta and the rest is basically 100% Austronesian. Your dad is the most Filipino Filipino to ever Filipino.

[East-Asia]

The autosomal DNA of Yumin (mtDNA C5d) can be equally used to model autosomal DNA of the ancient mtDNA CZ population and of the ancient mtDNA B4* population which is ca.44000 years old. CZ is a branch of M8'CZ, which as hg M should be related to those Papuans who are 100% hg M dominated. mtDNA CZ is younger than mtDNA B4*. mtDNA CZ settled inside mtDNA B4* population. Tianyuan is mtDNA B*, but he is younger than 44000-year-old B4* population included in Yumin, so Yumin can be the source of autosomal DNA of B4* population for Tianyuan. The mix of 44000-year-old B4* and CZ* is 56%Tianyuan+44% Papuan included in Native Americans in the article about Tianyuan.

So any B4 in China or Southeast Asia should be related to a small degree to 56%Tianyuan(+44% Papuan included in Native Americans in the article about Tianyuan. Because 40000-44000 year-old component migrated to Central Asia, Mongolian Plateau and to ANE, this fact creates he illusion that mtDNA B4-rich populations in China and Southeast Asia are related to ANE, Central Asia, Mongolian Plateau or even Native America.

The initial Salkhit people of Mongolia (basal female haplogroup mtDNA N*) is by far older than 40000-44000 year old mtDNA B4+mtDNA CZ population. 40000-44000 year old mtDNA B4+mtDNA CZ population which migrated to Siberia creates the illusion of northern gene flow in mtDNA B-rich Vietnam (yDNA O dominated) or mtDNA B4a-related Austronesian Ami population which is also dominated by yDNA O.

The TMRCA of various hg B clades in Vietnam are actually older than northern 40000-44000-year-old mtDNA B4*+ mtDNA CZ* population which got included in Native Americans.

[East-Asia]

Koreans who are relatively “pure East Asian” like samples from the Amur River

NE20 F AR33K AR33K 34,324–32,360CalBP B –
NE56 M AR19K AR19K 19,587–19,175CalBP G2 C2
NE34 M AR14.5K AR14K 14,932–14,176CalBP D4h3 C
NE-5 M AR14.1K AR14K 14,814–14,017CalBP [email protected] C
NE-1 F AR12K AR13-10K 12,735–12,486CalBP D4g –
NE-8 M AR11K AR13-10K 11,601–11,176CalBP D4o DE
NE57f M AR11K_deleted– 11,206–10,765CalBP – –
NE36f U AR10.6K_deleted– 11,065–10,513CalBP D4h4 –
NE-3 M AR10.6K AR13-10K 10,996–10,429CalBP M8 C
NE-4 M AR10.5K AR13-10K 10,740–10,302CalBP M8 C
NE30g M AR9.9K_2d.rel.AR10.5K_deleted- 10,167–9,676CalBP D4m C
NE45 M AR9.2K_o AR9.2K_o 9,425–9,029CalBP G1a1a P
NE44f M AR9.2K_deleted – 9,425–9,027CalBP D4c1 –
NE35 F AR8.9K ARpost9K 9,131–8,770CalBP R11 –
NE-16 F AR8.5K ARpost9K 8,723–8,421CalBP D4o –
NE49 F AR8.3K ARpost9K 8,425–8,204CalBP D4o –
NE39 M AR8.1K ARpost9K 8,340–8,029CalBP D4m2 C
NE58h M AR7.3K_LowCov – 7,425–7,168CalBP D4e5 –
NE-18f M AR7K_deleted– 7,245–6,894CalBP D4h3 –
NE19 M AR7K ARpost9K 7,167–6,854CalBP C
NE29 F AR6.87K ARpost9K 6,993–6,747CalBP D4c1 –
NE-22 M AR6.84K ARpost9K 6,986–6,676CalBP D4m C
NE-9 F AR6.33K ARpost9K 6,440–6,205CalBP C4a1a+195
NE-2 M AR6.32K ARpost9K 6,437–6,201CalBP D4m C
NE61h U AR3.4K_LowCov – 3,485–3,360CalBP D4b1a2a –

"The Z hg is relatively rare and originates from the Amur River region with an estimated coalescence date for the basal Z of 19,711.6 ± 8380.8 BCE (Behar et al., 2012); several of the eastern rivers of the Mongol Steppe, including the Kherlen River, are eventual tributaries of the Amur River. Samples AT323 and AT326 were assigned to M31a2 by Haplogrep2, but according to mtDNAmanager the M3 haplotype is just as likely (Lee et al., 2008). Although the SNPs at np16093, 16126, and 16223 compared with the revised Cambridge Reference Sequence (rCRS; Andrews et al., 1999) place this individual's mtDNA haplotype within the M hg, this combination of SNPs has not been found within an M haplotype in previously published studies. However, this combination was identified in a Bronze Age prone burial from central Mongolia (Rogers, 2016), suggesting recent ancestry between the two samples. Both M31 and M3 hgs are currently prevalent in the region of northern India and Nepal and may have arrived via migration from the southeast along the Central Asian Mountain Corridor (Endicott et al., 2007; Frachetti, 2015; Metspalu et al., 2004)."

mtDNA M31a2 in Slab Grave Mongol-Tungus is related to Great Andamanese Onge, a "brother" of Japanese D-M64. Some branches of M3 are present in China and formerly mtDNA M3 and mtDNA M31 possibly interacted.

So mtDNA M3 was not initially related to y chromosome hg N. Some y chromosome hg N could start to coexist with mtDNA M3 at a later period. However, initially those hg N who began to interact with mtDNA M3a, belonged to a different mtDNA haplogroup than mtDNA M3a. A different mtDNA haplogroup first interacted with mtDNA M75 population, and later one of its branches started to interact with M3a, while another of its branches started to interact with mtDNA B5b4 which split from mtDNA B5b1 of ancient Shandong Houli and Tonglin settlements in "Maternal genetic structure in ancient Shandong between 9500 and 1800 years ago".