black man wrote:There is one more Vietnamese sample the numbers of which can be read in the paper by Trejaut et al. from 2014...
In short, when M7+ and M95+ are characteristic to MKs in general, these mutations can be considered representative for "originally" MK lineages in the Kinh people until further notice.
Since IMS-JST002611+ is frequent in Karafet's Vietnamese sample only, I think that this mutation reflects the impact of Hoklo lineages first of all, Fujian being the epicentre of IMS-JST002611+ distribution within East Asia according to Trejaut et al.
C-M217+ should be studied in depth more often and more holistically. I have no idea whether it is local aboriginal, from TKs, from MKs or from standard Chinese people.
He et al. 2012
Karafet et al. 2010
Trejaut et al. 2014
As for the Vietnamese, differences in the frequencies of various Y-DNA haplogroups among different samples are not the only problem. MtDNA frequencies also have been significantly different (cf.
the high frequency of M7 in the "Kinh" sample of the Chinese He & Peng team (21/139 = 15.1%) vs.
the low frequency of M7 in the "Vietnamese" sample of the Korean Jin & Kim team (2/42 = 4.8%)). I think we simply need new (and, hopefully, larger) samples from a greater number of locations with greater geographical coverage (rather than one sample from every suburb of Hanoi, for example).
Nearly all C-M217 in East Asia south of Xinjiang, Mongolia, and Manchuria (except, as far as I recall, a few cases in Japan and Tibet) belongs to a subclade different from that to which the majority of "Altaic" and other North Asian and American C-M217 belongs. C-M217 in each of these two large regions (roughly North Asia + America vs.
East Asia + Southeast Asia) reflects a separate haplogroup, in other words, and they just happen to have been conflated under the "C-M217" designation because of poor phylogenetic resolution. I do think that researchers should quit lumping all these subclades together; judging from Figure 1a of "Comparison of Y-chromosomal lineage dating using either evolutionary or genealogical Y-STR mutation rates," conflating East/Southeast Asian C-M217 with North Asian/American C-M217 may be an even more egregious disregard for precision than conflating American Q1a2a1a1-M3 with African R1b1c-V88.
Speaking of insufficient geographical coverage and samples that are too small, I have noticed that Qi et al.
(2013) finally have sampled a population of Tibetans from Maqên County, the seat of Mgolog Tibetan Autonomous Prefecture in the southeast of Qinghai Province (erstwhile Amdo region of Tibet), but they have sampled 125 females and only 5 males
. Why do they need to have such an extreme gender bias in sampling? In any case, as it turns out, three of the five sampled males belong to C2-M217(xM48, P53.1), and two belong to O3a-M324(xM134), neither of which has been particularly frequent in samples of Tibetans from other areas.
The sample of Tibetans from Qinghai in Wen et al.
(2004), which is from Yulshul according to a subsequently published paper, is quite different. Yulshul is the western neighbor of Mgolog within Qinghai, but it is associated with the Kham region instead of the Amdo region:
13/92 = 14.1% C-M130
19/92 = 20.7% DE-YAP(xD1a-M15)
2/92 = 2.2% D1a-M15
13/92 = 14.1% F-M89(xK-M9)
20/92 = 21.7% K-M9(xO1a-M119, O2a1-M95, O3-M122, P-M45)
5/92 = 5.4% O3-M122(xM134)
13/92 = 14.1% O3a2c1-M134
1/92 = 1.1% O1a-M119
6/92 = 6.5% P-M45
Whereas the Kham Tibetans from Garzê Tibetan Autonomous Prefecture according to Wang et al.
(2014) have the following composition:
1/93 = 1.1% C-M130(xC1a1-M105, C1b-M356, C1c-M38, C1d-M347, C2-M217)
5/93 = 5.4% C2-M217(xC2a-M93, C2b1-P39, C2b2-M48, C2c-P53, C2e1a-M407)
7/93 = 7.5% D1a1-N1
1/93 = 1.1% D1c*-P99(xD1c1-P47)
19/93 = 20.4% D1c1-P47
1/93 = 1.1% F-M89(xG-M201, I-M258, J-M304, K-M9)
1/93 = 1.1% J-M304
1/93 = 1.1% N1c1a-M178
14/93 = 15.1% O1a1-P203
2/93 = 2.2% O2*-M268
2/93 = 2.2% O2a*-PK4(xO2a1-M95)
2/93 = 2.2% O2a1*-M95(xO2a1a-M88)
1/93 = 1.1% O2a1a-M88
1/93 = 1.1% O2b-M176
2/93 = 2.2% O3a-M324(xO3a1-L127/KL1, O3a2-P201)
7/93 = 7.5% O3a1c-JST002611
1/93 = 1.1% O3a1-L127/KL1/KL2(xO3a1a-M121, O3a1c-JST002611)
3/93 = 3.2% O3a2b-M7
6/93 = 6.5% O3a2c1*-M134(xM117)
14/93 = 15.1% O3a2c1a-M117
There seems to be a great deal of local diversity (because of founder effects, assimilation of other populations, or genetic drift) in Tibet. The Khampas in Garzê have a lot of O1a1-P203, the Khampas in Yulshul have a lot of C-M130, F-M89(xK-M9), and K-M9(xO1a-M119, O2a1-M95, O3-M122, P-M45) (probably some sort of N or O2(xO2a1-M95)), and the Amdowas in Mgolog may
have a lot of C2-M217 and O3a-M324(xM134), but it is impossible to be sure without further sampling.