search this blog

Wednesday, February 1, 2017

Strong genetic continuity in the Amur Basin since the Neolithic

Open access at Science Advances:

Abstract: Ancient genomes have revolutionized our understanding of Holocene prehistory and, particularly, the Neolithic transition in western Eurasia. In contrast, East Asia has so far received little attention, despite representing a core region at which the Neolithic transition took place independently ~3 millennia after its onset in the Near East. We report genome-wide data from two hunter-gatherers from Devil’s Gate, an early Neolithic cave site (dated to ~7.7 thousand years ago) located in East Asia, on the border between Russia and Korea. Both of these individuals are genetically most similar to geographically close modern populations from the Amur Basin, all speaking Tungusic languages, and, in particular, to the Ulchi. The similarity to nearby modern populations and the low levels of additional genetic material in the Ulchi imply a high level of genetic continuity in this region during the Holocene, a pattern that markedly contrasts with that reported for Europe.

Siska et al., Genome-wide data from two early Neolithic East Asian individuals dating to 7700 years ago, Science Advances, 01 Feb 2017: Vol. 3, no. 2, e1601877, DOI: 10.1126/sciadv.1601877


Rob said...

They're poorer quality I gather (lack of solid resolution even for mtdna).
Will you be accessing them ?

Davidski said...

Don't think so, unless someone else requests them, on my behalf as well, and sends them over.

Matt said...

From the other thread. Interesting paper. I think it would've been better to say "Europe and Japan" than Europe though (since adna indicates replacement of Jomon people).

Kind of looks like it walked things back in terms of the ambition of their analysis from the Wang et al abstract from Paleobarn ( where these samples were mentioned before.

One thing I would've liked to see which they didn't seem to include would've been the outgoup f3 between the two Devil's Gate individuals and also their conditional nucleotide diversity. Possibly this was not technically possible due to low quality genomes.

Where we have ancient dna from the Lapita, we find that there's a high degree of f3 sharing between them -, which is beyond the range for modern Taiwanese Aborigines.

Likewise with the early Boncuklu in the Neolithic people, we found that their f3 sharing was much higher (fig S3) and conditional nucleotide diversity was much lower than later Neolithic people (Fig2), and than moderns -

So in those instances, we could infer some kind of process whereby smaller local populations pooled to accrue more genetic diversity, or were replaced by more diverse local variants (even though, e.g. the later Barcin Neolithic don't look like mixes of Boncuklu and any other population in mixture f3 stats).

Also, I was a little surprised that they reported the following D-stats in their Extended Table 6:

D(MA1,DevilsGate1,Papuan, Khomani) = -0.0270, Z: -2.300
D(MA1,DevilsGate1,Onge, Khomani) = -0.0278, Z: -2.344

Would've expected them to be higher. What is the comparable on D(MA1,Ulchi,Bougainville/Onge, Khomani)?

Extended Table 7 has some odd results as they say. These are the stats of the form "D(African outgroup, X; Ulchi, Devil’s Gate)" to "test whether Devil’s Gate formed a clade with the Ulchi", "Because admixture f3 can be affected by demographic events such as bottlenecks". For example D(Khomani,MA1,Devilsgate1,Ulchi) = 0.0671, Z=5.009 / D(Khomani,Han,Devilsgate1,Ulchi) = 0.0347, Z=5.352. Though as this table seems to even show D(Khomani,Luo,Devilsgate1,Ulchi) = 0.0122, Z=-2.360, I'm not sure about the significance of these.

rozenfag said...

Yes, coverage in unfortunately low. But, there will be another study soon, with many more samples from approximately the same region: Reconstructing population history in East Asia, Wang et al.

Chad Rohlfsen said...

@ Matt,

Here, f3s are consistent with my qpAdm models of Barcin and Mentese as mostly Boncuklu with some Natufian/Levant_EN mixture.

result: Boncuklu Natufian Barcin_EN -0.004252 0.003197 -1.330 28893
result: Boncuklu Natufian Mentese_EN -0.002439 0.007134 -0.342 6448
result: Boncuklu Levant_EN Barcin_EN -0.005525 0.002117 -2.610 48620
result: Boncuklu Levant_EN Mentese_EN -0.002529 0.005208 -0.486 10490

Higher SNP counts should easily make these all Z>3.

Rob said...

matt & Chad

Yes I was going to say Chad's post on a few threads ago seem to be at odds with your quote from the study itself....

Chad Rohlfsen said...

They should've cleaned up deamination and ran transversions.

Veronika Siska said...


Results on transversions only are in the supplementaries.

Alberto said...

Nice to see some Asian ancient DNA finally coming. Though still from the Russian side fo the border. I wonder when will the Chinese start releasing more of their ancient DNA (the Tarim Basin mummies is another strange case of results not being published, even when tested for Y and mt DNA long ago).

In this one I would have liked to see the non-Tungusic Nivkh people included. And also a more explicit test for ANE in the samples, like D(Mbuti,DevilsGate1)(MA1, Loschbour).

It's also a bit strange the comparison with the European case. This area didn't pick up agriculture, so why would farmer's genes replace hunter-gatherer's ones when people continued to be hunter-gatherers?

Not that the differences are going to be big in East Asia, since whoever were the first farmers were probably genetically very similar to the other populations, so it won't be much of a genetic replacement in any case.

Matt said...

@ Rozenfag, ah, so that study is still on.

High coverage samples will also be able to calculate fst based differentiation, and that will be useful as well to see if the Devil's Gate individuals have similar fst behaviour as present day populations - e.g. Ulchi-Yoruba 0.187, Ulchi-Han 0.028, or if they have higher fsts, as WHG do compared to present day Europeans (WHG-Yoruba 0.201, Ukrainian-Yoruba 0.150, WHG-Ukrainian 0.057). Fst can calculated with only a few samples like with (2) CHG and (3) EHG in Lazaridis 2016, if the quality is OK.

@ Alberto: It's also a bit strange the comparison with the European case. This area didn't pick up agriculture, so why would farmer's genes replace hunter-gatherer's ones when people continued to be hunter-gatherers?

In a sense yes, I agree. Japan and Tibet, where populations moved to an agricultural base would suggest a high degree of replacement (Jomon adna seems to make it fairly certain for Japan, while the model for Tibet is currently dependent on "ghost" admixture and will benefit from adna). So I find it hard to understand the emphasis on East Asian->West Eurasian distinction of Neolithic process as much as farming vs non-farming Holocene transition.

It is defensible that there is still a Neolithic transition going on, if you define things the Russian way that pottery use counts as Neolithic (New Stone) Age. Though in that sense a similar transition may then be EHG->Steppe Eneolithic, rather than EHG->Yamnaya, and here the Ulchi-Devil's Gate comparison with ADMIXTURE suggests something maybe similar to the transition between EHG->Steppe Eneolithic, with about 30% extra admixture from their estimate. Or more speculatively SHG->Pitted Ware Culture.

(Theoretically: Farming creates bigger population expansions - more chance of local populations being absorbed - and/or requires different skills compared to adding pottery to a HG subsistence base - more chance of locals being replaced without admixture).

Though I might be inaccurate in equating the Ulchi to being potentially effectively equivalent to a "survivor" version of a PWC or Steppe Eneolithic as hunter gatherers who adopted pottery, and a minority Neolithic expansion ancestry, but not really other Neolithic cultural developments. By comparison, in subsistence even the Saami are basically herders, of an unusual sort, and not HG (herding replaced advanced / Neolithic HG in West Eurasia in the environments beyond suitability for farming - perhaps not so much in the North East, and thus the Ulchi).

Alberto: Not that the differences are going to be big in East Asia, since whoever were the first farmers were probably genetically very similar to the other populations, so it won't be much of a genetic replacement in any case.

In deep, neutral ancestry I agree - there doesn't seem to be likely to have been as much time for different clades to have differentiated in East Asia compared to the deep splits between Basal Eurasian / Main West Eurasian and within each of those groups. (Now that the evidence does not seem to support an early wave in East Asia basal to all other Eurasians). In selected alleles it may be more uncertain.

This study is interesting in suggesting similar pigmentation etc variants in the Devil's Gate individuals to present day people, similar EDAR, similar ALDH, but our understanding of some of these traits in East Asian groups may be incomplete.