PLoS One has a paper out on Korean (South) population genetics and phylogeography, Gene Flow between the Korean Peninsula and Its Neighboring Countries:

SNP markers provide the primary data for population structure analysis. In this study, we employed whole-genome autosomal SNPs as a marker set (54,836 SNP markers) and tested their possible effects on genetic ancestry using 320 subjects covering 24 regional groups including Northern ( = 16) and Southern ( = 3) Asians, Amerindians ( = 1), and four HapMap populations (YRI, CEU, JPT, and CHB). Additionally, we evaluated the effectiveness and robustness of 50K autosomal SNPs with various clustering methods, along with their dependencies on recombination hotspots (RH), linkage disequilibrium (LD), missing calls and regional specific markers. The RH- and LD-free multi-dimensional scaling (MDS) method showed a broad picture of human migration from Africa to North-East Asia on our genome map, supporting results from previous haploid DNA studies. Of the Asian groups, the East Asian group showed greater differentiation than the Northern and Southern Asian groups with respect to Fst statistics. By extension, the analysis of monomorphic markers implied that nine out of ten historical regions in South Korea, and Tokyo in Japan, showed signs of genetic drift caused by the later settlement of East Asia (South Korea, Japan and China), while Gyeongju in South East Korea showed signs of the earliest settlement in East Asia. In the genome map, the gene flow to the Korean Peninsula from its neighboring countries indicated that some genetic signals from Northern populations such as the Siberians and Mongolians still remain in the South East and West regions, while few signals remain from the early Southern lineages.

I can’t comment too much on the inferences they make from the results because I’m not familiar with the geography of South Korea, or particular historical details. But more generally the genetics of Korea are of particular interest for social reasons:

South Korea is one of the most ethnically homogeneous societies in the world with more than 99 per cent of inhabitants having Korean ethnicity…The Koreans call their ethnic homogeneousity of their society using the word, ?????? (Dan-il minjok gook ga, literally means the single race society.)

Korean racialism has recently gotten the spotlight in works such as The Cleanest Race: How North Koreans See Themselves and Why It Matters, articles in The New York Times about South Korean prejudice against dark-skinned people, and the rise of mixed-origin Koreans nationals due to the large number of Vietnamese brides in rural areas. Here’s an interesting comment on South Korean race consciousness, Western Mixed-Race Men Can Join Military:

Western mixed-race men can join the military beginning next year.

Currently, Asian mixed-race men, dubbed “Kosians,” are subject to the country’s conscription system, but “Amerasians” or “Eurasians” are exempted from the mandatory service.

The parliamentary approval of a bill proposed by Rep. Yoo Seung-min of the governing Grand National Party has paved the way for them to join the military.

Western mixed-race men, who have distinctive skin colors, had been exempted because they could have experienced difficulty mixing with Korean colleagues in barracks, the defense ministry had said previously.

The article was published in January of 2010. And that’s not the weirdest idea to come out of the Korean peninsula. With all that in mind, the distinctiveness, or lack thereof, of the Korean nation as adduced from scientific genetics is of particular curiosity, as it is a clear example of the intersection of science and culture. First, here’s the figure which shows where in Asia & South Korea they got their samples from:

And here’s a detailed breakdown of samples:

One point to note is that there seem to be some mixed-nationality individuals in the sample; Korean-Japanese, and Korean-Vietnamese. Here’s a MDS plot showing the relationship between the various East Asian groups:

And Structure (remember K = putative ancestral populations which contribute quanta to the genome of individuals):

I think it is important to note that their Chinese samples were all north Chinese; Beijing and Manchurian. Fujianese and Cantonese would span the Vietnamese and Chinese cluster. The outliers are probably due to the moderately cosmopolitan nature of the Beijing HapMap sample. The Han Chinese are less diverse than Europeans as a whole, but not inordinately so (using pairwise Fst’s a measure). There is an asymmetry when talking about China and any other East Asian nation because it is feasible that Han groups from various regions of China are more genetically similar to non-Han groups which are geographical neighbors. This is what L. L. Cavalli-Sforza found in History and Geography of Human Genes. The northern Chinese clustered with northern Asians, while the southern Chinese clustered with Southeast Asian groups. There have been conflicting results since that initial finding, but I think that points to the sensitivity of some of the inferences to the geographical and linguistic biases of sampling (different dialect groups in Guangdong may be very genetically distinct).

With all that said it’s pretty clear from the above figure that the Japanese and Korean samples are close enough that you need to zoom in on them specifically. So here you go:

KB_Japanese = Kobe Japanese. In the paper itself they’re testing a few historical hypotheses. So I’ll leave it to them for the interpretation:

The gene flow events of the three selected models for SW, MW and SE Korea can be assessed using the genome map. The populations in Model I (SW Korea) are closer to Mongolians than are the other two regions in the genome map (Fig. 2B). Historically, some of the loyal families and their subjects in the Goguryeo Empire moved to this region and formed the BaekJae Empire in BC18-22. This region also showed connections with populations in Tokyo (JPT), as illustrated in Fig. 4. Certain outliers in Model II (SE Korea) display some similarity to the people of Kobe, a port city near Osaka, indicating that there may have been links between the two regions. In addition, considering that the SE Korea region has some connections with Siberian lineages, with respect to grave patterns and culture, it is possible that the outliers in the GU and Kobe (KB) populations could be of Siberian lineage. On the other hand, the GR and US populations showed average signals in the Korean Peninsula. Historically, the Kaya Empire, with its southern lineages, was formed in the GR region and then the Shilla and Kaya Empires became united around AD532. Very recently, the US region became one of the rapidly developing regions, and people from other provinces moved to this region. This might explain why it shows an average signal in South Korea. Model III (MW Korea): the Middle West area formed a melting pot in the Korean Peninsula because populations moving from South to North, North to South, and from Eastern China, including the SanDung peninsula, to the Middle West in Korea all came together in this region. In the genome map, the signals for MW Korea are also close to those for Peking (CHB) in China. The overall result for the Korea-Japan-China genome map indicates that some signals for Mongolia and Siberia remain in SW Korea and SE Korea, respectively, while MW Korea displays an average signal for South Korea.

The connections between coastal southern Korea and the western islands of Japan are well known. It seems like that the Yayoi people, who probably contributed the preponderance of the ancestry of modern Japanese, arrived in Kyushu approximate ~2,500 years ago. And were originally a group within the Korean peninsula. Over the past 2,000 years Korea has gone through a process of ethnic-linguistic homogenization during the ethnogenesis of the modern Korean nation, but it seems possible that the original group(s) which gave rise to the Yayoi existed in southern Korea to facilitate contact between the islands and the peninsula into the historical era.

Citation: Jung J0, Kang H, Cho YS, Oh JH, & Ryu MH (2010). Gene Flow between the Korean Peninsula and Its Neighboring Countries PLoS ONE : 10.1371/journal.pone.0011855