Genetic history of East Asians

Over the past decade, Chinese archaeologists have published several reviews regarding the results of excavations in Xinjiang. They imply the genetic composition of Xiongnu's supreme ruling class. Particularly interesting are the tombs in the cemetery at Heigouliang, Xinjiang (the Black Gouliang cemetery, also known as the summer palace of the Xiongnu king), east of the Barkol basin, near the city of Hami. By typing results of DNA samples during the excavation of one of the tombs, it was determined that of the 12 men: 6 Q1a* (not Q1a1-M120, not Q1a1b-M25, not Q1a2-M3), 4 Q1b-M378, 2 Q* (not Q1a, not Q1b: unable to determine subclades):[4]

In a paper (Lihongjie 2012), the author analyzed the Y-DNAs of the ancient male samples from the 2nd or 1st century BCE cemetery at Heigouliang in Xinjiang – which is also believed to be the site of a summer palace for Xiongnu kings – which is east of the Barkol basin and near the city of Hami. The Y-DNA of 12 men excavated from the site belonged to Q-MEH2 (Q1a) or Q-M378 (Q1b). The Q-M378 men among them were regarded as hosts of the tombs; half of the Q-MEH2 men appeared to be hosts and the other half as sacrificial victims.

Genetic studies published in 2006 and 2015 revealed that the mitochondrial haplogroups of Xianbei remains were of East Asian origin. According to Zhou (2006) the haplogroup frequencies of the Tuoba Xianbei were 43.75% haplogroup D, 31.25% haplogroup C, 12.5% haplogroup B, 6.25% haplogroup A and 6.25% "other."[6]

Zhou (2014) obtained mitochondrial DNA analysis from 17 Tuoba Xianbei, which indicated that these specimens were, similarly, completely East Asian in their maternal origins, belonging to haplogroups D, C, B, A and haplogroup G.[7]

The research also found a relation between Xianbei individuals with modern Oroqen, Ewenki and Outer Mongolian people. Especially Tungusic Oroqen show close relation to Xianbei.[8]

Y-chromosome haplogroup O2-M122 is a common DNA marker in Han Chinese, as it appeared in China in prehistoric times. It is found in more than 50% of Chinese males, and ranging up to over 80% in certain regional subgroups of the Han ethnicity.[20] Other Y-DNA haplogroups that have been found with notable frequency in samples of Han Chinese include O-P203 (15/165 = 9.1%, 47/361 = 13.0%), C-M217 (10/168 = 6.0%, 27/361 = 7.5%, 187/1730 = 10.8%, 20/166 = 12.0%), N-M231 (6/166 = 3.6%, 18/361 = 5.0%, 117/1729 = 6.8%, 17/165 = 10.3%), O-M268(xM95, M176) (54/1147 = 4.7%,[21] 8/168 = 4.8%, 23/361 = 6.4%, 12/166 = 7.2%), and Q-M242 (2/168 = 1.2%, 49/1729 = 2.8%, 12/361 = 3.3%, 48/1147 = 4.2%[21]). However, the mitochondrial DNA (mtDNA) of Han Chinese increases in diversity as one looks from northern to southern China, which suggests that male migrants from northern China married with women from local peoples after arriving in modern-day Guangdong, Fujian, and other regions of southern China.[12][22] Despite this, tests comparing the genetic profiles of northern Han, southern Han and southern natives determined that haplogroups O1b-M110, O2a1-M88 and O3d-M7, which are prevalent in southern natives, were only observed in some southern Han (4% on average), but not in northern Han. Therefore, this proves that the male contribution of southern natives in southern Han is limited, assuming that the frequency distribution of Y lineages in southern natives represents that before the expansion of Han culture that started two-thousand years ago.[12][23] In contrast, there are consistent strong genetic similarities in the Y chromosome haplogroup distribution between the southern and northern Chinese population, and the result of principal component analysis indicates almost all Han populations form a tight cluster in their Y chromosome. However, other research has also shown that the paternal lineages Y-DNA O-M119,[24] O-P201,[25] O-P203[25] and O-M95[26] are found in both southern Han Chinese and South Chinese minorities, but more commonly in the latter. In fact, these paternal markers are in turn less frequent in northern Han Chinese.[27][28]

The mitochondrial-DNA haplogroups of the Han Chinese can be classified into the northern East Asian-dominating haplogroups, including A, C, D, G, M8, M9, and Z, and the southern East Asian-dominating haplogroups, including B, F, M7, N*, and R.[12] These haplogroups account for 52.7% and 33.85% of those in the Northern Han, respectively. Among these haplogroups, D, B, F, and A were predominant in the Northern Han, with frequencies of 25.77%, 11.54%, 11.54%, and 8.08%, respectively. However, in the Southern Han, the northern and southern East Asian-dominating haplogroups accounted for 35.62% and 51.91%, respectively. The frequencies of haplogroups D, B, F, and A reached 15.68%, 20.85%, 16.29%, and 5.63%, respectively.[10][29][30][31][32]

Jōmon people is the generic name of people who lived in the Japanese archipelago during the Jōmon period. Today most Japanese historians believe that the Jomon people were not one homogeneous people but were at least two or three distinct groups.[33]

The origin of the Jōmon people and their ancestors is disputed. Several theories suggested Southeast Asia or Northeast Asia as possible place of origin. Another theory supported an origin in East Asia itself. Newest genetic studies (since 2017) conclude that the Jōmon are the last descendants of a unique group of ancient people. The study suggests an origin in modern Central Asia.[34][35]

Another study by Hideaki Kanzawa showed that the Jōmon people of Hokkaido and Honshu have a genome that is commonly found in Arctic populations but is rare in Yamato people. The study further suggests that the Jōmon drank alcohol and had wet earwax that is more common in western Eurasians.[36]

According to Mitsuru Sakitani the Jōmon people are an admixture of two distinct ethnic groups: A more ancient group (carriers of Y chromosome D-M55) that were present since more than 30,000 years in Japan and a more recent group (carriers of Y chromosome C1a) that migrated to Japan about 13,000 years ago.[37]

Haplogroup D1 arrived from Central Asia to northern Kyushu via the Altai Mountains and the Korean Peninsula more than 40,000 years before present, and Haplogroup D-M55 (D1b) was born in Japanese archipelago. D-M55 is distinct from other D-branches since more than 53,000 years and has five unique mutations not found the others.[37][38] C1a1's ancestral type reached Japan over the Korean Peninsula via the Altai Mountains from Western Asia. Although its age of arrival is unknown, the spread of the existing subgroup is about 12,000 years ago, which is almost consistent with the start of the Jōmon period. The next relative C1a2 was common in ancient European and West Asian samples and is still found in small numbers of modern Europeans, Armenians, Algerians (Kabyle Berbers), and Nepalis.[37][39]

Recent Y chromosome haplotype testing has led to the hypothesis that male haplogroups D-M55 and C1a1, which have been found in different percentages of samples of modern Japanese, Ryukyuan, and Ainu populations, may reflect patrilineal descent from members of pre-Jōmon and Jōmon period of the Japanese Archipelago.[35]

Haplogroup D-M55 is found in about 36% and haplogroup C1a1 in about 5%[35][40] of modern Japanese people. C1a1 has its highest amount in Tokushima Prefecture at about 10%, followed by Okinawa Prefecture, Aomori Prefecture, and Tokyo at about 7-8%.[35][41] In addition, it is assumed that the haplogroup C2 existed in a small amount of Jōmon people.[42]

Recently it was confirmed that the Japanese branch of haplogroup D-M55 is distinct and isolated from other D-branches since more than 53,000 years. The split between D1a (which is common in Tibet and has a medium distribution in Central Asia) happened likely in Central Asia, while some others suggest an instant split during the origin of haplogroup D itself, as the Japanese branch has five unique mutations not found in any other D-branch.[43][38]

A recent DNA study in 2019 suggests that haplogroup D-M55 was carried by about 70% and haplogroup C (C1a1) by about 30% of the ancient Jōmon people. A specific Japanese-Jōmon clade is only found in ancient Jōmon and some modern Japanese. No other population was found to carry this specific clade, which support the distinct position of the Jōmon population.[44]

MtDNA Haplogroup of Jōmon people is characterized by the presence of haplogroups M7a and N9b.

M7a is estimated to share a most recent common ancestor with M7b'c, a clade whose members are found mainly in Japan (including Jōmon people), other parts of East Asia, and Southeast Asia, 33,500 (95% CI 26,300 <-> 42,000) years before present.[45] All extant members of haplogroup M7a are estimated to share a most recent common ancestor 20,500 (95% CI 14,700 <-> 27,800) years before present.[45] Haplogroup M7a now has its highest frequency in Okinawa.

Haplogroup N9b is estimated to share a most recent common ancestor with N9a and Y, two clades that are widespread in eastern Asia, 37,700 (95% CI 29,600 <-> 47,300) years before present.[45] All extant members of haplogroup N9b are estimated to share a most recent common ancestor 21,100 (95% CI 16,700 <-> 26,200) years before present.[45] Haplogroup N9b now has its highest frequency among Tungusic peoples in southeastern Siberia (especially Udeges), but it has been found to be very common in skeletal remains of Jōmon people of northern Japan (Tōhoku and Hokkaidō).

In addition, haplogroups D4, D5, M7b, M8, M9a, M10, G, A, B, and F have been found in Jōmon people as well.[46][47][48] These latter haplogroups are all distributed widely among populations of East Asia (including modern Japanese, Ryukyuans, and Ainus) and Southeast Asia, but some of their subclades are distributed almost exclusively in Japan.[47]

Analysis of the mitochondrial DNA ("mtDNA") of Jōmon skeletons from Hokkaido, Okinawa Island and Tōhoku region indicates that haplogroups N9b and M7a may reflect maternal Jōmon contribution to the modern Japanese mtDNA pool.[49][50][51][52][53][54] In another study of ancient DNA published by the same authors in 2011, both the control and coding regions of mitochondrial DNA (mtDNA) recovered from Jōmon skeletons excavated from the northernmost island of Japan, Hokkaido, were analyzed in detail, and 54 mtDNA samples were confidently assigned to relevant haplogroups. Haplogroups N9b, D4h2, G1b, and M7a were observed in these individuals.[55] According to 2013 study, there was mtDNA sub-haplogroups inter-regional heterogeneity within the Jōmon people, specifically between studied Kantō, Hokkaido and Tōhoku Jōmon.[49] According to 2011 study all major East Asian mtDNA lineages expanded before 10,000 YBP, except for two Japanese lineages D4b2b1 and M7a1a which population expanded around 7,000 YBP unequivocally during the Jōmon Period (14–2.3 kya), thousands of years before intensive agriculture which imply that the growth of population and depletion of food resources was the reason for population expansion and not agriculture.[56]

A study about maternal DNA of Jōmon individuals resulted in similarities between Jomon people and ancient Siberian people. Interestingly, the study also suggests a relation of some Jomon people to at least some Native American groups.[57]

A 2017 study on ancient Jomon aDNA from Sanganji shell mound in Tōhoku region estimates that the modern mainland Japanese population probably inherited less than ~20% of Jōmon peoples' genomes.[50]

The first full genomic DNA analysis of Jōmon individuals by Hideaki Kanzawa-Kiriyama of the department of genetics in the University for Advanced Studies (SOKENDAI) showed that the Jōmon people are not closely related to any modern ethnic group. His analysis groups the Jōmon people into a unique genetic cluster far away from any modern ethnic groups. Hideaki says that some Jōmon DNA is found in modern ethnic groups, such as Japanese people, Udege people, Nivkh people, Ainu people and Ryukyuan people. From all ethnic groups, the Ainu and Ryukyuans show the closest relation to ancient Jōmon people. Compared with populations worldwide, the Jomon are relatively close to modern Ryukyuans, Ainu and Yamato Japanese.[58]

A DNA-based reconstruction of a 3,800-year old Jomon woman of Rebun Island in Hokkaido showed that she had slightly darker skin than modern Japanese people but a lighter eye colour. She also had freckles and thin brown hair.[59]

A full genome analysis (Kanzawa-Kiriyama et al. 2019),[60] using high-confidence SNPs and functional SNP assessments to assign possible phenotypic characteristics as well as Y-chromosome polymorphisms, analysed a male and a female Jomon sample. The study results suggest that the Jōmon are their own distinct population and not closely related to other populations. The Funadomari Jōmon are not related to Australo-Melanesians (including Andamanese) or Africans. The Jōmon are closer to Eurasian populations and form a cluster near the “Basal East Asians”.

Modern Japanese share about 9% to 13% of their genome with the Jōmon. Jōmon specific genome is also found in minor percentage in populations of Northeast Asia and Southeast Asia, suggesting gene-flow from Jōmon related groups. Additionally, the Jōmon share specific gene alleles with populations in the Arctic regions of Eurasia and northern America.

Tests using phylogenetic relationship suggests that the Funadomari Jōmon have about 86% East Asian related ancestry and about 14% West Asian/European related ancestry. According to the scientists, more data is needed to explain these results.

The Yayoi people were migrants to the Japanese archipelago from Asia (Korea or China) during the Yayoi period (1000 BCE–300 CE) and Kofun period (250–538 CE). They are seen as direct ancestors of the modern Yamato people, the majority of Japanese and of the Ryukyuan people. It is estimated that modern Japanese share in average about 90% of their genome with the Yayoi.[61]

It is estimated that Yayoi people mainly belonged to Haplogroup O-M176 (O1b2) (found in ~32% of present-day Japanese males), Haplogroup O-M122 (O2, formerly O3, today ~20%), Haplogroup O-K18 (O1b1, today ~1%), and Haplogroup O-M119 (O1a, today ~1%), which are typical for East and Southeast Asians.[62][63] Mitsuru Sakitani suggests that haplogroup O1b2, which is common in today's Japanese, Koreans, and some Manchu, and O1a are one of the carriers of Yangtze civilization. As the Yangtze civilization declined several tribes crossed westward and northerly, to the Shandong peninsula, the Korean Peninsula and the Japanese archipelago.[64] One study calls haplogroup O1b1 as a major Austroasiatic paternal lineage and the haplogroup O1b2 (of Koreans and Japanese) as a "para-Austroasiatic" paternal lineage.[65]

A recent study confirms that modern Japanese are predominantly descendants of the Yayoi. The mitochondrial chromosomes of modern Japanese are nearly identical with the Yayoi and differ significantly from the Jomon population (see below).[66]

The main paternal haplogroups of modern Yamato Japanese are Haplogroup D-M55 (today ~33%, with the frequency in various samples ranging from 18/70 = 25.7% in a sample from Tokushima Prefecture[67] to 24/59 = 40.7% in a sample of Japanese male volunteers[68] and 11/27 = 40.7% in a sample from Aomori Prefecture[69]), Haplogroup O-M176 (O1b2) (today ~32%, range 37/142 = 26.1% in a sample of Japanese[70] to 35/97 = 36.1% in a sample from Western Japan[71]), Haplogroup O-M122 (O2, formerly O3) (today ~20%, ranging from 4/59 = 6.8% in a sample of Japanese volunteers[68] and 11/102 = 10.8% in a sample of adults from Fukuoka[72] to 38/157 = 24.2% in a sample of Japanese[73] and 14/53 = 26.4% in a sample from Kyushu[67]), Haplogroup C-M217 (C2, today ~6%, ranging from 0/26 = 0.0% in a sample from Aomori,[67] 1/61 = 1.6% in a sample from Shizuoka,[67] 1/47 = 2.1% in a sample of Japanese,[74] and 3/137 = 2.2% in a sample from the Kantō region[71] to 15/206 = 7.3% in a sample from Sapporo,[72] 18/241 = 7.5% in a sample from Osaka,[72] 4/53 = 7.5% in a sample from Kyushu,[67] and 8/102 = 7.8% in a sample from Fukuoka[72]), and Haplogroup C-M8 (C1a1, today ~5%, ranging from 0/53 = 0.0% in a sample from Kyushu[67] to 7/70 = 10.0% in a sample from Tokushima[67]).[62][63][75][76] Haplogroups N-M231, O-M119, O-K18, and Q-M242 also have been observed with low frequency among present-day Japanese.

A comprehensive study of worldwide Y-DNA diversity (Underhill et al. 2000) included a sample of 23 males from Japan, of whom eight (35%) belonged to haplogroup D-M174, six (26%) belonged to O-M175, five (22%) belonged to O-M122, three (13%) belonged to C-M8 and C-M130, and one (4.3%) belonged to N-M128.[77]

Among 259 males from Japan (70 from Tokushima, 61 from Shizuoka, 53 from Kyūshū, 45 from Okinawa, 26 from Aomori, and 4 Ainus) whose Y-DNA has been examined in a 2005 study by Michael F. Hammer, ninety (34.7%) belong to haplogroup D-M55, eighty-two (31.7%) belong to haplogroup O-P31 (including 22% O-47z, 7.7% O-M176(x47z), and 1.9% O-M95(xM111)), fifty-two (20.1%) belong to haplogroup O-M122, fourteen (5.4%) belong to haplogroup C-M8, ten (3.9%) belong to haplogroup NO-M214(xO-M175) (including 2.3% NO-M214(xO-M175, N-LLY22g), 1.2% haplogroup N-LLY22g(xM128, P43, M178), and 0.4% haplogroup N-M178), and eight (3.1%) belong to haplogroup C-M217 (including 1.9% haplogroup C-M217(xM86) and 1.2% haplogroup C-M86).[35]

The patrilines belonging to D-P37.1 were found in all the Japanese samples, but were more frequently found in the Ainu (75.0%) and Okinawa (55.6%) samples and less frequently found in the Tokushima (25.7%) and Kyūshū samples (26.4%).[35] Haplogroups O-M175 and C-M8 were not found in the small Ainu sample of four individuals, and C-M217 was not found in the Okinawa sample of 45 individuals.[35] Haplogroup N was detected in samples of Japanese from Aomori (2/26 N-LLY22g(xM128, P43, M178)), Shizuoka (1/61 N-LLY22g(xM128, P43, M178)), and Tokushima (1/70 N-M178), but was not found in the Kyūshū, Okinawa, or Ainu samples.[35] This study, and others, report that Y-chromosome patrilines crossed from the Asian mainland into the Japanese archipelago, and continue to make up a large proportion of the Japanese male lineage.[78] If focusing haplogroup O-P31 in those researches, the patrilines derived from its subclade O-SRY465 are frequently found in both Japanese (mean 32%,[79] with frequency in various samples ranging from 26%[80][81] to 36%[82]) and Koreans (mean 30%,[83] with frequency in various samples ranging from 19%[80][84] to 40%[82]). According to the research, these patrilines have undergone extensive genetic admixture with the Jōmon period populations previously established in Japan.[35]

A 2007 study by Nonaka et al. reported that among a total of 263 healthy unrelated Japanese male individuals born in 40 of the 47 prefectures of Japan, but especially Tokyo (n=51), Chiba (n=45), Kanagawa (n=14), Saitama (n=13), Shizuoka (n=12), and Nagano (n=11), the frequencies of the D2, O2b, and O3 lineages were 38.8%, 33.5%, and 16.7%, respectively, which constituted approximately 90% of the Japanese population. Haplogroup diversity for the binary polymorphisms was calculated to be 86.3%.[41]

Poznik et al. (2016) have reported that the males in the JPT (Japanese in Tokyo, Japan) sample[85] of the 1000 Genomes Project are 20/56 = 36% D2-M179, 18/56 = 32% O2b-M176, 10/56 = 18% O3-M122, 4/56 = 7.1% C1a1-M8, 2/56 = 3.6% O2a-K18, and 2/56 = 3.6% C2-M217.[86]

In a project approved by the Ethics Committee of Tokai University School of Medicine, Ochiai et al. (2016) have reported finding D-M174 (rs2032602 T>C) in 24/59 (40.7%), O-M268 (rs13447443 A>G) in 21/59 (35.6%), C-M130 (rs35284970 C>T) in 8/59 (13.6%), O-P198 (rs17269816 T>C) in 4/59 (6.8%), N-M231 (rs9341278 G>A) in 1/59 (1.7%), and O-P186(xM268, P198) (rs16981290 C>A, rs13447443 A, rs17269816 T) in 1/59 (1.7%) of a sample obtained through buccal swabs from Japanese male volunteers (n = 59) who had given informed consent to participate in the study.[68]

According to an analysis of the 1000 Genomes Project's sample of Japanese collected in the Tokyo metropolitan area, the mtDNA haplogroups found among modern Japanese include D (42/118 = 35.6%, including 39/118 = 33.1% D4 and 3/118 = 2.5% D5), B (16/118 = 13.6%, including 11/118 = 9.3% B4 and 5/118 = 4.2% B5), M7 (12/118 = 10.2%), G (12/118 = 10.2%), N9 (10/118 = 8.5%), F (9/118 = 7.6%), A (8/118 = 6.8%), Z (4/118 = 3.4%), M9 (3/118 = 2.5%), and M8 (2/118 = 1.7%).[56]

A 2011 SNP consortium study done by the Chinese Academy of Sciences and Max Planck Society consisting of 1719 DNA samples determined that Koreans and Japanese clustered near to each other, confirming the findings of an earlier study that Koreans and Japanese are related. However, the Japanese were found to be genetically closer to South Asian populations as evident by a genetic position that is significantly closer towards South Asian populations on the principal component analysis (PCA) chart. Some Japanese individuals are also genetically closer to Southeast Asian and Melanesian populations when compared to other East Asians such as Koreans and Han Chinese, indicating possible genetic interactions between Japanese and these populations.[87]

A 2008 study about genome-wide SNPs of East Asians by Chao Tian et al. reported that Japanese along with other East Asians such as Joseon Koreans and Han Chinese are genetically distinguishable from Southeast Asians and that the Japanese are related to Koreans, who in turn are more closely related to Han Chinese. However, the Japanese are relatively genetically distant from Han Chinese, compared to Koreans.[88] Another study (2017) shows a relative strong relation between all East and Southeast Asians.[89]

Hideo Matsumoto, professor emeritus at Osaka Medical College tested Gm types, genetic markers of immunoglobulin G, of Japanese populations for a 2009 study.[90] According to this study, the Gm ab3st gene is found at notably high frequencies across eastern Siberia, northern China, Korea, Mongolia, Japan, and Tibet.[90] The mean frequency of Gm ab3st for the mainstream Japanese population was found to be 26.0%, with a peak in the Yaeyama Islands (36.4% Yonaguni, 32.1% Ishigaki) among all populations in Japan and peaks in Akita (29.5%) and Shizunai (28.3%) among mainstream Japanese.[90] On mainland Asia, peak frequencies of Gm ab3st were found among Oroqen (44.0%) and Tungus (30.0%) in northeast China and among the north Baikal Buryats (30.7%); however, this gene is also frequent among Eskimos (25.4% Alaska, 24.7% Greenland, 20.5% Chaplin, Russia), Luoravetlans (Koryak 20.0%, Chukchi 15.3%), and Athabaskans (New Mexico Apache 19.7%, Alaska Athabascan 14.3%), and it is not uncommon even as far west as the south shore of the Caspian Sea (8.8% Gilani, 8.5% Mazanderani).[90] Minimum frequencies of Gm ab3st were found in Yakushima (22.0%) among all populations in Japan and in Tsu (23.3%) and Ōita (23.6%) among mainstream Japanese.[90] The data from small, isolated island populations, such as those of Yonaguni, Ishigaki, and Yakushima, were not used when calculating the mean for the mainstream Japanese population.[90] The study also considered Ainu and Korean populations and found Gm ab3st with a frequency of 25.2% among Ainu in Hidaka, Hokkaido and a mean frequency of 14.5% (range 13.1% Pusan, South Korea to 18.6% Yanji, China) among Koreans.[90]

Gm afb1b3, on the other hand, is a southern marker gene possibly originating in southern China on the background of the fb1b3 gene (the modal Gm type among Caucasoids) and found at very high frequencies across southern China, Southeast Asia, Taiwan, Sri Lanka, Bangladesh, Nepal, Assam, and the Pacific Islands.[90] Professor Matsumoto has remarked that the center of dispersal of the Gm afb1b3 gene may be in the Yunnan and Guangxi area of southern China; extremely high frequencies of this gene have been observed in samples of mostly Daic peoples from this region (95.2% Shui in Sandu, Guizhou, 94.2% Zhuang in Guangxi, 91.4% Bouyei in Duyun, Guizhou, 87.5% Miao in Guizhou, 84.0% Dai in Luxi, Yunnan) and from neighboring Laos (97.0% Laotian) and Thailand (89.9% Thai).[90] However, Gm afb1b3 is almost equally common among people in Malaysia (97.3% Kadazan on Borneo, 85.0% Malay), Indonesia (76.6% Sulawesi, 75.2% Java), the Philippines (83.6% Luzon Filipinos, 76.4% Luzon Negritos, 67.2% Mindanao Negritos), Karen people in Thailand (82.3%), Kacharis in Assam (80.9%), Cambodians (76.7%), Taiwanese aborigines (76.2%), Micronesians (88.7%), Melanesians (74.6%), and Polynesians (74.7% Cook Islands, 69.4% Hawaii).[90] The study found that the mean frequency of Gm afb1b3 was 10.6% (range 7.8% Shizunai to 13.0% Osaka) for the general Japanese population. Minimum frequencies (4.0% to 4.4%) of Gm afb1b3 were found among the native people in the Yaeyama and Miyako islands in the extreme south of Japan and among the Ainu (4.3%) in the extreme north of Japan. The author suggested that the somewhat elevated frequency of the Gm afb1b3 gene among the mainstream Japanese compared to the Sakishima islanders and the Ainu may have resulted from some admixture of the mainstream Japanese population at rates as low as 7–8% with southern Asian (from southern China or Southeast Asia as far west as Bangladesh and Nepal) populations having the Gm afb1b3 gene in high frequency.[90]

The other Gm types observed among Japanese are ag (45.8%) and axg (17.6%), which are not so useful for discerning human migrations and genetic relationships because they appear to be retained from a common ancestor of most modern humans and are found in similar proportions (with the frequency of ag being significantly greater than the frequency of axg) in many populations all over the world (aboriginal Australians and Americans, South Asians, Caucasoids, etc.).[90]

A 2017 study conducted by Fumihiko Takeuchi, Tomohiro Katsuya, Ryosuke Kimura and Norihiro Kato compared three genetically distinct Japanese groups, Hondu (Honshu), Ryukyu and Ainu to 26 other Asian populations to analyze the shared ancestry and genetic differentiation between the Japanese people and other Asians.[91] The study revealed for the Japanese as a whole, some genetic components from all of the Central, East, Southeast and South Asian populations are prevalent in the Japanese population with the major components of ancestry profile coming from the Korean and Han Chinese clusters.[91] The major components of the Japanese Hondo cluster is similar to the Korean (87–94%), followed by Han Chinese 1 (0–8%) clusters.[91] The genetic components from the Southeast Asian (Thais, Vietnamese and Malays) and South Asian (Sinhalese and Tamils) clusters were larger for the Ryukyu cluster – Southeast Asian (4–6%) and South Asian (4–6%) – in comparison to the results found in the Hondo cluster – Southeast Asian (0–1%) and South Asian (1–2%).[91]

A recent study (2018) shows that the Japanese are predominantly descendants of the Yayoi people and are closely related to other modern East Asians, especially Koreans and Han Chinese.[92][93] It is estimated that the majority of Japanese only has about 12% Jōmon ancestry or even less.[94]

Recent studies suggest that the Japanese people are predominantly descendants of the Yayoi people, and that the Yayoi largely displaced the local Jōmon.[95]

A genome research (Takahashi et al. 2019) shows that modern Japanese (Yamato) do not have much Jōmon ancestry at all. Nuclear genome analysis of Jōmon samples and modern Japanese samples show strong differences.[66]

Jin Han-jun et al. (2003) said that the distribution of Y-chromosomal haplogroups shows that Koreans have a complex origin that results from genetic contributions from range expansions, most of which are from southern-to-northern China, and genetic contributions from the northern Asian settlement.[80]

Korean males display a high frequency of Haplogroup O-M176 (O1b2, formerly O2b), a subclade that probably has spread mainly from somewhere in the Korean Peninsula or its vicinity,[73][99] and Haplogroup O-M122 (O2, formerly O3), a common Y-DNA haplogroup among East and Southeast Asians in general.[100][12] Haplogroup O1b2-M176 has been found in approximately 30% (ranging from 20%[80][74] to 37%[35]) of sampled Korean males, while haplogroup O2-M122 has been found in approximately 40% of sampled Korean males.[74][101][84] Korean males also exhibit a moderate frequency (approximately 15%) of Haplogroup C-M217.

About 2% of Korean males belong to Haplogroup D-M174 (0/216 = 0.0% DE-YAP,[84] 3/300 = 1.0% DE-M145,[102] 1/68 = 1.5% DE-YAP(xE-SRY4064),[74] 8/506 = 1.6% D1b-M55,[73] 3/154 = 1.9% DE, 18/706 = 2.55% D-M174,[103] 5/164 = 3.0% D-M174[82] 1/75 D1b*-P37.1(xD1b1-M116.1) + 2/75 D1b1a-M125(xD1b1a1-P42) = 3/75 = 4.0% D1b-P37.1,[35] 3/45 = 6.7% D-M174[104]). The D1b-M55 subclade has been found with maximal frequency in a small sample (n=16) of the Ainu people of Japan, and is generally frequent throughout the Japanese Archipelago.[105] Other haplogroups that have been found less commonly in samples of Korean males are Y-DNA haplogroup N-M231 (approx. 4%), haplogroup O-M119 (approx. 3%), haplogroup O-M268(xM176) (approx. 2%), haplogroup Q-M242 and Haplogroup R1 (approx. 2% total), J, Y*(xA, C, DE, J, K), L, C-RPS4Y(xM105, M38, M217), and C-M105.[73][74][73]

Korea Foundation Associate Professor of History, Eugene Y. Park, said that there is no correlation between a Korean person's Y-chromosome DNA haplogroup and their surname or ancestral seat.[106][107]

He Miao et al. (2009) created an artificial combination of equal parts of the Y-chromsomes of the HapMap samples of Han Chinese in Beijing and Japanese in Tokyo. The study said that this artificial combination resembled five populations which included Koreans in South Korea and Koreans in China.[108]

Studies of Korean mitochondrial DNA lineages have shown that there is a high frequency of Haplogroup D4, ranging from approximately 23% (11/48) among ethnic Koreans in Arun Banner, Inner Mongolia[109] to approximately 32% (33/103) among Koreans from South Korea.[81][110] Haplogroup D4 is the modal mtDNA haplogroup among northern East Asians (Japanese, Ryukyuans, Koreans, Manchus, Oroqens, Manchurian Evenks, Daurs, Mongols, northern Han Chinese, Tibetans) in general, with a peak frequency among Japanese and Ryukyuans in Japan. Haplogroup B, which occurs very frequently in many populations of Southeast Asia, Polynesia, and the Americas, is found in approximately 10% (5/48 ethnic Koreans from Arun Banner, Inner Mongolia) to 20% (21/103 Koreans from South Korea) of Koreans.[109][110] Haplogroup A has been detected in approximately 7% (7/103 Koreans from South Korea) to 15% (7/48 ethnic Koreans from Arun Banner, Inner Mongolia) of Koreans.[109][110][81] Haplogroup A is the most common mtDNA haplogroup among the Chukchi, Eskimo, Na-Dene, and many Amerind ethnic groups of North and Central America.

The other half of the Korean mtDNA pool consists of an assortment of various haplogroups, each found with relatively low frequency, such as G, N9, Y, F, D5, M7, M8, M9, M10, M11, R11, C, and Z.

A study of the mtDNA of 708 Koreans sampled from six regions of South Korea (134 from Seoul-Gyeonggi, 118 from Jeolla, 117 from Chungcheong, 114 from Gangwon, 113 from Jeju, and 112 from Gyeongsang) found that they belonged to haplogroup D (35.5%, including 14.7% D4(xD4a, D4b), 7.8% D4a, 6.5% D5, 6.4% D4b, and 0.14% D(xD4, D5)), haplogroup B (14.8%, including 11.0% B4 and 3.8% B5), haplogroup A (8.3%), haplogroup M7 (7.6%), haplogroup F (7.1%), haplogroup M8'CZ (6.5%), haplogroup G (6.1%), haplogroup N9a (5.2%), haplogroup Y (3.8%), haplogroup M9 (2.7%), haplogroup M10 (1.6%), haplogroup M11 (0.42%), haplogroup N(xN9, Y, A, F, B4, B5) (0.28%), and haplogroup N9(xN9a) (0.14%).[111]

A study of 1094 individuals in the Korean Genome Project found that they belonged to haplogroup D (34.19%), haplogroup B (13.89%), haplogroup M(xC, D, G, Z) (13.8%), haplogroup A (8.32%), haplogroup G (8.23%), haplogroup F (7.86%), haplogroup N(xA, B, F, R, Y) (5.76%), haplogroup C (3.02%), haplogroup R(xB, F) (2.01%), haplogroup Y (1.74%), and haplogroup Z (1.19%).[112] The individuals sampled for the Korean Genome Project are mostly from the Ulsan metropolitan region.[112]

Jin Han-jun et al. (1999) said that, based on genetic studies of classic genetic markers of protein and nuclear DNA, Koreans tend to be closely genetically related to Mongols among East Asians, which is supported by the following studies: Goedde et al. (1987); Saha & Tay (1992); Hong et al. (1993); and Nei & Roychoudhury (1993). The study said that the mtDNA 9‐bp deletion frequency in the intergenic COII/tRNA^Lys region of Mongols (5.1%) is lower than that of Chinese (14.2%), Japanese (14.3%) and Koreans (15.5%). The study said that these 9‐bp deletion frequencies suggest that Koreans are closely related to Japanese and Chinese and that Koreans are not so closely related to Mongols. The study said that the homogeneity in the 9-bp deletion frequencies among Chinese (14.2%), Japanese (14.3%) and Koreans (15.5%), only spanning from a low of 14.2% for Chinese to a high of 15.5% for Koreans, indicates that very few mtDNA are differentiated in these three populations. The study said that the 9‐bp deletion frequencies for Vietnamese (23.2%) and Indonesians (25.0%), which are the two populations constituting Mongoloid Southeast Asians in the study, are relatively high frequencies when compared to the 9-bp deletion frequencies for Mongols(5.1%), Chinese (14.2%), Japanese (14.3%) and Koreans (15.5%), which are the four populations constituting Northeast Asians in the study. The study said that these 9-bp deletion frequencies are consistent with earlier surveys which showed that 9-bp deletion frequencies increase going from Japan to mainland Asia to the Malay Peninsula, which is supported by the following studies: Horai et al. (1987); Hertzberg et al. (1989); Stoneking & Wilson (1989); Horai (1991); Ballinger et al. (1992); Hanihara et al. (1992); and Chen et al. (1995). The study said that Cavalli-Sforza's chord genetic distance (4D), from Cavalli-Sforza & Bodmer (1971), which is based on the allele frequencies of the intergenic COII/tRNA^Lys region, showed that Koreans are more genetically related to Japanese than Koreans are genetically related to the other East Asian populations which were surveyed. The Cavalli-Sforza's chord genetic distance (4D) between Koreans and other East Asian populations in the study, from least to greatest, are as follows: Korean to Japanese (0.0019), Korean to Chinese (0.0141), Korean to Vietnamese (0.0265), Korean to Indonesian (0.0316) and Korean to Mongols (0.0403). The study said that the close genetic affinity between present-day Koreans and Japanese is expected due to the Yayoi migration from China and the Korean Peninsula to Japan which began about 2,300 years ago, a migration which is supported by the following studies: Chard (1974); Hanihara (1991); Hammer & Horai (1995); Horai et al. (1996); Omoto & Saitou (1997). The study said that Horai et al. (1996) detected mtDNA D-loop variation which supports the idea that a large amount of maternal lineages came into Japan from immigrants from the Korean Peninsula after the Yayoi period.[113]

Wook et al. (2000) said that Chu et al. (1998) found that phylogeny which was based on 30 microsatellites indicated that Korean people were closely related to Chinese people from Manchuria and Yunnan, but Kim Wook et al. (2000) found that the high incidence of the DXYS156Y-null variant in northeast Chinese implied that it is possible to exclude these northeastern Chinese populations from being sources which are significant in Korean people. The phylogenetic analysis done by Wook et al. (2000) indicated that Japanese people are genetically closer to Korean people than Japanese people are genetically related to any of the following peoples: Mongolians, Chinese, Vietnamese, Indonesians, Filipinos and Thais. The study said that mainland Japanese having Koreans as their closest genetic population is consistent with the following previous studies: Hammer and Horai (1995); Horai et al. (1996); and Kim et al. (1998). The study found that Koreans are more genetically homogenous than the Japanese, and the study said that this might be due to different sizes of the founding populations and range expansions. The study said that the moderate mean Y-chromosome haplotype diversity value for Koreans might be the result of migrations from East Asia that had a homogenizing influence. The study said that it is more probable that Koreans descend from dual infusions of Y-chromosomes from two different waves of East Asians rather than a single East Asian population due to the dual patterns of the Y-chromosome haplotype distribution found in Koreans.[114]

Kim Jong-jin et al. (2005) did a study about the genetic relationships among East Asians based on allele frequencies, particularly focusing on how close Chinese, Japanese and Koreans are genetically related to each other. Most Koreans were hard to distinguish from Japanese, and the study was not able to clearly distinguish Koreans and Japanese. Koreans and Japanese clustered together in the principal component analysis and the best least-squares tree. The study said that "[c]ommon ancestry and/or extensive gene flow" historically between Koreans and Japanese appears to be "likely" and results in a lot of difficulty finding population-specific alleles that could assist in differentiating Koreans and Japanese.[115]

Jung Jongsun et al. (2010) used the following Korean samples for a study: Southeast Korean (sample regions: Gyeongju, Goryeong and Ulsan), Middle West Korean (sample regions: Jecheon, Yeoncheon, Cheonan and Pyeongchang) and Southwest Korean (sample regions: Gimje, Naju and Jeju). Due to political reasons, the study said that it did not use North Korean samples, but the study said that the "historical migration event of Baekje from Goguryeo Empire (BC 37–AD 668) in Northern Korea imply that Northern lineages remain in South Korea." The study said that the "Northern people of the Goguryeo Empire" are closely related to Mongolians, and the study said that this group of people ruled most of Southwest Korea. The study said that "some of the royal families and their subjects in the Goguryeo Empire moved to this region and formed the Baekje Empire in BC 18–22." Southwest Koreans are closer to Mongolians in the study's genome map than the other two Korean regions in the study are close to Mongolians. Southwest Koreans also display genetic connections with the HapMap sample of Japanese in Tokyo, and, in the neighbor joining tree, the nodes for Southwest Korea are close to Japan. In the study's Korea-China-Japan genome map, some Southwest Korean samples overlap with samples from Japan. The study said that the fairly close relationship, in both the study's genetic structure analysis and genome map, of the Jeju Southwest Korean sample and the HapMap sample of Japanese in Tokyo, Japan, has made the evolutionary relationship of Chinese, Japanese and Koreans become clearer. Southeast Koreans display some genetic similarity with people of Kobe, Japan, which indicates that there might have been links between these regions. The study said that it is possible that outliers in the Gyeongju sample, one of the sampled Southeast Korean regions, and outliers in the Kobe, Japan, sample both have Siberian lineage due to Southeast Koreans having connections with Siberian lineages with respect to grave patterns and culture. The overall result for the study's Korea-Japan-China genome map indicates that some signals for Siberia remain in Southeast Korea. In contrast to the Gyeongju sample, the Goryeong and Ulsan samples, which are both Southeast Korean samples, displayed average signals for the Korean Peninsula. The study said that Middle West Korea was a melting pot in the Korean Peninsula with people traveling from North to South, South to North, and people traveling from East China, including from the Shandong Peninsula. Western Chinese, which included those in the Shandong Peninsula, travelled across the Yellow Sea, and these Western Chinese lived and traded in both China and Korea. In the study's genome map, Middle West Koreans are close to the HapMap sample of Han Chinese in Beijing and, in the neighbor joining tree, the nodes for Middle West Korea are close to China. The overall result for the study's Korea-Japan-China genome map indicates that Middle West Korea displays an average signal for South Korea. Chinese people are located between Korean and Vietnamese people in the study's genome map.[116][117]

Kim Young-jin and Jin Han-jun (2013) said that principal component analysis had Korean HapMap samples clustering with neighboring East Asian populations which were geographically nearby them such as the Chinese and Japanese. The study said that Koreans are genetically closely related to Japanese in comparison to Koreans' genetic relatedness to other East Asians which included the following East and Southeast Asian peoples: Tujia, Miao, Daur, She, Mongols, Naxi, Cambodians, Oroqen, Yakuts, Yi, Southern Han Chinese, Northern Han Chinese, Hezhen, Xibo, Lahu, Dai and Tu. The study said that the close genetic relatedness of Koreans to Japanese has been reported in the following previous studies: Kivisild et al. (2002); Jin et al. (2003); Jin et al. (2009); and Underhill and Kivisild (2007). The study said that Jung et al. (2010) said that there is a genetic substructure in Koreans, but the study said that it found Korean HapMap individuals to be highly genetically similar. The study said that Jin et al. (2009) found that Koreans from different populations are not different in a significant way which indicates that Koreans are genetically homogenous. The study said that the affinity of Koreans is predominately Southeast Asian with an estimated admixture of 79% Southeast Asian and 21% Northeast Asian for Koreans, but the study said that this does not mean that Koreans are heterogenous, because all of the Koreans which were analyzed uniformly displayed a dual pattern of Northeast Asian and Southeast Asian origins. The study said that Koreans and Japanese displayed no observable difference between each other in their proportion of Southeast Asian and Northeast Asian admixture. The study said the 79% Southeast Asian and 21% Northeast Asian admixture estimate for Koreans is consistent with the interpretation of Jin et al. (2009) that Koreans descend from a Northeast Asian population which was subsequently followed by a male-centric migration from the southern region of Asia which changed both the autosomal composition and Y-chromosomes in the Korean population.

Veronika Siska et al. (2017) said that the Ulchi people are genetically closest in the study's panel to the human remains from the Devil's Gate Cave which are dated to about 7,700 years ago. Modern Korean and Japanese, the Oroqen people and the Hezhen people display a high affinity to the human remains from Devil's Gate Cave. Considering the geographic distance of Amerindians from Devil's Gate Cave, Amerindians are unusually genetically close to the human remains from Devil's Gate Cave. Korean genomes display similar traits to Japanese genomes on genome-wide SNP data. In an admixture analysis, when the genes of Devil's Gate is made into a unique genetic component, this new Devil's Gate genetic component is highest in peoples of the Amur Basin, including Ulchi, and makes up about more than 50% of Koreans and Japanese. It also has a sporadic distribution among other East Asians, Central Asians and Southeast Asians.[118]

Hideo Matsumoto, professor emeritus at Osaka Medical College, tested Gm types, genetic markers of immunoglobulin G, of Korean populations for a 2009 study. The Korean populations were populations in Jeju Island, Busan, Gwangju, Kongsan, Jeonju, Wonju, the Kannung of South Korea and a Korean population in Yanji. Matsumoto said that the Gm ab3st gene is a marker for northern Mongoloid possibly originating in Siberia and found at high frequencies across northeast Asia and Tibet. Matsumoto said that the average frequency of Gm ab3st for Koreans was 14.5% which was intermediate between an average frequency of 26% for general Japanese and a frequency of 11.7% which was for a Han Chinese population in Beijing. Matsumoto said that Gm afb1b3 is a southern marker gene possibly originating in southern China and found at high frequencies across Southeast Asia, southern China, Taiwan, Sri Lanka, Bangladesh, Nepal, Assam and parts of the Pacific. However, given the result that the Okinawans being genetically most northern among the Japanese with the highest frequency of the Gm ab3st gene which is assigned to be northern, the term northern and southern used in his study is controversial. Matsumoto said that the average frequency of Gm afb1b3 for Koreans was 14.7% which was intermediate between a frequency of 10.6% for general Japanese and a frequency of 24.1% for Beijing Han Chinese. Matsumoto said that Koreans displayed the northern Mongoloid pattern, but Matsumoto said that Koreans displayed a higher frequency of the southern marker gene, Gm afb1b3, than the Japanese. Matsumoto said that "Japanese and Korean populations were originally identical or extremely close to each other", and Matsumoto said, "It seemed to be during the formation of the contemporary Korean population that such a Gm pattern intermediate between Japanese and the northern Han in China emerged." Matsumoto said that the different Gm pattern between Japanese and Koreans most likely came about from frequent inflows of Chinese and/or northern populations into the Korean Peninsula.[119]

The majority of Mongolians belong to the y-DNA Haplogroup C-M217. Haplogroup C-M217 among the Mongols is characterized by very deep total diversity that dates back to the very origin of haplogroup C-M217 (TMRCA 33,900 [95% CI 31,300 <-> 36,500] ybp[45]) and very shallow diversity in each of the frequently observed subclades: C-M504 (TMRCA 2,900 [95% CI 2,200 <-> 3,700] ybp), C-M86 (TMRCA 3,700 [95% CI 3,000 <-> 4,500] ybp[45]), C-M407 (TMRCA 4,400 [95% CI 3,500 <-> 5,200] ybp[45]), and C-F1756 (TMRCA 5,000 [95% CI 4,200 <-> 5,800] ybp[45]). Of these four subclades, C-M407 is phylogenetically extremely divergent from the others, and is more closely related to subclades of C-M217 that are found among present-day Chinese, Koreans, Japanese, and other East and Southeast Asians; however, among Mongols, C-M407 is found most frequently toward the north and west, among western Buryats[121] and Dorbet Kalmyks.[122][123]

Haplogroup O-M175 and Haplogroup N-M231 are found at medium rates among present-day Mongols. The subclades of Haplogroup O-M175 that have been observed among Mongols tend to be similar to those found among Han Chinese, whereas the subclades of Haplogroup N-M231 that have been observed among Mongols tend to be similar to those found among Nenets, Nganasans, Khakasses, and Tuvans (N-B478) on the one hand or those found among Chukchi, Koryaks, and Asian Eskimos on the other (N-B197).[124] In addition, some members of a wide variety of other Y-DNA haplogroups have been found among present-day Mongols, including Haplogroup Q-M242, Haplogroup R-M207 (Haplogroup R1b-M478, Haplogroup R1b-M269, Haplogroup R1a-M17, Haplogroup R2a-M124), Haplogroup D-M174, Haplogroup J2a-M410, Haplogroup J1-Page8, Haplogroup G1-M285, and Haplogroup I2a2-M436.[125][126][127]

The maternal haplogroups are diverse but similar to other northern Asian populations. The most common maternal haplogroups in Mongolians are haplogroup D4, Haplogroup A, and Haplogroup B.[128]

West Eurasian mtDNA haplogroups Haplogroup HV, Haplogroup U, Haplogroup K, Haplogroup I, Haplogroup J, represents 14% in western Xingjang Mongolian, 10% in Mongolia, 8.4% in central Inner Mongolian samples, 2% in eastern Xin Barage Zuoqi County samples.[129]

A study in 2010 suggested that the majority of the Tibetan gene pool may have diverged from the Zang around 15,000 years ago.[131] However, there are possibilities of much earlier human inhabitation of Tibet,[132][133] and these early residents may have contributed to the modern Tibetan gene pool.[134]

The date of divergence between Tibetans and Sherpas was estimated to have taken place around 11,000 to 7,000 years ago.[130]

After modern Oceanic populations, modern Tibetan populations show the highest rate of allele sharing with archaic hominins at over 6%.[130] Modern Tibetans show genetic affinities to three archaic populations: Denisovans, Neanderthals, and an unidentified archaic population.[130]

In comparison to modern Han populations, modern Tibetans show greater genetic affinity to Denisovans; however, both the Han and Tibetans have similar ratios of genetic affinity to general Neanderthal populations.[130]

Modern Tibetans were identified as the modern population that has the most alleles in common with Ust'-Ishim man.[130]

The distribution of Haplogroup D-M174 (subclade Haplogroup D-Z27276) is found among nearly all the populations of Central Asia and Northeast Asia south of the Russian border, although generally at a low frequency of 2% or less. A dramatic spike in the frequency of D-M174 occurs as one approaches the Tibetan Plateau. D-M174 is also found at high frequencies among Japanese people, but it fades into low frequencies in Korea and China proper between Japan and Tibet. The claim that the Navajo people and Tibetans are related, while discussed among linguists since Edward Sapir, has not found support in genetic studies. Some light has been shed on their origins, however, by one genetic study[135] in which it was indicated that Tibetan Y-chromosomes had multiple origins, one from Central Asia and the other from East Asia.

Common yDNA haplogroups in Turkic peoples are Haplogroup N-M231, Haplogroup C-M217, Haplogroup Q-M242 and Haplogroup O-M175. Some groups also have Haplogroup R1b, Haplogroup J-M172 and Haplogroup D-M174. Ancient samples show that Turks have mostly East-Asian lineages, similar to Mongolian and Han-Chinese samples.[139][140]

The modern Turkic groups in Anatolia (Turkey) and Europe have less relation to East-Asian groups than their Central-Asian relatives. Various studies estimate about 15-30% East-Asian lineages in Anatolian/European Turks with the average at 21.7%.[141]