Skip to main content
NCBI home page
BMC Biology logoLink to BMC Biology
. 2026 Feb 6;24:68. doi: 10.1186/s12915-026-02541-7

Reconstructing the genetic formation of Han Chinese from ancient genomes

Limin Qiu 1,#, Haodong Chen 2,#, Hui Chen 3,#, Chaochao Qin 4,#, Mengting Xu 1, Yiling Jiang 1, Hao Ma 1, Qu Shen 5, Yanying Peng 5, Xiaolu Mao 5, Yilan Liu 1, Baitong Wang 2, Zhi Ji 5, Kongyang Zhu 1, Yu Xu 1, Le Tao 2,6, Haifeng He 1, Hongming Zhou 5, Xinyue Zou 5, Xinyi Wang 5, Jiaxin Tang 5, Daohua Hao 4,✉, Xiaokun Wang 3,✉, Chuan-Chao Wang 7,✉
PMCID: PMC12977919  PMID: 41652444

Abstract

Background

The formation of the Han Chinese is deeply rooted in the Neolithic cultures of the Yellow River basin, particularly the pivotal Longshan cultural sphere which bridged prehistoric societies and early dynastic civilization. However, the genetic impact of Longshan-era populations on subsequent historical groups remains largely unexplored due to a critical lack of ancient genomic data from this key transitional period. This gap hinders a clear understanding of how early cultural integration in the heartland shaped the genetic structure of later Chinese populations.

Results

This study reports 28 newly sequenced ancient human genomes from the Han Dynasty Xujiacundong and Zhouhe archaeological sites in Shandong Province, which are integrated with previously published regional datasets to investigate the genetic legacy of Neolithic Longshan populations in the formation of Han Chinese ancestry. Our analyses reveal pronounced genetic differentiation between Longshan populations from the Central Plain and lower Yellow River basin during the Late Neolithic period. Most individuals from the Xujiacundong site exhibit mixed ancestry, predominantly derived from Central Plain Longshan-related ancestry (93.8%) with a minor contribution from southeastern coastal China-related ancestry (6.2%). In contrast, all individuals from the Zhouhe site exhibit genetic homogeneity with Central Plain Longshan-related ancestry. These results indicate substantial genetic heterogeneity within the lower Yellow River basin during the Han Dynasty. Moreover, we found a high degree of genetic homogeneity between ancient Han Dynasty populations and modern Han Chinese from Shandong. Admixture modeling and f-statistics further demonstrate that Longshan-related ancestries—particularly those associated with the Central Plain—played a dominant role in shaping the genetic structure of historical populations across a wide geographic range, including the Upper Yellow River, the West Liao River Basin, and Southwest China, etc.

Conclusions

These findings underscore the profound and pervasive genetic influence of the Central Plain Longshan populations on surrounding regions, driving the demographic expansion and genetic homogenization of the Han Chinese. This interplay of population movements and cultural diffusion highlights the central role of Longshan-era demic expansion in shaping the genetic landscape and cohesion of the Han people.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-026-02541-7.

Keywords: Ancient genome, Demic diffusion, Han Chinese, Xujiacundong site, Zhouhe site

Background

As the world’s largest ethnic group, the Han Chinese constitute approximately 91.11% of China’s population and represent roughly 17.5% of the global demographic composition. This demographic majority has played a pivotal role in shaping the historical trajectory and cultural evolution of Chinese civilization. Emerging from the fertile cradle of the Central Plain (Zhongyuan, 中原), the Han Chinese trace their ancestral origins to the Huaxia (华夏) civilization—an ancient agricultural tribal confederation that developed along the middle and lower Yellow River basin during the Neolithic era. The Yellow River culture formed the main vein of the origin of Chinese civilization. With in-depth exchange and integration of the middle and lower reaches as the main body, the Yellow River culture has gathered the advanced factors of the Yangtze River and various similar regional cultures [1]. Since the Neolithic Age, ancestral communities, including the Huaxia (华夏), Dongyi (东夷), and Xirong (西戎) tribes, had thrived along the Yellow River [2]. The Huaxia are the progenitors of Chinese civilization and ancestors of the modern Han Chinese. By the Qin, Han, Wei, and Jin dynasties (221 BCE-420 CE), the middle and lower Yellow River basin gradually evolved into the core cultural zone of the Han ethnicity and China’s dominant ethnic group [2]. Throughout the prolonged formative period of Chinese civilization, this cultural system primarily emerged through profound interactions and syncretism between midstream and downstream cultures, while continuously assimilating advanced elements from the Yangtze River basin and other regional cultures, thereby establishing itself as the foundational matrix of Chinese civilizational origins [1]. From the dawn of civilization through the Tang and Song dynasties (618–1279 CE), the Yellow River valley consistently maintained its status as the nation’s political, economic, and cultural center, solidifying its position as the primary cradle of Chinese civilization and laying the foundation for the formation of the Han Chinese.

The lower reaches of the Yellow River witnessed the early civilizations of the Yellow River basin, characterized by continuous development from the Houli culture (circa 7000–5000 BCE) through the Beixin culture (circa 5400–4400 BCE), the Dawenkou culture (circa 4200–2600 BCE) and the Shandong Longshan culture (circa 2500–2200 BCE), etc. The Longshan culture, named after its discovery in Longshan Town, Shandong Province, broadly refers to a type of cultural relics from the late Neolithic period in the middle and lower reaches of the Yellow River in China. The initial Shandong Longshan culture was inherited from the Dawenkou culture and preceded the Yueshi culture (circa 2500–2000 BCE). Subsequently, archaeologists have discovered culturally related relics in various locations: I. the Miaodigou Second Phase culture (radiocarbon calibrated to 2900–2800 BCE) in the Central Plain, derived from the Yangshao culture (circa 5000–3000 BCE), which evolved from the Yangshao Culture and represents the early stage of the Longshan Culture in the Central Plain; II. Henan Longshan Culture, which later developed into the early Bronze Age civilization of the Central Plains (radiocarbon calibrated to 2600–2000 BCE); III. Shaanxi Longshan Culture, also known as Keshengzhuang Second Phase Culture (radiocarbon calibrated to 2300–2000 BCE), etc. [3]. The Longshan culture is, in fact, a complex comprising many distinct archaeological features. For instance, the Longshan cultures of the Central Plain and Shandong regions are considered to be potentially independent cultural systems with separate origins and development [4]. The Longshan Culture maintained varying degrees of interaction with contemporaneous cultures. From a perspective emphasizing these interconnections, this period may be termed the “Longshan Era”, chronologically situated between the Miaodigou Second Phase/Dawenkou Cultures and the Erlitou/Yueshi cultures [3]. The Longshan Era marked the transition from the Neolithic to the Bronze Age, and corresponded to the emergence of the legendary Three Dynasties (Xia, Shang, Zhou, 夏商周) [3]. Archaeological research suggests that the early socio-economic development of the Haidai region, located in the lower reaches of the Yellow River, was significantly ahead of that of the Central Plain [5]. However, in the later period, a great flood (possibly related to the legend of “Yu the Great Tames Floods”, 大禹治水) caused a cultural discontinuity [6], evidenced by the lack of a clear succession between the Yueshi and Shandong Longshan culture [7, 8]. In contrast, the Longshan culture of the Central Plain continued to develop in a relatively stable environment, ultimately giving rise to the Xia Dynasty civilization during the Erlitou culture period [9]. Gradually, as a pivotal cultural phase of the pre-Qin period, the Longshan Era established a centripetal distribution pattern centered around the core area of the Central Plain Longshan culture. This cultural framework not only influenced the development of regional cultures but also laid the groundwork for the “pluralistic unity” configuration of the Chinese nation. The cultural achievements of the Longshan Era, such as advanced pottery-making techniques, the emergence of early urban settlements, and the development of social hierarchies, were crucial steps in the prehistoric cultural evolution that eventually led to the formation of early Chinese civilization.

A series of analyses based on WGS or SNP arrays have revealed genetic differences among modern Han populations from different regions [10–12], with smaller genetic variation within northern populations and increasing genetic diversity as one moves southward. Genetic evidence supports the Qinling-Huaihe Line and the Nanling Mountains as geographical boundaries for Han subgroups, with northern and southern Han populations sharing a common ancestry [12]. Studies based on Y-chromosomes and mitochondrial DNA (mtDNA) from modern Han populations suggest that the large-scale migration of northern Han populations led to the demographic expansion of Han populations and their culture [13]. This aligns with historical records indicating that the ancestors of the Han people can be traced back to the Huaxia tribes in the middle reaches of the Yellow River, who continuously expanded into surrounding areas. However, due to the lack of ancient genomic data, the genetic patterns of the Han population during its early formative period remain unclear. The term “Han” not only refers to a specific ethnic collective but also points to a shared ancestry, history, and cultural identity.

The name “Han Chinese”, alternatively the “Han people” (汉人; Hànrén), first appeared during the Northern and Southern period and was inspired by the Han dynasty, which is considered to be one of the first golden ages of China. The Han dynasty’s prestige and prominence led many of the ancient Huaxia to identify themselves as Han people. The Han Dynasty did not emerge in a cultural vacuum. It absorbed, refined, and systematized core elements that originated in the prehistoric era and were transmitted through intermediate dynasties. Previous ancient genome studies have revealed the profound genetic influence of the ancient population from the middle Yellow River region on ancient populations from the upper/lower Yellow River, northeastern China and southwest China, a legacy traceable back to the Yangshao Culture period (contemporaneous with the Dawenkou Culture period in the lower Yellow River region) that predates even the Longshan Culture Era [14–20]. In this study, we newly reported 28 ancient individuals from the Han Dynasty, which were discovered at the Xujiacundong and Zhouhe sites in Shandong Province. These two archaeological sites are geologically close to published ancient populations (Qinglanfu_Historic, Shandong_HE). They were all excavated in Linzi City, Shandong Province. By integrating these with previously published ancient genomes from other regions of East Asia in the post-Longshan Era [14–19, 21, 22], we systematically investigated the genetic impact of Longshan culture-related ancestral components on historical populations (Fig. 1A). We aimed to uncover the deeper connections between cultural transitions and population dynamics, offering new perspectives for research on the origins of Chinese civilization, as well as the genetic influence of ancient populations from the lower Yellow River region during the Han Dynasty on surrounding present-day Han Chinese.

Fig. 1.

Fig. 1

Open in a new tab

Geographic locations and the genetic profile of newly generated Xujiacundong and Zhouhe ancients from the lower reaches of Yellow River. The relevant reference data can be found in Additional file 1: Table S1C and Additional file 1: Table S1D. A Geographical locations of 68 ancient populations in this study, including our new XJCDM and ZHM samples. The map was made with Natural Earth: free vector and raster map data @naturalearthdata.com. B Principal component analysis (PCA) with most populations in eastern Eurasia. We projected the ancient samples onto dimensions computed by present-day populations. Modern individuals are represented by gray dots in the figure. C Calibrated radiocarbon dates of ancient samples used in PCA, see corresponding legend below

Results

Newly generated ancient genome-wide data

For this study, we extracted DNA from 58 ancient human remains from the Han Dynasty at the Xujiacundong and Zhouhe sites in Zibo City, Shandong Province (Additional file 1: Table S1A). These two archaeological sites are geographically adjacent to each other. We constructed double-strand libraries for each sample without treatment with uracil-DNA glycosylase (UDG), and then performed shotgun sequencing. All genomes displayed typical damage patterns of ancient DNA (Additional file 2: Fig. S1). Pseudo-haploid genotypes were called on the targeted SNPs by randomly sampling a single allele at each position on a 1.24 million SNP panel. We removed 25 individuals for either high levels of contamination (> 3%) or low coverage of SNPs (< 30,000). The number of SNPs covered by the 1240K panel for the remaining individuals ranges from 36,682 to 1,006,484. Additionally, we identified five pairs of individuals with genetic relatedness up to the third degree, occurring both within each archaeological site and between the two sites. Given the close geographical proximity of Xujiacundong and Zhouhe, these kinship links suggest the presence of inter-site interactions, potentially including intermarriage, among the ancient populations (Additional file 2: Fig. S2, Additional file 1: Table S1A). To avoid bias in population genetic analyses, only the higher-coverage individuals from each related pair were retained for downstream analyses. Finally, twenty-eight unrelated individuals remained for further analysis. Three human teeth of samples XJCDM110, XJCDM354, and ZHM3 were analyzed for conventional radiocarbon dating (Additional file 2: Fig. S3). The calibrated radiocarbon dates fall within approximately 100–200 BCE, corresponding to the Han Dynasty, and are consistent with the archaeological chronology inferred from the associated cultural context (Additional file 1: Table S1A). We initially divided the studied individuals into two groups based on their respective archaeological sites and labeled them as XJCDM_Han_Dynasty and ZHM_Han_Dynasty. Subsequently, based on preliminary PCA and individual-level qpWave analyses, we identified one genetically distinct individual from the Xujiacundong Han Dynasty site, which was designated as XJCDM_Han_Dynasty_o1 (see Methods for details on outlier detection using qpWave).

Y chromosomal and mtDNA haplogroup assignment

Among the remaining 28 individuals, 10 were biologically determined to be male, and 18 were biologically determined to be female. Mitochondrial haplogroups exhibit high diversity but are all characteristic of East Asian lineages (Fig. 2A). For instance, individuals classified under haplogroup M include subclades such as M7 and M8, which are broadly distributed in East Asia [23–25]. Haplogroup D is predominantly found in northern East Asia [26], while haplogroup B is more prevalent along the Pacific coast of eastern Asia [27], etc. The Y chromosomal haplogroups primarily fall into four categories: O, N, Q, and C (Fig. 2B). Haplogroup O, the predominant paternal lineage in East Asia [28], has the highest frequency in these samples (40%). Haplogroup N, widely distributed across northern China, Mongolia, and Siberia [29], constitutes a significant proportion (30%). The two individuals assigned to haplogroup Q belong to its subclade Q1a1a, which is hypothesized to have originated in ancient Siberian populations and later integrated into the paternal ancestry of Han Chinese over two millennia ago [30]. One individual belongs to haplogroup C2b1a2a, which is primarily observed in northern China and North Asia [31].

Fig. 2.

Fig. 2

Open in a new tab

The genetic origin of newly generated Xujiacundong and Zhouhe ancients. More information can be found in Additional file 1: Table S1A. A Mitochondrial haplogroups of ancient samples in XJCDM and ZHM (n = 28). B. Y-chromosome haplogroups of ancient male samples in XJCDM and ZHM (n = 10). C Outgroup-f3 analysis in the form of f3 (Target, X; Mbuti), in which X represent the ancient populations in East Asia. The top 30 f3 values are displayed here. D f4 analysis in the form of f4 (Mbuti, X; Studied pop, YR_LN) based on "merged 1240K" dataset, in which X (on the x-axis) denotes other ancient/present-day populations unrelated to middle-lower YR, such as ANA, ANE, southwest China and southeast China, etc. Studied pop (on the y-axis) represents ancient populations from diverse regions of East Asia dating to the post-Longshan Era

Genetic structure of populations in the Central Plain and the lower Yellow River region

We initially performed principal component analysis (PCA) by projecting ancient genomes onto the first two principal components of variation constructed from modern East Asians. The populations were distributed across three distinct clusters, corresponding to coastal and inland populations in southeast China, Northeast Asian populations, and high-altitude populations (Fig. 1B, Additional file 2: Fig. S4). Notably, individuals associated with the Shandong Longshan cultural context did not genetically cluster with contemporaneous Longshan populations from the Central Plain, indicating pronounced regional differentiation during the Late Neolithic. Our newly reported ancient individuals, together with previously published ancient individuals from historical periods in the Lower Yellow River region (e.g., Shandong_1k, Shandong_2k, Shandong_HE), did not cluster with earlier Dawenkou- and Longshan-related populations from the same region (e.g., Liulin_MDWK, Xixiahou_LDWK, Sanlihe_LDWK, Sanlihe_LS, and Wutai_LS). Instead, these historical-period individuals exhibited closer genetic affinity to earlier ancient populations from the Central Plain, such as YR_LBIA and YR_LN.

In the unsupervised ADMIXTURE analysis, the lowest cross-validation (CV) error was observed at K = 3, suggesting that three ancestral components provide the best-supported model for the selected populations (Additional file 2: Fig. S5A). At this resolution, individuals from our study population are primarily composed of three ancestry components whose overall profiles and relative proportions are comparable to those observed in ancient populations from the middle Yellow River region during the Longshan period (Additional file 2: Fig. S5B). These components broadly correspond to genetic affinities related to ancient populations from the eastern coastal regions of China (shown in yellow), coastal and inland southeastern China (shown in orange), and northern East Asia and Northeast Asia (shown in blue).

The genetic differentiation between the Central Plain Longshan and Shandong Longshan ancestries

To further investigate the genetic influence of individuals from the Central Plain Longshan Culture and Shandong Longshan Culture periods on our target populations, we performed f4-statistics analyses. For consistency in subsequent discussions, we first grouped published Longshan-related ancestral populations into two categories: one is YR_LS, encompassing punished Central Plain Longshan-related ancestries YR_LN (n = 8) and YR_Yangshaocun_Longshan (n = 7), another is SD_LS, encompassing Shandong Longshan-related ancestries Sanlihe_LS (n = 4), Chengzi_LS (n = 1), and Wutai_LS (n = 6). First, f4(Mbuti, coastal and inland southeastern China; YR_LS, SD_LS) yielded significantly positive results (Z-scores > 3; Additional file 1: Table S2A). This indicates that SD_LS shared more alleles with populations from coastal and inland southeastern China than YR_LS did. Second, we compared these two groups against early local hunter-gatherers in Shandong (Shandong_EN). The f4(Mbuti, Shandong_EN; YR_LS, Wutai_LS) showed significantly positive values (Z-scores > 3; Additional file 1: Table S2A), whereas no significant signals were detected in f4(Mbuti, Shandong_EN; YR_LS, Sanlihe_LS/Chengzi_LS). This suggests that among three SD_LS populations, only Wutai_LS shared more alleles with Shandong_EN than YR_LS did—while Sanlihe_LS and Chengzi_LS showed no such pattern. Collectively, these results demonstrate that Shandong Longshan Culture ancestries received significantly more gene flow from two sources—southeastern coastal Chinese populations and local Shandong hunter-gatherers—compared to Central Plain Longshan Culture ancestries. This provides genetic evidence to support the archaeological hypothesis that the Central Plain Longshan Culture and Shandong Longshan Culture were likely two distinct cultural systems, each with independent origins and developmental trajectories [4].

The genetic formation of ancient Han Chinese in the lower Yellow River region

The outgroup f3 analysis in the form of outgroup-f3(Target, X; Mbuti) revealed that our studied population shared the greatest genetic drift with ancestral populations along with the Yellow River (Fig. 2C, Additional file 1: Table S2B).

Integrated with PCA, ADMIXTURE, and outgroup-f3 analyses—alongside historical records documenting increasing cultural exchange and eventual integration between the middle and lower Yellow River regions from the Longshan Era to the Han Dynasty—the Central Plain Longshan-related ancestries likely contributed genetically to populations along the lower Yellow River during the post-Longshan Era. The significantly negative values of f4(Mbuti, middle YR-related; XJCDM_Han_Dynasty/ZHM_Han_Dynasty, SD_LS) (Z-scores < − 3) (Additional file 1: Table S2C) also suggested that XJCDM_Han_Dynasty/ZHM_Han_Dynasty shared more alleles with middle Yellow River-related ancestries compared to Shandong Longshan-related ancestries.

We detected no significantly negative values in f4(Mbuti, Shandong_EN; Target, SD_LS) (Additional file 1: Table S2D), indicating that our target populations showed no detectable genetic influence from Shandong_EN. Additionally, Sanlihe_LS (n = 4) offers more robust genetic data than Chengzi_LS (n = 1, a single individual). Combining these f4-statistic results and sample size considerations, we tentatively selected Sanlihe_LS as the representative of SD_LS. A previous study compared YR_LN and YR_Yangshaocun_Longshan (both belonging to YR_LS), finding that YR_Yangshaocun_Longshan contained more Ancient Northeast Asian (ANA)-related ancestry, whereas YR_LN showed no detectable ANA-related ancestry [32]. We then tested our target populations using f4(Mbuti, ANA-related; Target, SD_LS) (with published Amur River ancient populations as ANA representatives; Additional file 1: Table S2D) and also detected no significant ANA-related ancestry. Given the consistency in ANA profiles between our targets and YR_LN, we selected YR_LN as the representative of YR_LS.

When using YR_LN as the representative of YR_LS, we detected significantly negative values in the form of f4(Mbuti, X; XJCDM_Han_Dynasty, YR_LN) (Z-scores < − 3) when X represented populations in southeastern coastal and inland China (e.g. Liangdao2, Ami) (Additional file 1: Table S2D), but no significantly negative values occurred in f4(Mbuti, X; ZHM_Han_Dynasty, YR_LN) (|Z-scores|< 3) (Additional file 1: Table S2D). Besides, no significant values were detected in f4(Mbuti, X; XJCDM_Han_Dynasty_o1, YR_LN) (|Z-scores|< 3) (Additional file 1: Table S2E).

To further validate the ancestral sources of our target populations, we expanded this analysis with qpAdm modeling. We selected three potential ancestral sources representing distinct regions: YR_LN (for middle Yellow River basin ancestry), Sanlihe_LS (for lower Yellow River basin ancestry), and the Ami (for southeastern coastal/inland ancestry). Using a rotating strategy to ensure model robustness, we found that XJCDM_Han_Dynasty could be modeled by ~ 93.8% YR_LS-related + ~ 6.2% southeastern coastal/inland China-related for a successful fit. In comparison, the outlier XJCDM_Han_Dynasty_o1 could be successfully modeled as one-way deriving from 100% ancestry from YR_LS-related. Additionally, ZHM_Han_Dynasty could also be successfully modeled as deriving entirely from YR_LS-related in a one-way model. We also performed ancestral modeling on published Historical Era ancient individuals from various archaeological sites in the lower Yellow River region, and the results showed that the majority could either be modeled as one-way of YR_LS or as a two-way modeling of ~ 56.6%–62.3% YR_LS-related + ~ 37.7%–43.4% SD_LS-related (Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

The well-fitted qpAdm modelling for XJCDM_Han_Dynasty, XJCDM_Han_Dynasty_o1, ZHM_Han_Dynasty, and other published ancient East Asians that own YR_LS-related ancestries (more details in Additional file 1: Table S2G). The relevant reference data can be found in Additional file 1: Table S1C and Additional file 1: Table S1D

Population genetic structure of ancient Chinese in other regions

During the later period of the Longshan Era, the Central Plain Longshan civilization had already formed the cultural core of the broader Chinese territory [33]. To investigate the regional genetic landscape systematically, we combined published ancient genomic datasets from across China with our newly sequenced individuals [17–19, 22], aiming to assess the contributions of Longshan-related ancestral components to these populations.

For the populations from the upper Yellow River region, including individuals excavated from the Dacaozi, Heishuiguo, and Foyemiaowan sites, we performed f4-statistics of the form f4(Mbuti, X; upper Yellow River, YR_LN), where X represents populations from Eurasia outside the Yellow River Basin, and YR_LN represents the Central Plain Longshan-related ancestry. No significant results were detected (|Z-scores|< 3), except for a moderately negative value in f4(Mbuti, southeast China; Heishuiguo_H, YR_LN) (− 3 < Z-scores < − 2) (Fig. 2D, Additional file 1: Table S2F). Using qpAdm, Dacaozi_H (also referred to as Upper_YR_IA) and Foyemiaowan_H could be modeled entirely as deriving from YR_LS-related ancestry, whereas Heishuiguo_H required a two-way model composed of ~ 92.9% YR_LS-related and ~ 7.1% southeast coastal China-related ancestries, represented by Ami (Fig. 3, Additional file 1: Table S2G).

In the middle Yellow River region during the post-Longshan Era (e.g., YR_Spring_and_Autumn_period, YR_Western_Zhou_Dynasty; corresponding archaeologically to the Bronze Age), f4-statistics revealed no significant deviations (|Z-scores|< 3) when compared with YR_LN (Fig. 2D, Additional file 1: Table S2F). qpAdm modeling further confirmed that these populations could be represented as deriving entirely from YR_LS-related ancestry (Fig. 3, Additional file 1: Table S2G).

Previous ancient genomic studies have documented frequent population turnovers in the Yellow River (YR) and the West Liao River (WLR) regions [18]. Consistent with these observations, Bronze Age individuals from the Longtoushan site in the WLR exhibited significantly negative f4-statistics with Ancient Northeast Asian (ANA)-related populations (f4(Mbuti, ANA-related; WLR_BA, YR_LN), Z-scores < − 3) (Fig. 2D, Additional file 1: Table S2F), indicating that WLR_BA shared more alleles with ANA-related ancestries compared to YR_LS-related. qpAdm modeling estimated that WLR_BA derived approximately 48.2% of its ancestry from YR_LS-related sources and 51.8% from ANA-related ancestry, represented by AR_EN here (Fig. 3, Additional file 1: Table S2G).

Finally, for the Songshan population from southwestern China, significantly negative values were detected in f4(Mbuti, southern China; Songshan, YR_LN) (Z-scores < − 3) (e.g., BaBanQinCen, Dushan, Shenxian, etc.), and moderately significant negative values were observed in f4(Mbuti, coastal and inland southeastern China; Songshan, YR_LN) (− 3 < Z-scores < − 2) (e.g., Liangdao2, Baojianshan, Taiwan_Hanben, etc.), indicating that the Songshan population shared more alleles with southern Chinese ancestries compared to YR_LN (Fig. 2D, Additional file 1: Table S2F). Songshan could be successfully modeled as ~ 70.9% YR_LS-related + ~ 17.7% southeastern coastal China-related + ~ 11.4% southern East Asia (sEA) ancestries (represented by Baojianshan in Guangxi) by qpAdm (Fig. 3, Additional file 1: Table S2G).

Genetic influence of XJCD_Han_Dynasty/ZHM_Han_Dynasty on present-day Han Chinese in the Shandong Province

The Han Chinese have played a central role in shaping the development and continuity of Chinese civilization. Originating in the Central Plain (Zhongyuan), they trace their ancestry to the Huaxia people [34]. Han Chinese people and culture later spread around the Chinese mainland, driven by large and sustained waves of migration during successive periods of Chinese history, for example, the Qin (221–206 BCE) and Han (202 BCE–220 CE) dynasties, and absorbed various non-Han ethnic groups over the centuries at various points in Chinese history. Thus began the gradual emergence of long-term genetic stability in China. Consistent with this historical framework, principal component analysis shows that modern northern Han Chinese cluster closely with individuals from the Xujiacundong and Zhouhe Han Dynasty populations (Additional file 2: Fig. S4). To assess the contribution of ancient populations in the lower Yellow River region to contemporary Shandong Han, we performed qpAdm modeling using the newly sequenced Han Dynasty genomes. Notably, The modern Shandong Han population can be successfully modeled using individuals from the Zhouhe site—who represent a genetic continuity of the Central Plains Longshan ancestry—as the sole ancestral source. (Additional file 1: Table S2H).

Discussion

The emergence of the Han Chinese as a pan-regional ethnic entity reflects a millennia-long process of cultural integration and demographic expansion, rooted in Neolithic interactions among diverse populations across the Yellow River basin. Here, we integrate archaeological and genomic evidence to elucidate how Longshan culture-associated genetic components, which were instrumental in the formation of early state-level societies in China, shaped the genetic landscape of historical Han populations. Through the analysis of 28 ancient genomes from the Han Dynasty in present-day Zibo, Shandong, we uncover a nuanced interplay between regional genetic diversity and the predominant influence of Central Plain Longshan ancestry, providing novel insights into the “demic diffusion” model underlying Han cultural and genetic expansion.

Regional genetic dynamics from the late Neolithic to the Han Dynasty in the lower Yellow River

The genetic distinction between the Central Plain Longshan and Shandong Longshan populations during the late Neolithic aligns with archaeological evidence suggesting these as distinct cultural entities with independent origins [3, 4]. The Central Plain Longshan culture, centered in present-day Henan and Shanxi, laid the foundations for the Erlitou and Xia Dynasty civilizations, whereas the Shandong Longshan culture, inheriting traits from the Dawenkou tradition, developed a distinctive material culture in the Haidai region. Although archaeological findings indicate cultural interactions, such as the presence of Central Plain-style pottery in Shandong, genomic data reveal almost direct genetic continuity between the late Dawenkou and Shandong Longshan populations.

By the Han Dynasty, most individuals from Xujiacundong exhibited admixture of Central Plain Longshan-related (~ 93.8%) and southeastern coastal/inland China-related (~ 6.2%) ancestries, reflecting centuries of population blending. In contrast, the outlier individual from the Xujiacundong site and individuals from the Zhouhe site exhibit genetic homogeneity with Central Plain Longshan-related ancestry, consistent with a direct demographic expansion from the Central Plain, likely driven by population movements associated with the centralized governance of the Han Dynasty. Furthermore, the genetic profiles of historical-period individuals from other archaeological sites across the Lower Yellow River region reflect notable genetic diversity in the area, alongside the enduring and substantial influence of Central Plains Longshan-related ancestry on ancient Shandong populations during the Historical Era. Notably, we detected no significant genetic contributions from earlier local Shandong hunter-gatherers to these Han Dynasty populations.

The Central Plain as a genetic hub: Pan-Chinese diffusion of Longshan ancestry

The widespread prevalence of Central Plain Longshan-related ancestry across diverse regions—from the Upper Yellow River to Southwest China—underscores its role as a foundational genetic substrate for Han populations. In the Upper Yellow River Basin, individuals from archaeological sites such as Dacaozi and Foyemiaowan derived entirely from Central Plain Longshan-related ancestry, consistent with archaeological evidence for the westward spread of millet agriculture and associated cultural practices during the Neolithic. The modest admixture of southeastern coastal ancestry observed in Heishuiguo (~ 7.1%) likely reflects later interactions along trade networks or southward migrations, yet the Central Plain component remained overwhelmingly dominant.

In the West Liao River Basin, the Upper Xiajiadian Culture population at Longtoushan exhibits a roughly equal mixture of Central Plain Longshan-related (~ 48.2%) and ANA-related ancestry (~ 51.8%), highlighting the genetic interplay between southern agriculturalists and northeastern hunter-gatherer farmers. This admixture corresponds with archaeological evidence for the diffusion of millet farming into the Liao River valley, indicating that while Central Plain migrants introduced agricultural technologies, they also intermarried with local populations, producing a hybrid genetic profile.

In Southwest China, the Songshan population during the Historical Era shows ~ 70.9% Central Plain Longshan-related ancestry combined with southern Chinese components, consistent with historical records of Han expansion into the Yunna-Guizhou Plateau during the Han and Tang dynasties. Previous studies have associated this with the spread of millet farming along the “Tibetan-Burman Corridor” [20], where Central Plain migrants intermixed with indigenous groups, leading to a genetic landscape dominated by but not exclusive to Longshan-related ancestry.

Demic diffusion, cultural integration, and Han ethnic formation

Our findings provide strong support for the “demic diffusion” model proposed by Wen et al. (2004), in which large-scale migrations of agriculturalists from the Central Plain drove both cultural and genetic expansion. The genetic homogeneity of Han populations in the Yellow River basin—even 2000 years ago—contrasts with the increasing diversity observed in southern Han populations, likely resulting from successive waves of north–south migration and admixture with local populations [12, 35]. Central Plain Longshan ancestry appears to have functioned as a “genetic glue”, promoting pan-regional homogenization, while localized admixture events—such as contributions from Shandong Longshan in the east, Ancient Northeast Asian (ANA) populations in the northeast, and southern Chinese populations in the southwest—added regional diversity.

This process was not purely demographic but closely intertwined with cultural assimilation. The Longshan culture’s technological innovations (e.g., bronze metallurgy, urbanization) and its ideological systems (e.g., ancestor worship, statecraft) likely attracted or compelled local populations to adopt Han identities, even as genetic mixing occurred. The absence of Northeast Asian hunter-gatherer ancestry in historical Shandong populations, despite their presence in the Neolithic [16], suggests that cultural integration was accompanied by selective genetic replacement or dilution, favoring agriculturalist lineages.

During the Qin and Han periods, state power became increasingly centralized, while the Han ethnicity expanded outward from the Central Plains, exerting profound influence on surrounding populations. The Han Dynasty, standing as a pivotal formative period in the emergence of the Han Chinese, is considered to be one of the first golden ages in Chinese history. As a unified and cohesive empire that succeeded the short-lived Qin dynasty, Han China established itself as the center of the East Asian geopolitical order at the time, projecting its power and influence unto Asian neighbors. It was comparable with the contemporary Roman Empire in population size, geographical extent, and cultural reach [36–38]. These historical processes of demographic expansion and consolidation provide the context for understanding modern genetic patterns. When using ancestries in the lower Yellow River during the Han Dynasty as potential ancestral sources for modern Han Chinese in Shandong, we found that they could be fully modeled by a one-way admixture (100% fit), demonstrating a high degree of genetic homogeneity between ancient populations from the Han Dynasty and modern Han Chinese in Shandong.

Shandong’s geographic location at the eastern reaches of the Yellow River, bridging the Central Plain with northern and southern China, positioned it as a vital corridor for both population movements and cultural exchange. Situated along the coast, facing the Liaodong Peninsula, Korean Peninsula, and Japanese Archipelago, Shandong also connected westward to the Central Plain and southward to the Jiangnan region, making it a strategic transportation and cultural hub. Whether Shandong experienced additional genetic turnover through interactions with other regions over the subsequent two millennia, and how this shaped the genetic profile of modern Han populations, remains an important avenue for future investigation.

Implications for Chinese civilization studies

Yan Wenming’s (1987) “multi-petal flower” model, where the Central Plain served as the core integrating surrounding cultures, finds genomic corroboration in our study. While not the sole contributor, the Central Plain Longshan ancestry provided the foundational genetic and cultural framework, with regional “petals” (e.g., Shandong Longshan, southern Chinese) enriching the whole. This interplay of core-periphery interactions, driven by both peaceful exchange and demographic expansion, created the genetic and cultural homogeneity that defines the Han Chinese as a cohesive ethnic group, despite their vast geographical spread.

Future studies integrating ancient genomes from more regions (e.g., the Yangtze River basin and Lingnan) and periods (e.g., the Tang, Song dynasties) will further refine our understanding of how historical events like the “Disaster of Yongjia” (Sixteen Kingdoms period) or “The Hu-Guang to Sichuan Migration” (Ming-Qing migrations) shaped the Han genetic landscape. Additionally, functional genomic analyses of Longshan-related populations could uncover adaptive traits associated with millet agriculture, disease resistance, or environmental adaptation, shedding light on the biological consequences of demographic expansion.

Conclusions

By analyzing 28 ancient genomes during the Han Dynasty, sourced from two archaeological sites in present-day Zibo, Shandong Province, in conjunction with previously published historical populations, this study explores the genetic contributions of Longshan culture-related ancestries to the formation of Han Chinese populations across different regions of China. Our results demonstrate that the emergence of the Han Chinese was a dynamic process of admixture, driven in most part by demic diffusion from the Central Plain Longshan populations and further shaped by regional interactions. The widespread prevalence of Central Plain Longshan-related ancestry across diverse regions underscores the pivotal role of the Yellow River basin as both a genetic and cultural hearth, while local admixture events highlight the complexity of ethnic formation. Moreover, during the formative stage of Han Chinese identity, ancient Han dynasty populations from the lower Yellow River region already exerted a profound genetic influence on present-day Shandong Han Chinese. These findings integrate archaeological narratives with genomic evidence, illustrating how population movements contributed to the emergence of one of humanity’s largest ethnic groups.

Methods

Archaeological information

Xujiacundong Site

The Xujiacundong Site is located at No. 51 Linzi Avenue, Linzi District, Zibo City, Shandong Province. The site lies approximately 4 km south of the southern city wall of the capital of the ancient Qi State (齐国). About 1.5 km to the north of the site is the Nanmafang Warring States Tomb Site (南马坊战国墓遗址), while 4.5 km to the east sits the Taigong Lake (太公湖). This site constitutes an important Han Dynasty family cemetery, the presence of which reflects the stability of social structures and the settled nature of communities in Han Dynasty Shandong. A total of 387 tombs were uncovered at this site, dating from the Warring States period to the Han Dynasty. The predominant tomb type was the vertical earthen pit grave, and the burial goods in most tombs were relatively simple and of limited amounts. The cemetery primarily consisted of commoner burials, with some belonging to mid-to-lower-ranking officials. Based on excavated seals and soul jars, the tomb occupants included individuals with surnames such as “Feng Da” (冯达), “Dongguo” (东郭), “Chunyu” (淳于), and “Wang” (王). Apart from “Feng Da”, the other three surnames were prominent clans of the Qi State. Among them, “Dongguo” and “Chunyu” were of the Jiang lineage (姜姓), with Dongguo being descendants of the Qi ruling family—the offspring of Duke Huan of Qi (齐桓公).

Zhouhe Site

The Zhouhe Site is situated northeast of the intersection of Linzi Qidu Road and Linzi Avenue. The two sites are approximately 2 km apart. Excavations here uncovered 140 Han Dynasty tombs, yielding over 50 intact or restorable pottery items, primarily including jars, pots, basins, and a small number of bronze swords, mirrors, belt hooks, and coins. The tombs can be categorized into three types: vertical earthen pit graves, vertical earthen pit graves with stone accumulations, and brick-chambered tombs. The tombs were small in scale, densely distributed, and clearly laid out, with few burial goods, all of which belonged to commoners.

Ancient DNA sample processing

We collected 58 human skeletal remains from the Xujiacundong and Zhouhe site in Shandong, China. All samples were processed in the dedicated ancient DNA clean room at the Institute of Anthropology, Xiamen University, following established precautions for working with ancient human DNA [39–41]. The human skeletal remains were first cleaned with 75% ethanol, and the surface was cleaned using a drill bit. Next, the samples were washed with 10% NaClO and exposed to ultraviolet light for 30 min. We collected 80–120 mg of bone powder by drilling deep into the petrous part of the temporal bone. After adding 1 mL of 0.5 mM EDTA and 0.25 mg/mL Proteinase K to the bone powder, the mixture was agitated at 300 rpm and incubated overnight at 37 °C for lysis. After centrifugation, the precipitate was discarded, and 12.5 mL of binding buffer was added to the supernatant. The binding buffer contains 5 M guanidine hydrochloride, 40% isopropanol, 25 mM sodium acetate, and 0.05% Tween-20 (Sigma Aldrich, Germany) at pH 5.5. Then, we purified samples using the MinElute kit (Qiagen, Germany) and eluted DNA extract by 0.1*TE [42]. We used the NEBNext Ultra II DNA Library Prep Kit with an adaptor from blunt-ended ligation-based approaches [43] instead of a circular NEBNext Adaptor to prepare double-stranded libraries. Next, we purified the DNA library with the AMPure XP beads (Beckman Coulter, USA) and employed the conventional agarose gel electrophoresis method to inspect the library strips. Finally, sequencing was performed on the DNBSEQ-T7 platform. All libraries were sequenced using the DNBSEQ-T7 platform to produce pair-end reads (2 × 150 bp for nuclear DNA).

Radiocarbon dating

The preparation of bone samples involves standard acid–base-acid (ABA) procedure and collagen extraction. Firstly, bone samples are ultrasonicated in ultrapure water, dried, grinded and sieved to get the appropriately sized sample fraction (0.5–1 mm). Then, the samples are treated with 0.5 M hydrochloric acid (~ 18 h), 0.1–0.2 M sodium hydroxide (30 min−1 h), and 0.5 M hydrochloric acid (1 h). Bone collagen gelatinization is performed in a pH 3 solution at 70 °C for 20 h. The gelatine solution is filtered using a cleaned filter and freeze-dried. The quality of collagen is monitored by measuring the carbon and nitrogen content in collagen, determining the atomic C/N ratio, and assessing the collagen yield. Samples are dated if the collagen yield is above 1% and the C:N ratio of the collagen is between 2.9 and 3.5. The samples that deviate from these ratios are deemed unsuitable for dating [44]. The pure collagen samples were combusted with CuO and silver in vacuo to CO2 at 900 °C for 2 h. The CO2 was purified, graphitized and measured on a compact 0.5 MeV NEC Accelerator Mass Spectrometry in the Guangzhou Institute of Geochemistry, Chinese Academy of Sciences [45]. All three radiocarbon dates were carried out using OxCal v4.4.4 (https://c14.arch.ox.ac.uk/oxcal/OxCal.html), applying the latest consensus calibration curve, IntCal20 [46].

Sequence data processing

We first trimmed the sequencing adapter and merged the paired-end reads into one sequence, employing AdapterRemoval v2.3.2 [47]. We also used FastQC v0.11.9 to do quality control checks on the adapter-removed sequencing data. We then mapped the merged reads onto the human reference genome hs37d5 (GRCh37 with decoy sequences) using BWA v0.7.17 [48] samse, with parameters -l 1024 and -n 0.01. Dedup v0.12.8[49] was used to remove the PCR duplicates. Moreover, we filtered the sequences using the view implemented in SAMtools v1.7 [50] with the parameters -q 30 and -Fox 4. Low mapping quality reads were removed in this step. We checked whether each library showed the increased C-to-T misincorporations at 5′ end and G-to-A misincorporations at 3′ end and expected for double-stranded libraries. To minimize the bias due to ancient DNA deamination, we further clipped six bases from both ends of each read using trimBam implemented in BamUtil v1.0.15, ensuring the deamination rate remains below 5% (Additional file 1: Table S1B) [51]. We then used mpileup implemented in SAMtools to filter the alignment quality with the parameters -q30 and -Q30. After the quality filtering procedure, the number of reads for some samples containing many low-quality bases would be reduced. Finally, we generated pseudohaploid calls for each sample employing the pileupCaller (https://github.com/stschiff/sequenceTools) software with the parameter -RandomHaploid, using the 1240 K dataset as a reference.

Authentication of ancient DNA

We used PMDtools v0.50 [52] to quantify DNA damage patterns among ancient DNA and estimate misincorporation rates (5′C > T and 3′G > A). Then, we used Schmutzi v1.5.5.5 [53] to estimate the mtDNA contamination rate. For the assigned male sample, we used ANGSD v1.2.11 [54] to estimate X-chromosomal contamination [55]. We also conducted the ContamLD analysis which was used to estimate nuclear contamination for all individuals [56]. The warning “Model_Misspecified” usually means the coverage is very low and the estimate might not be reliable, which occurred in all of our individuals. Therefore, the modern human contamination estimation was primarily based on Schmutzi and ANGSD outputs.

Sex determination and uniparental haplogroup assignment

We used three methods to assign the genetic gender to our samples. The first relies on the ratio of alignments to chromosome X compared to the ratio of alignments of all autosomes (Rx) [57]. The second computes the ratio of the total number of sequences aligned to either sex chromosome (RY) [58]. The third compares the genome coverage of the X- and Y-chromosomes with that of autosomes [59]. We combined these methods to produce a more reliable result.

We used the log2fasta program implemented in Schmutzi [53] to call the mtDNA consensus sequences. Then, mitochondrial haplogroups were assigned using Haplogrep2 [60]. Y-chromosome haplogroups were examined by aligning a set of positions in the ISOGG (International Society of Genetic Genealogy, http://isogg.org/, July 2019), in which we only restrict our analysis to reads with base and mapping quality higher than 30. Haplogroup determination was performed with Yleaf v2.2[61] software, which provide outputs for allele counts of ancestral and derived SNPs along a path of branches of the Y-chromosome tree.

Data merging

We merged our data with published modern and ancient populations from the Human Origins data set (HO) and the 1240K dataset included in the Allen Ancient DNA Resource (AADR) [62], and other published ancient individuals using mergeit from EIGENSOFT. The “1240K” dataset is based on the Illumina array 1240K dataset, which includes 1,233,013 SNPs. The “Human Origin” dataset, based on the Affymetrix Human Origins array, which consists of 597,573 SNPs, contains various modern populations. To see more information about ancient and present-day populations used in this study in Additional file 1: Table S1C and Additional file 1: Table S1D. We used the “merged HO” dataset in modern Chinese analyses involving modern populations, like smartpca and ADMIXTURE. We also used the “merged 1240K” dataset in analyses for ancient individuals, like outgroup-f3, f4, qpWave, and qpAdm analyses [63].

Principal components analysis

We performed principal component analysis (PCA) based on the “merged HO” dataset using smartpca v16000 in the EIGENSOFT package. The PCA analysis was performed at the individual level to characterize the genetic structure of all samples. The option “lsqproject: YES” was set to project ancient individuals to the calculated principal components based on modern populations. The ggplot2 package in the R software (http://www.r-project.org/) was used for visualization.

ADMIXTURE analysis

We ran ADMIXTURE v1.3.0 [64] to perform an unsupervised admixture analysis on both present-day and ancient populations after pruning for linkage disequilibrium in plink v1.90 [65] using the parameters –indep-pairwise 200 25 0.4 [66, 67]. We selected 94 ancient and present-day populations comprising 665 samples for ADMIXTURE analysis. We retained 258,434 SNPs to perform the analysis after pruning for linkage disequilibrium.

f-statistics

qp3Pop v651 was implemented to calculate the outgroup-f3 analysis using the “inbreed: YES” parameter. qpDstat v980 was performed to calculate the f4 statistics using the parameter “f4mode: YES” [63, 66]. We calculated outgroup-f3 statistics in the form of f3(X, Y; Mbuti) to measure the shared genetic drift between X and Y, where Mbuti is a distant African population and acts as an outgroup, and we excluded populations with fewer than 30,000 SNPs used for calculation. Besides, we calculated f4 statistics in f4(X, Y; Z, Mbuti) to explore the additional gene flow from Z into X or Y, where Mbuti acts as an outgroup. Both qp3Pop and qpDstat programs are implemented in the ADMIXTOOLS package [63].

qpWave analysis

We performed qpWave v410 as implemented in ADMIXTOOLS parameters “allsnps: YES” and “inbreed: YES” on the “merged 1240 K” dataset to detect the minimum number of independent gene pools to explain a set of target populations. We also utilized qpWave to identify outlier individuals in the studied populations.

We initially included all individuals from the Xujiacundong site in the “left pop”. When rank = 0, a tail value < 0.05 suggested potential outlier presence. Through iterative screening, we excluded individuals XJCDM354, XJCDM110, XJCDM15, XJCDM353, and XJCDM85, after which the remaining individuals showed homogeneity (tail > 0.05). The remaining individuals were treated as a provisional population (designated as XJCDM_tmp), and we conducted pairwise qpWave analysis with XJCDM_tmp and the five excluded individuals. Only XJCDM354 exhibited a tail value < 0.05. Consequently, we reintegrated XJCDM110, XJCDM15, XJCDM353, and XJCDM85 into XJCDM_tmp. Final pairwise qpWave analysis between XJCDM_tmp and XJCDM354 yielded a tail value of 0.0027, conclusively identifying XJCDM354 as an outlier. The same methodology applied to the Zhouhe site revealed no statistically significant outliers.

Admixture modelling

We used qpAdm v810 implemented in ADMIXTOOLS using parameters “allsnps: YES and inbreed: YES” with the “merged 1240 K” dataset to estimate the admixture proportion of a target population as the combination of the source populations [68, 69]. We used the rotating strategy [69]. We used a set of outgroups as fixed outgroups and added the population to the outgroup when a source population was not chosen as a potential source.

Supplementary Information

12915_2026_2541_MOESM1_ESM.zip (276.3KB, zip)

Additional file 1: Table S1A. A summary of 58 ancient samples in this study (the individuals for the downstream analysis are highlighted in yellow). Table S1B. (a) C-to-T misincorporations at the 5’ end of 58 ancient individuals. (b) G-to-A misincorporations at the 3’ end of 58 ancient individuals. Table S1C. Published present-day genomes co-analyzed in this study. Table S1D. Published ancient genomes co-analyzed in this study. Table S2A. The results of f4(Mbuti, X; YR_LS, SD_LS) using “merged 1240K” dataset, in which X denotes populations around East Asia. Table S2B. The results of outgroup-f3 (Target, X; Mbuti) using the “merged 1240K” dataset, in which X denotes the ancient populations in East Asia. Table S2C. The results of f4(Mbuti, middle YR-related; XJCDM_Han_Dynasty/ZHM_Han_Dynasty, SD_LS) using “merged 1240K” dataset. Table S2D. The results of f4(Mbuti, X; XJCDM_Han_Dynasty/ZHM_Han_Dynasty, YR_LN) using the “merged 1240K” dataset, in which X denotes populations around East Asia. Table S2E. The results of f4(Mbuti, X; XJCDM_Han_Dynasty_o1, YR_LN) using the “merged 1240K” dataset, in which X denotes ancient populations around East Asia. Table S2F. The results of f4(Mbuti, X; Studied, YR_LN) using the “merged 1240K” dataset, X is related to populations outside the Yellow River Basin. Table S2G. qpAdm models to estimate the genetic composition of some post-Longshan Era populations in different regions (the results in Fig. 3 are highlighted in yellow). Table S2H. qpAdm models to estimate the genetic composition of present-day Han Chinese in the Shandong Province.

12915_2026_2541_MOESM2_ESM.zip (6.6MB, zip)

Additional file 2: Fig. S1. Ancient DNA damage patterns for 28 newly sequenced ancient individuals from Linzi City, Shandong Province during the Han Dynasty. Curves on the left side and right side represented C->T misincorporations at the 5’ end and G->A misincorporations at the 3’ end, respectively. Fig. S2. Kinship analysis by READv2. We detected five pairs of individuals within third-degree relatives and excluded one individual from each pair. Fig. S3. Radiocarbon dating results of 3 samples: XJCDM110, XJCDM354, and ZHM3. Fig. S4. Principal component analysis (PCA) of ancient and present-day East Asians. Different colors and shapes represent different populations. Present-day individuals were color-coded based on the language family they belonged to. Fig. S5. The results of ADMIXTURE analysis. A. Cross-validation (CV) error in ADMIXTURE analysis. The lowest CV errors were observed for K=3. B. ADMIXTURE analysis with “merged HO” dataset among multiple ancient/present-day populations in East Asia. C. The ADMIXTURE result with the “merged HO” dataset among multiple ancient/present-day populations from K=2 to K=8.

Acknowledgements

We thank the editors and reviewers for their contributions and suggestions. We thank all participants in these studies. SF and ZX from the Information and Network Center of Xiamen University are acknowledged for their help with high-performance computing.

Abbreviations

XJCDM

Xujiacundong site

ZHM

Zhouhe site

SNPs

Single nucleotide polymorphisms

HO

Human Origin

PCA

Principal components analysis

SD_LS

Shandong Longshan populations

YR_LS

Central Plain Longshan populations in the middle reaches of Yellow River

sEA

Southern East Asian

EA

East Asian

YR

Yellow River

WLR

West Liao River

ANA

Ancient Northeast Asian

EN

Early Neolithic

MN

Middle Neolithic

LN

Late Neolithic

BA

Bronze Age

IA

Iron Age

LBIA

Late Bronze and Iron Ages

H

History Era

BCE

Before Common Era

CE

Common Era

Authors’ contributions

C.C.W., X.W and D.H. conceived and supervised the project. H.C. and C.C.Q. provided the samples. Q.S., Y.Y.P., X.L.M., Y.L.L., B.T.W. and Z.J. H.M.Z., X.Y.Z., X.Y.W. and J.X.T. collected the samples from the archaeological sites. L.M.Q., M.T.X, H.M., L.T. and H.F.H. performed the majority of the experiments. H.D.C, Y.L.J, K.Y.Z, Y.X. performed the genetic data analysis and prepared the figures. H.D.C, L.M.Q and C.C.W. wrote and edited the manuscript. All authors read and approved the final manuscript.

Funding

The work was funded by the National Key Research and Development Program of China (2024YFC3306701, 2023YFC3303701-02), National Natural Science Foundation of China (T2425014 and 32270667), the Natural Science Foundation of Fujian Province of China (2023J06013), the Major Project of the National Social Science Foundation of China (21&ZD285), and the Open Research Fund of the Forensic Genetics Key Laboratory of the Ministry of Public Security (2023FGKFKT07).

Data availability

All data generated or analysed during this study are included in this published article, its supplementary information files and publicly available repositories. The alignment files (BAM format) reported in this paper have been deposited in the Genome Sequence Archive in the National Genomics Data Center, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA-Human: HRA015816), which is publicly accessible at https://ngdc.cncb.ac.cn/gsa-human/browse/HRA015816. Other data and code are available from the corresponding author on reasonable request. All modern and ancient reference populations used in this study are listed in Additional file 1: Table S1C and Additional file 1: Table S1D. Genotype data of ancient individuals in this study on the 1240 K panel are available in the EIGENSTRAT format from the following link: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/FFIDCW. Other ancient genomic datasets utilized in this study are available for download via the following links: https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA007869, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA008852, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA006574, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA004375, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA006067, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA005990.

Declarations

Ethics approval and consent to participate

The procedures related to human subjects were approved by the Medical Ethical Committee of Xiamen University (XDYX202412K88).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Limin Qiu, Haodong Chen, Hui Chen, and Chaochao Qin contributed equally to this work.

Contributor Information

Daohua Hao, Email: 51983964@qq.com.

Xiaokun Wang, Email: wangxiaokun@ruc.edu.cn.

Chuan-Chao Wang, Email: chuanchaowang@fudan.edu.cn.

References

  • 1.Wang Z. Main vein of the Yellow River culture (in Chinese). J Shandong Norm Univ (Social Sciences). 2022.
  • 2.Cheng Y. Why the Yellow River culture is considered the root and soul of the Chinese nation (in Chinese). Haidai Xuekan. 2023.
  • 3.Zhang C. Longshan culture, Longshan period and Longshan era: rereading Longshan culture and Longshan era (in Chinese). Cultural Relics in Southern China. 2021.
  • 4.Yang Z, Wang S. A brief discussion on the Longshan culture (in Chinese). Archaeology. 1963.
  • 5.Jin S, Zhao X. The historical significance of Shandong Lungshan culture and the reasons for its decline (in Chinese). J Zhengzhou Univ (Philosophy and Social Sciences Edition). 1994.
  • 6.Han J. Exploring the land formation process of the Tianjin Region (in Chinese). Proceedings of the First Annual Meeting of the Chinese Archaeological Society. 1979.
  • 7.Xu Q. Examining the geographic and topographic changes in northern Shandong from 7000 to 3000 years ago through the distribution of ancient cultural sites. Archaeology. 1992.
  • 8.Zhang W, Zhang Z, Fan W, Wang S. Preliminary report on the excavation of the Luantai Site in Luyi, Henan (in Chinese). Huaxia Archaeology. 1989.
  • 9.Zou H. The birth of Chinese civilization (in Chinese). Cultural Relics. 1987.
  • 10.Lan A, Kang K, Tang S, Wu X, Wang L, Li T, Weng H, Deng J, WeGene Research T, Zheng Q et al: Fine-scale population structure and demographic history of Han Chinese inferred from haplotype network of 111,000 genomes. bioRxiv. 2020:166413.
  • 11.Liu S, Huang S, Chen F, Zhao L, Yuan Y, Francis SS, et al. Genomic analyses from non-invasive prenatal testing reveal genetic associations, patterns of viral infections, and Chinese population history. Cell. 2018;175(2). [DOI] [PubMed] [Google Scholar]
  • 12.Cong P, Bai W, Li J, Yang M, Khederzadeh S, Gai S, et al. Genomic analyses of 10,376 individuals in the Westlake BioBank for Chinese (WBBC) pilot project. Nat Commun. 2022;23(1):2939. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wen B, Li H, Lu D, Song X, Zhang F, He Y, et al. Genetic evidence supports demic diffusion of Han culture. Nature. 2004;431(7006):302–5. [DOI] [PubMed] [Google Scholar]
  • 14.Shen Q, Wu Z, Zan J, Yang X, Guo J, Ji Z, Wang B, Liu Y, Mao X, Wang X et al: Ancient genomes illuminate the demographic history of Shandong over the past two millennia. Journal of Genet Genomics 2024. [DOI] [PubMed]
  • 15.Wang B, Hao D, Xu Y, Zhu K, Wang R, Yang X, et al. Population expansion from central plain to northern coastal China inferred from ancient human genomes. iScience. 2024;27(12):111405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Du P, Zhu K, Wang M, Sun Z, Tan J, Sun B, et al. Genomic dynamics of the Lower Yellow River Valley since the Early Neolithic. Curr Biol. 2024;34(17):3996-4006.e3911. [DOI] [PubMed] [Google Scholar]
  • 17.Xiong J, Wang R, Chen G, Yang Y, Du P, Meng H, et al. Inferring the demographic history of Hexi Corridor over the past two millennia from ancient genomes. Sci Bull (Beijing). 2024;69(5):606–11. [DOI] [PubMed] [Google Scholar]
  • 18.Ning C, Li T, Wang K, Zhang F, Li T, Wu X, et al. Ancient genomes from northern China suggest links between subsistence changes and human migration. Nat Commun. 2020;11(1):2700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Zhu K, Hu C, Yang M, Zhang X, Guo J, Xie M, et al. The demic diffusion of Han culture into the Yunnan-Guizhou plateau inferred from ancient genomes. Natl Sci Rev. 2024;11(12):nwae387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Tao L, Yuan H, Zhu K, Liu X, Guo J, Min R, He H, Cao D, Yang X, Zhou Z et al: Ancient genomes reveal millet farming-related demic diffusion from the Yellow River into southwest China. Curr Biol 2023, 33(22):4995–5002 e4997. [DOI] [PubMed]
  • 21.Wang F, Wang R, Ma H, Zeng W, Zhao Y, Wu H, et al. Neolithization of Dawenkou culture in the lower Yellow River involved the demic diffusion from the Central Plain. Sci Bull (Beijing). 2024;69(23):3677–81. [DOI] [PubMed] [Google Scholar]
  • 22.Ma H, Zhou Y, Wang R, Yan F, Chen H, Qiu L, et al. Ancient genomes shed light on the long-term genetic stability in the Central Plain of China. Sci Bull (Beijing). 2025;70(3):333–7. [DOI] [PubMed] [Google Scholar]
  • 23.Zheng H, Yan S, Qin Z, Wang Y, Tan J, Li H, et al. Major population expansion of East Asians began before Neolithic time: evidence of mtDNA genomes. PLoS ONE. 2011;6(10):e25835. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Umetsu K, Tanaka M, Yuasa I, Adachi N, Miyoshi A, Kashimura S, et al. Multiplex amplified product-length polymorphism analysis of 36 mitochondrial single-nucleotide polymorphisms for haplogrouping of East Asian populations. Electrophoresis. 2005;26(1):91–8. [DOI] [PubMed] [Google Scholar]
  • 25.Davila A: Mitochondrial DNA variant associated with Leber hereditary optic neuropathy and high-altitude Tibetans Proceedings of the National Academy of Sciences 2012. [DOI] [PMC free article] [PubMed]
  • 26.Pimenoff VN, Comas D, Palo JU, Vershubsky G, Kozlov A, Sajantila A. Northwest Siberian Khanty and Mansi in the junction of West and East Eurasian gene pools as revealed by uniparental markers. Eur J Hum Genet. 2008;16(10):1254–64. [DOI] [PubMed] [Google Scholar]
  • 27.Yao YG, Kong QP, Bandelt HJ, Kivisild T, Zhang YP. Phylogeographic differentiation of mitochondrial DNA in Han Chinese. Am J Hum Genet. 2002;70(3):635–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Yan S, Wang C, Zheng H, Wang W, Qin Z, Wei L, Wang Y, Pan X, Fu W, He Y et al: Y Chromosomes of 40% Chinese are descendants of three Neolithic super-grandfathers. PLoS ONE. 2014;9(8). [DOI] [PMC free article] [PubMed]
  • 29.Shi H, Qi X, Zhong H, Peng Y, Zhang X, et al. Genetic evidence of an East Asian origin and Paleolithic northward migration of Y-chromosome haplogroup N. PLoS ONE. 2013;8(6):e66102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zhao Y, Zhang Y, Zhang Q, Li H, Cui Y, Xu Z, et al. Ancient DNA reveals that the genetic structure of the northern Han Chinese was shaped prior to 3,000 years ago. PLoS ONE. 2015;10(5):e0125676. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Malyarchuk B, Derenko M, Denisova G, Wozniak M, Grzybowski T, Dambueva I, et al. Phylogeography of the Y-chromosome haplogroup C in northern Eurasia. Ann Hum Genet. 2010;74(6):539–46. [DOI] [PubMed] [Google Scholar]
  • 32.Li S, Wang R, Ma H, Tu Z, Qiu L, Chen H, et al. Ancient genomic time transect unravels the population dynamics of Neolithic middle Yellow River farmers. Sci Bull. 2024;69(21):3365–70. [DOI] [PubMed] [Google Scholar]
  • 33.Fan Y. The position and action of the Central-Plains culture in the course of the form and development of Chinese civilization (in Chinese). J Zhengzhou Univ (Philosophy and Social Sciences Edition). 2006.
  • 34.Liu X. Han. In: Ember, C.R., Ember, M. (eds) Encyclopedia of Medical Anthropology. Springer, Boston, MA. 2004:703–17.
  • 35.Yang MA, Fan X, Sun B, Chen C, Lang J, Ko YC, et al. Ancient DNA indicates human population shifts and admixture in northern and southern China. Science. 2020;369(6501):282–8. [DOI] [PubMed] [Google Scholar]
  • 36.Kang DC. East Asia before the West: five centuries of trade and tribute. Columbia University Press; 2010. [Google Scholar]
  • 37.Minahan J: Ethnic groups of North, East, and Central Asia: an encyclopedia; 2014.
  • 38.Cohen WI. East Asia at the center four thousand years of engagement with the world. Columbia University Press; 2000. [Google Scholar]
  • 39.Shapiro B, Hofreiter M. A paleogenomic perspective on evolution and gene function: new insights from ancient DNA. Science. 2014;343(6169). [DOI] [PubMed] [Google Scholar]
  • 40.Knapp M, Clarke AC, Horsburgh KA, Matisoo-Smith EA. Setting the stage - building and working in an ancient DNA laboratory. Ann Anat - Anat Anz. 2012;194(1):3–6. [DOI] [PubMed] [Google Scholar]
  • 41.Zhu K, He H, Tao L, Ma H, Yang X, Wang R, et al. Protocol for a comprehensive pipeline to study ancient human genomes. STAR Protoc. 2024;5(2):102985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Dehasque M, Pecnerova P, Kempe Lagerholm V, Ersmark E, Danilov GK, Mortensen P, Vartanyan S, Dalen L: Development and optimization of a silica column-based extraction protocol for ancient DNA. Genes (Basel) 2022, 13(4). [DOI] [PMC free article] [PubMed]
  • 43.Meyer M, Kircher M. Illumina sequencing library preparation for highly multiplexed target capture and sequencing. Cold Spring Harb Protoc. 2010;2010(6):pdb prot5448. [DOI] [PubMed] [Google Scholar]
  • 44.van Klinken GJ. Bone collagen quality indicators for palaeodietary and radiocarbon measurements. J Archaeol Sci. 1999;26(6):687–95. [Google Scholar]
  • 45.Zhu S, Ding P, Wang N, Shen C, Jia G, Zhang G. The compact AMS facility at Guangzhou Institute of Geochemistry, Chinese Academy of Sciences. Nucl Instrum Methods Phys Res, Sect B. 2015;361:72–5. [Google Scholar]
  • 46.Reimer PJ, Austin WEN, Bard E, Bayliss A, Blackwell PG, Ramsey CB, et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0-55 cal kBP). Radiocarbon. 2020;62(4):725–57. [Google Scholar]
  • 47.Schubert M, Lindgreen S, Orlando L. AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res Notes. 2016;9:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):1754–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Peltzer A, Jager G, Herbig A, Seitz A, Kniep C, Krause J, et al. EAGER: efficient ancient genome reconstruction. Genome Biol. 2016;17:60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Danecek P, Bonfield JK, Liddle J, Marshall J, Ohan V, Pollard MO, et al. Twelve years of SAMtools and BCFtools. Gigascience. 2021;10(2):1–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Jun G, Wing MK, Abecasis GR, Kang HM. An efficient and scalable analysis framework for variant extraction and refinement from population-scale DNA sequence data. Genome Res. 2015;25(6):918–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Skoglund P, Northoff BH, Shunkov MV, Derevianko AP, Paabo S, Krause J, et al. Separating endogenous ancient DNA from modern day contamination in a Siberian Neandertal. Proc Natl Acad Sci U S A. 2014;111(6):2229–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Renaud G, Slon V, Duggan AT, Kelso J. Schmutzi: estimation of contamination and endogenous mitochondrial consensus calling for ancient DNA. Genome Biol. 2015;16:224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Korneliussen TS, Albrechtsen A, Nielsen R. ANGSD: analysis of next generation sequencing data. BMC Bioinformatics. 2014;15(1):356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Moreno-Mayar JV, Korneliussen TS, Dalal J, Renaud G, Albrechtsen A, Nielsen R, et al. A likelihood method for estimating present-day human contamination in ancient male samples using low-depth X-chromosome data. Bioinformatics. 2020;36(3):828–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Nakatsuka N, Harney E, Mallick S, Mah M, Patterson N, Reich D. ContamLD: estimation of ancient nuclear DNA contamination using breakdown of linkage disequilibrium. Genome Biol. 2020;21(1):199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Mittnik A, Wang CC, Svoboda J, Krause J. A molecular approach to the sexing of the triple burial at the upper Paleolithic site of Dolni Vestonice. PLoS ONE. 2016;11(10):e0163019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Skoglund P, Storå J, Götherström A, Jakobsson M. Accurate sex identification of ancient human remains using DNA shotgun sequencing. J Archaeol Sci. 2013;40(12):4477–82. [Google Scholar]
  • 59.Fu Q, Posth C, Hajdinjak M, Petr M, Mallick S, Fernandes D, et al. The genetic history of Ice Age Europe. Nature. 2016;534(7606):200–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Weissensteiner H, Pacher D, Kloss-Brandstatter A, Forer L, Specht G, Bandelt HJ, et al. HaploGrep 2: mitochondrial haplogroup classification in the era of high-throughput sequencing. Nucleic Acids Res. 2016;44(W1):W58-63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Ralf A, Montiel González D, Zhong K, Kayser M, Su B. Yleaf: software for human Y-chromosomal haplogroup inference from next-generation sequencing data. Mol Biol Evol. 2018;35(5):1291–4. [DOI] [PubMed] [Google Scholar]
  • 62.Mallick S, Micco A, Mah M, Ringbauer H, Lazaridis I, Olalde I, et al. The Allen Ancient DNA Resource (AADR) a curated compendium of ancient human genomes. Sci Data. 2024;11(1):182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Patterson N, Moorjani P, Luo Y, Mallick S, Rohland N, Zhan Y, et al. Ancient admixture in human history. Genetics. 2012;192(3):1065–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 2009;19(9):1655–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, Lee JJ. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience. 2015;4(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Peter BM. Admixture, population structure, and F-statistics. Genetics. 2016;202(4):1485–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Lawson DJ, van Dorp L, Falush D. A tutorial on how not to over-interpret STRUCTURE and ADMIXTURE bar plots. Nat Commun. 2018;9(1):3258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Haak W, Lazaridis I, Patterson N, Rohland N, Mallick S, Llamas B, et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature. 2015;522(7555):207–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Harney E, Patterson N, Reich D, Wakeley J. Assessing the performance of qpAdm: a statistical tool for studying population admixture. Genetics. 2021;217(4). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

12915_2026_2541_MOESM1_ESM.zip (276.3KB, zip)

Additional file 1: Table S1A. A summary of 58 ancient samples in this study (the individuals for the downstream analysis are highlighted in yellow). Table S1B. (a) C-to-T misincorporations at the 5’ end of 58 ancient individuals. (b) G-to-A misincorporations at the 3’ end of 58 ancient individuals. Table S1C. Published present-day genomes co-analyzed in this study. Table S1D. Published ancient genomes co-analyzed in this study. Table S2A. The results of f4(Mbuti, X; YR_LS, SD_LS) using “merged 1240K” dataset, in which X denotes populations around East Asia. Table S2B. The results of outgroup-f3 (Target, X; Mbuti) using the “merged 1240K” dataset, in which X denotes the ancient populations in East Asia. Table S2C. The results of f4(Mbuti, middle YR-related; XJCDM_Han_Dynasty/ZHM_Han_Dynasty, SD_LS) using “merged 1240K” dataset. Table S2D. The results of f4(Mbuti, X; XJCDM_Han_Dynasty/ZHM_Han_Dynasty, YR_LN) using the “merged 1240K” dataset, in which X denotes populations around East Asia. Table S2E. The results of f4(Mbuti, X; XJCDM_Han_Dynasty_o1, YR_LN) using the “merged 1240K” dataset, in which X denotes ancient populations around East Asia. Table S2F. The results of f4(Mbuti, X; Studied, YR_LN) using the “merged 1240K” dataset, X is related to populations outside the Yellow River Basin. Table S2G. qpAdm models to estimate the genetic composition of some post-Longshan Era populations in different regions (the results in Fig. 3 are highlighted in yellow). Table S2H. qpAdm models to estimate the genetic composition of present-day Han Chinese in the Shandong Province.

12915_2026_2541_MOESM2_ESM.zip (6.6MB, zip)

Additional file 2: Fig. S1. Ancient DNA damage patterns for 28 newly sequenced ancient individuals from Linzi City, Shandong Province during the Han Dynasty. Curves on the left side and right side represented C->T misincorporations at the 5’ end and G->A misincorporations at the 3’ end, respectively. Fig. S2. Kinship analysis by READv2. We detected five pairs of individuals within third-degree relatives and excluded one individual from each pair. Fig. S3. Radiocarbon dating results of 3 samples: XJCDM110, XJCDM354, and ZHM3. Fig. S4. Principal component analysis (PCA) of ancient and present-day East Asians. Different colors and shapes represent different populations. Present-day individuals were color-coded based on the language family they belonged to. Fig. S5. The results of ADMIXTURE analysis. A. Cross-validation (CV) error in ADMIXTURE analysis. The lowest CV errors were observed for K=3. B. ADMIXTURE analysis with “merged HO” dataset among multiple ancient/present-day populations in East Asia. C. The ADMIXTURE result with the “merged HO” dataset among multiple ancient/present-day populations from K=2 to K=8.

Data Availability Statement

All data generated or analysed during this study are included in this published article, its supplementary information files and publicly available repositories. The alignment files (BAM format) reported in this paper have been deposited in the Genome Sequence Archive in the National Genomics Data Center, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA-Human: HRA015816), which is publicly accessible at https://ngdc.cncb.ac.cn/gsa-human/browse/HRA015816. Other data and code are available from the corresponding author on reasonable request. All modern and ancient reference populations used in this study are listed in Additional file 1: Table S1C and Additional file 1: Table S1D. Genotype data of ancient individuals in this study on the 1240 K panel are available in the EIGENSTRAT format from the following link: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/FFIDCW. Other ancient genomic datasets utilized in this study are available for download via the following links: https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA007869, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA008852, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA006574, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA004375, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA006067, https://ngdc.cncb.ac.cn/search/specific?db=hra&q=HRA005990.

Articles from BMC Biology are provided here courtesy of BMC

RESOURCES