Unveiling the Genetic History of the Maniq, a Primary Hunter-Gatherer Society

Keywords: Southeast Asia, Semang, Negritos, population genetics, genetic ancestry

Genetic Affiliation of the Maniq

To determine the population genetic relationships of the Maniq, we performed a PCA on populations of Africa, Europe, East Asia, and Oceania (fig. 1B). The Maniq individuals form a cluster close to the Malay Semang, on a cline between the Andamanese and East Asian population clusters, indicating an admixture between Andamanese-related and East Asian-related ancestral sources. We then performed a PCA restricted to Asian and Oceanian populations (fig. 1C). Eigenvector 1 spans from East Asian to Australasian populations, whereas the second Eigenvector is defined by the Oceanian versus Maniq populations. The Maniq forms a separate cluster at the polar opposite of East Asians, where Malay Semang and non-Semang populations lie in between in a cline.

We performed a clustering approach implemented in ADMIXTURE to identify the population structure of Maniq in relation to other Asian and Oceanian populations (fig. 2). The earliest K (K = 2) is defined by the East Asian-related versus Australasian-related (or Negrito-Australopapuan) clusters, where Maniq appears to be largely affiliated with the latter. Notably, as early as K = 3, Maniq already forms its own distinct cluster, separating them from Onge, Jarawa, Papuan, and Bougainville Islanders. The Maniq largely and consistently retained its own distinct cluster up to the optimal K = 7, the K with the lowest cross-validation error and high consistency of iterated runs (50 consistent replicates out of 50 runs). The lack of admixture among the Maniq in this structure-based analysis may be attributed to the high degree of genetic drift as discussed below.

Maniq Exhibit High Degree of Genetic Differentiation

To determine the degree of genetic differentiation in Maniq in relation to other populations in the Asia-Pacific region, we estimated the in between population F_ST of our data set panel (supplementary figs. 1–3, Supplementary Material online). We find that the Maniq have experienced intense genetic drift, given that they exhibit very high genetic distance based on pairwise F_ST when compared with any other reference population. The degree of genetic drift in Maniq is even higher than that of Andamanese or Philippine Negritos, who were previously shown as drifted populations (Mondal et al. 2016; Larena, Sanchez-Quinto, et al. 2021). Moreover, the degree of genetic drift is comparable to that of Surui of Brazil, even higher than the Mangyan Buhid of Philippines (supplementary fig. 3, Supplementary Material online), all of whom were previously demonstrated to have one of the highest levels of genetic drift relative to a worldwide set of populations (Mallick et al. 2016; Larena, Sanchez-Quinto, et al. 2021). The evidence on the extreme genetic drift observed in Maniq is also supported by our analysis on the ROH. The length and the number of ROH can provide insights on a population’s demographic history. High number of long tract ROH can be indictive of a population bottleneck or largely inbred population of small size (Ceballos et al. 2018). Except for Africans, all other populations including the Maniq displayed high number of short segment ROH, which was argued as an evidence for a shared out-of-Africa bottleneck (Kirin et al. 2010). Interestingly, compared with all other populations, the Maniq consistently exhibit high numbers of ROH in each segment length category of more than 1 Mb (fig. 3B). Moreover, when plotting the total number versus the total length of ROH segments (fig. 3A), the Maniq exhibit high values for both, a pattern that was also observed in other highly drifted populations such as the Surui of the Americas and the Mangyan Buhid of the Philippines (supplementary fig. 4AandB, Supplementary Material online). Accordingly, these aforementioned populations as well as the Maniq possess similar population characteristics: small population size and historical evidence of endogamy (Brandt 1961; Mallick et al. 2016; Larena, Sanchez-Quinto, et al. 2021). Thus, it is not surprising to observe that these highly drifted populations also demonstrate to possess the highest values of inbreeding coefficient (supplementary table 2, Supplementary Material online).

Maniq Are One of the Hòabìnhian-Related Populations of MSEA

It was previously argued that the ancient hunter-gatherers of the Hòabìnhian cultural complex are regarded as the ancestors of present-day hunter-gatherers of MSEA (McColl et al. 2018). This was recently supported by an ancient DNA data, where among the ancient individuals, Onge and Jarawa were shown to be most genetically affiliated with an 8,000-year-old Hòabìnhian individual from Pha Faen, Laos and a 4,000-year-old Hòabìnhian individual from Gua Cha, Malaysia (McColl et al. 2018). Using the test f3(Mbuti; Maniq, Ancient), we also find the Maniq to share high levels of genetic drift with the Lao Hòabìnhian, besides the ∼8,000-year-old Liangdao 2 individual and the Malay and Lao Neolithic individuals (supplementary fig. 5 and supplementary table 3, Supplementary Material online), indicating a combination of Hòabìnhian-related and East Asian-related ancestries for the Maniq. We then formally tested the presence of Hòabìnhian-related ancestry in the Maniq, other Semang, and Andamanese populations with f4(Mbuti, Lao Hòabìnhian; Liangdao 2, X) and f4(Mbuti, Lao Hòabìnhian; Balangao, X), where we examine which X population share alleles with the Lao Hòabìnhian relative to a least admixed East Asian individual/population, Liangdao 2 or Cordilleran Balangao (Larena, Sanchez-Quinto, et al. 2021). As expected and consistent with previous observations (McColl et al. 2018), Onge and Jehai displayed significant levels of Hòabìnhian-related ancestry. Looking at the remaining Semang populations, the Maniq displayed also high levels of Hòabìnhian-related ancestry (fig. 4AandB).

Examining the genetic relationships of present-day populations, our PCA and Admixture analysis indicate that the Maniq are affiliated to the Semang populations of Malaysia. We investigated this further with an outgroup f3 statistics, using the test f3(Mbuti; Maniq, X), where X is any other population in the Asia-Pacific region. Evidently, Maniq shares the most drift with the Malay Semang, followed by the Senoi (supplementary fig. 6, Supplementary Material online), suggesting a shared history among these populations. These findings are also consistent when we test f4(Mbuti, Maniq; X, Malay Semang/Andamanese), where the Maniq consistently forms a clade with the Malay Semang, relative to any other reference Asia-Pacific population (supplementary fig. 7A–F, Supplementary Material online). Moreover, using the tests f4(Mbuti, Papuan; Balangao, X) and f4(Mbuti, Onge; Balangao, X), we find the Maniq and Malay Semang populations to share alleles with Papuans and Andamanese, highlighting the shared Basal Australasian ancestry and the deep historical relationships among these populations (fig. 4EandF). Last, we find the Maniq to share more alleles with Andamanese relative to any other Malay Semang population (supplementary fig. 8, Supplementary Material online).

Maniq Display Evidence for Recent Admixture with East Asians

East Asian gene flow was previously observed among Malay Semang of MSEA (Aghakhanian et al. 2015). Ancient East Asian gene flow, represented by the Liangdao 2 sample, is present in Malay Semang, but to a lower extent in Maniq (fig. 4C). To formally investigate present-day gene flow in the Maniq, we performed the test f4(Mbuti, Balangao; Onge, X), where we examine whether any X of Asia-Pacific populations have gene flow with East Asians, represented by Balangao Cordillerans (fig. 4D). We find that Maniq, together with Malay Semang, exhibit significant gene flow with East Asians. This is supported by the implementation of f-statistical estimations in qpAdm (Harney et al. 2021), where we find that the most plausible model for the ancestral source of Maniq is a combination of both Andamanese and East Asian-related ancestries, 65% and 35%, respectively (fig. 5 and supplementary table 5, Supplementary Material online). Furthermore, we estimated the date of admixture using a weighted LD statistic-based method, MALDER (Loh et al. 2013), and revealed that Maniq had an admixture event between East Asian- and Andamanese-related ancestries which is dated to 709 years ago (95% CI: 475–944 years, assumed generation time of 25 years; supplementary table 6, Supplementary Material online).

To gain understanding on the phylogenetic relationships between Maniq and other MSEA populations, we implemented TreeMix which utilizes allele frequency data together with a Gaussian approximation for genetic drift (Pickrell and Pritchard 2012). Our analyses consistently reveal the significant genetic drift observed in Maniq relative to other populations, as well as the presence of Andamanese gene flow into the common ancestral branch of Semang groups (fig. 6). Additionally, using qpGraph, we explicitly tested the possible models of admixture history for the Maniq (fig. 7). A model, where Maniq is exclusively a clade of Hòabìnhian or East Asian-related groups were rejected (worst fitting Z of 8.3743 and 16.0049, respectively (see supplementary figs. 9 and 10, Supplementary Material online). A model that fits the data is when Maniq forms a clade with Onge and Lao Hòabìnhian, and later received admixture with East Asian-related populations (worst fitting Z = 0.788), providing additional support for the dual ancestral source for the Maniq.

No Evidence for Elevated Levels of Denisovan Ancestry among Maniq

Using the direct f4-ratio estimation of Neanderthal ancestry: f4(Neanderthal(Altai), Chimp; X, Yoruba)/f4(Neanderthal(Altai), Chimp; Neanderthal(Vindija), Yoruba), we find that the Maniq to possess ∼1.9% Neanderthal ancestry, which is expected for a population outside of Africa, (supplementary table 7 and supplementary fig. 11, Supplementary Material online). Likewise, the results are also consistent when we use the test f4(Chimp, Neanderthal; Mbuti, X) (supplementary table 8, Supplementary Material online). Using f4(Chimp, Denisovan; European, X) and the f4-ratio test f4(Chimp, Neanderthal; Southern Han, X)/f4(Chimp, Neanderthal; Southern Han, Denisovan), we find no detectable signal of Denisovan ancestry among the Maniq, as well as other Malay Semang populations (Z scores <1 or even negative). This holds also true for the Malay non-Semang, Andamanese, Indonesian, Cordilleran, and mainland East Asian populations (supplementary tables 9 and 10, Supplementary Material online). This is in contrast to the high levels of Denisovan ancestry found among Philippine Negritos and Australopapuans (Jinam et al. 2017; Jacobs et al. 2019), indicating that the Denisovan introgression event among these groups likely occurred in Island Southeast Asia and the Oceania (Jacobs et al. 2019; Choin et al. 2021; Larena, Sanchez-Quinto, et al. 2021).

Ethical Considerations and Sample Collection

This study was approved by the Ethics Committee of the University of Vienna (Reference No. 00444) and the Khon Kaen University Ethics Committee for Human Research (Reference No. HE622223). All study participants self-identify as belonging to the Maniq and are at least 18 years old. Consent process was conducted through a series of consultations, where the aims and rationale of the study were thoroughly explained and discussed with the Maniq. Saliva samples were collected from each individual who provided an informed consent (n = 21), using the Oragene DNA (OG-500) collection kit by DNA Genotek Inc. (Ontario, Canada). DNA isolation and genome-wide array-based genotyping was carried out by the Life & Brain Genomic Center (Bonn, Germany). We used the Infinium Omni2.5Exome-8 BeadChip (Illumina Inc., San Diego, CA) array which includes about 2.6 million variants, encompassing SNPs on all chromosomes as well as from the mtDNA. This array is essentially a combination of the Infinium Omni 2.5-8 Kit and the Infinium Exome-24 BeadChip (both Illumina Inc., San Diego, CA). The array was analyzed on the iScan system (Illumina Inc.). The results of this study will be discussed together with the Maniq in a future fieldwork trip (a planned visit in 2021 had to be canceled due to the coronavirus disease 2019 pandemic). We are especially interested in their opinions on their long isolation and relationships with other Semang.

Quality Control and Data Preparation

The newly generated genotyped data were processed for quality check and filtering using the PLINK v1.9 software (Purcell et al. 2007). Duplicate samples, duplicated SNPs, and mitochondrial and sex chromosomal SNPs were removed. All SNPs with a high missing rate (–geno 0.001) as well as all SNPs which failed the Hardy-Weinberg Exact test (P = 10⁻⁶) were excluded. No sample had to be excluded because of high missing call rates (–mind 0.01). We used the R program version 4.0.3 (R Core Team 2013) to run different R packages. In order to only use unrelated samples for the analyses, we used the R package “SNPRelate” (Zheng et al. 2012). First, we pruned the data set to exclude markers which were in linkage disequilibrium (LD) (snpgdsLDpruning(x, ld.threshold = 0.5), using the default composite measure), then we used the Identity-by-Descent Estimation using a Maximum Likelihood Estimation approach (IBD-MLE) with the following parameters (snpgdsIBDMLE(x, snp.id=snpset_pruned_05, maf = 0.01, missing.rate = 0.05)). A kinship cutoff of 0.125 was applied to identify the related samples, where one of the pair of related samples with the higher ID number was removed, resulting in a base Maniq data set containing 11 unrelated individuals and 2,333,374 SNPs spanning across the 22 autosomes. The base data set was merged with the populations from the published 1000 Genomes Phase 3 (Auton et al. 2015), Simons Genome Diversity Project (Mallick et al. 2016), Human Genome Diversity Project (Bergström et al. 2020), Andamanese (Mondal et al. 2016), Malaysian (Teo et al. 2009; Aghakhanian et al. 2015), Indonesian (Pierron et al. 2014; Mörseburg et al. 2016; Kusuma et al. 2017), and Philippine (Larena, Sanchez-Quinto, et al. 2021) data sets to generate a panel with 320,032 SNPs. In addition, the base data set was also merged with the published ancient DNA data (Lipson et al. 2018; McColl et al. 2018; Larena, Sanchez-Quinto, et al. 2021) to generate panel with 67,094 SNPs for transversion-sites only. All exome SNPs were excluded following QC and merging with other available data sets.

Population Genetic Analysis

Runs of homozygosity (ROH) and inbreeding coefficient were estimated using the –homozyg and –het flags of PLINK v1.9, respectively. The number of ROH for each trach length category for each population was plotted to infer demographic history (Ceballos et al. 2018), where long tracts of ROH are characteristic of genetic drift caused by recent bottlenecks, founder events, or endogamy. Principal component analysis (PCA) and estimation of between population F_ST were performed using EIGENSOFT v7.1 (Patterson et al. 2006; Price et al. 2006). The heat maps based on the F_ST values were plotted using the R package “pheatmap” (Kolde and Kolde 2015). The accompanying dendrograms were achieved using the default hierarchical clustering method (pheatmap(…, clustering_method = “complete”)). The lsq project function of EIGENSOFT v7.1 was also utilized to perform PCA on present-day populations and project ancient individuals. Population structure was analyzed using the unsupervised clustering approach implemented in ADMIXTURE v1.3 (Alexander et al. 2009), and limited the number of individuals to a maximum of 10 per population (except for Maniq, where we used all 11 samples). Using the default settings, Admixture analyses were run for 50 iterations for each of the 12 clusters, and the common modes of replicates were ascertained using the CLUMPP program (Jakobsson and Rosenberg 2007). For each K cluster, the major mode was visualized using Pong v1.4 (Behr et al. 2016). To create potential phylogenies with gene flow events we used the TreeMix software (Pickrell and Pritchard 2012). SNPs were pruned beforehand with PLINK, and the trees were computed with Mbuti as outgroup and grouping together 200 SNPs to account for potential LD. Migration edges were added sequentially from 0 to 9. The R package of Admixtools (Patterson et al. 2012) ADMIXTOOLS 2 version 2.0.0 (https://github.com/uqrmaie1/admixtools, last accessed February 27, 2022), was utilized to model admixture in populations (qpGraph), to estimate admixture proportions (qpAdm), to estimate shared genetic drift between populations based on outgroup f3 statistics (qp3Pop), and to formally test presence of admixture in populations with f4 and f4-ratio statistics (qpDstat and qpF4Ratio, respectively). A Z score of >3 was set as a threshold for statistical significance. The default parameters set by ADMIXTOOLS 2 were used in all computations except for the qpGraph computation qpgraph(…, lsqmode = TRUE, return_fstats = TRUE). With these settings the analyses run should be equal to the common settings for the classic Admixtools: outpop (NULL), blgsize (0.05), diag (0.0001), lsqmode (YES), and allsnps (YES). For an initial assessment of shared drift, we calculated f3 statistics with the following ancient samples. For Africa, we used “baa001” (Schlebusch et al. 2017) and “Mota” (Llorente et al. 2015), for Europe we used “Bichon” (Jones et al. 2015), “Ust’-Ishim” (Fu et al. 2014, 00), “Loschbour” (Lazaridis et al. 2014), and in Southeast Asia we used “Th521,” “Ma555,” “La368,” “La364,” “Ma912” (McColl et al. 2018), and “Liangdao 2” (Larena, Sanchez-Quinto, et al. 2021). To construct admixture graph models, we incorporated into the model the African Mbuti as an outgroup, the ancient Lao Hòabìnhian (“La368”) individual (McColl et al. 2018) and present-day Onge individuals to represent Andamanese-related ancestry, and the ancient Liangdao 2 individual (Larena, Sanchez-Quinto, et al. 2021) and present-day Balangao Cordilleran individuals to represent East Asian-related ancestry. For the implementation of qpAdm, we utilized Balangao Cordilleran of the Philippines as the surrogate for the least admixed East Asian source and Onge as the surrogate for the least admixed Andamanese source, and utilized Mbuti, Yoruba, French, Iberian, Surui, and Karitiana (Bergström et al. 2020) as the “right” or outgroup populations. To estimate the date of East Asian and Hòabìnhian-related admixture among Maniq and other related populations, we implemented a weighted LD statistic-based method, MALDER (Loh et al. 2013). From the two main data sets (present-day and archaic) we derived subsets of populations for analysis data sets if needed. Wherever applicable the SNPs were pruned before the analysis with PLINK using –indep-pairwise 50 10 0.5. A special data set was created for the ROH analysis, to achieve maximum SNP overlap before pruning. The list of populations, number of SNPs, and number of samples used in the study is detailed in supplementary table 1AandB, Supplementary Material online.

evac021_Supplementary_Data

Data Availability

The data generated in this study will be made available to researchers after review and approval by the Data Access Committee led by the corresponding authors.

1. Aghakhanian F, et al. 2015. Unravelling the genetic history of Negritos and indigenous populations of Southeast Asia. Genome Biol Evol. 7(5):1206–1215. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Alexander DH, Novembre J, Lange K. 2009. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 19(9):1655–1664. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Auton A, et al. ; 1000 Genomes Project Consortium. 2015. A global reference for human genetic variation. Nature 526(7571):68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Barrows DP. 1910. The Negrito and allied types in the Philippines. Am Anthropol. 12(3):358–376. [Google Scholar]
5. Behr AA, Liu KZ, Liu-Fang G, Nakka P, Ramachandran S. 2016. Pong: fast analysis and visualization of latent clusters in population genetic data. Bioinformatics 32(18):2817–2823. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Bergström A, et al. 2020. Insights into human genetic variation and population history from 929 diverse genomes. Science 367(6484):eaay5012. [DOI] [PMC free article] [PubMed] [Google Scholar]
7. Brandt JH. 1961. The Negrito of peninsular Thailand. J Siam Soc. 49:123–158. [Google Scholar]
8. Carlhoff S, et al. 2021. Genome of a middle Holocene hunter-gatherer from Wallacea. Nature 596:543–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Ceballos FC, Joshi PK, Clark DW, Ramsay M, Wilson JF. 2018. Runs of homozygosity: windows into population history and trait architecture. Nat Rev Genet. 19(4):220–234. [DOI] [PubMed] [Google Scholar]
10. Choin J, et al. 2021. Genomic insights into population history and biological adaptation in Oceania. Nature 592(7855):583–589. [DOI] [PubMed] [Google Scholar]
11. Clark DW, et al. 2019. Associations of autozygosity with a broad range of human phenotypes. Nat Commun. 10(1):4957. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Eberhard DM, Simons GF, Fennig CD. 2019. Ethnologue: languages of Asia. Dallas (TX): Sil International. [Google Scholar]
13. Evans IHN. 1968. The Negritos of Malaya. Hove: Psychology Press. [Google Scholar]
14. Fu Q, et al. 2014. Genome sequence of a 45,000-year-old modern human from western Siberia. Nature 514(7523):445–449. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Harney É, Patterson N, Reich D, Wakeley J. 2021. Assessing the performance of qpAdm: a statistical tool for studying population admixture. Genetics 217(4):iyaa045. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. HUGO Pan-Asian SNP Consortium, et al. 2009. Mapping human genetic diversity in Asia. Science 326:1541–1545. [DOI] [PubMed] [Google Scholar]
17. Jacobs GS, et al. 2019. Multiple deeply divergent Denisovan ancestries in Papuans. Cell 177(4):1010–1021.e32. [DOI] [PubMed] [Google Scholar]
18. Jakobsson M, Rosenberg NA. 2007. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics 23(14):1801–1806. [DOI] [PubMed] [Google Scholar]
19. Jinam TA, et al. 2017. Discerning the origins of the Negritos, first Sundaland people: deep divergence and archaic admixture. Genome Biol Evol. 9(8):2013–2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Jones ER, et al. 2015. Upper Palaeolithic genomes reveal deep roots of modern Eurasians. Nat Commun. 6:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Kelly RL. 2013. The lifeways of hunter-gatherers: the foraging spectrum. Cambridge: Cambridge University Press. [Google Scholar]
22. Kirin M, et al. 2010. Genomic runs of homozygosity record population history and consanguinity. PLoS One 5(11):e13996. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Kolde R, Kolde MR. 2015. Package ‘pheatmap’. R Package 1:790. [Google Scholar]
24. Kricheff DA, Lukas H. 2015. Being Maniq: practice and identity in the forests of Southern Thailand. Hunt Gatherer Res. 1(2):139–155. [Google Scholar]
25. Kusuma P, et al. 2017. The last sea nomads of the Indonesian archipelago: genomic origins and dispersal. Eur J Hum Genet. 25(8):1004–1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Kutanan W, et al. 2018. Contrasting maternal and paternal genetic variation of hunter-gatherer groups in Thailand. Sci Rep. 8(1):1536. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Larena M, McKenna J, et al. 2021. Philippine Ayta possess the highest level of Denisovan ancestry in the world. Curr Biol. 31:4219–4230.e10. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Larena M, Sanchez-Quinto F, et al. 2021. Multiple migrations to the Philippines during the last 50,000 years. Proc Natl Acad Sci U S A. 118:e2026132118. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Lazaridis I, et al. 2014. Ancient human genomes suggest three ancestral populations for present-day Europeans. Nature 513(7518):409–413. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Lipson M, et al. 2018. Ancient genomes document multiple waves of migration in Southeast Asian prehistory. Genetics 361:92–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Llorente MG, et al. 2015. Ancient Ethiopian genome reveals extensive Eurasian admixture in Eastern Africa. Science 350(6262):820–822. [DOI] [PubMed] [Google Scholar]
32. Loh P-R, et al. 2013. Inferring admixture histories of human populations using linkage disequilibrium. Genetics 193(4):1233–1254. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Lukas H. 2004. Can “They” save “Us”, the Foragers? Indonesian and Thai Hunter-Gatherer Cultures under Threat from Outside. Südostasien Working Papers Band 2. doi: 10.1553/soawp2. [DOI]
34. Mallick S, et al. 2016. The Simons genome diversity project: 300 genomes from 142 diverse populations. Nature 538(7624):201–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Matsumura H, et al. 2008. Morphometric affinity of the late Neolithic human remains from Man Bac, Ninh Binh Province, Vietnam: key skeletons with which to debate the ‘two layer’ hypothesis. Anthropol Sci. 116:0802160030. [Google Scholar]
36. Matsumura H, et al. 2019. Craniometrics reveal “two layers” of prehistoric human dispersal in eastern. Eurasia Sci Rep. 9:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. McColl H, et al. 2018. The prehistoric peopling of Southeast Asia. Science 361(6397):88–92. [DOI] [PubMed] [Google Scholar]
38. Mondal M, et al. 2016. Genomic analysis of Andamanese provides insights into ancient human migration into Asia and adaptation. Nat Genet. 48(9):1066–1070. [DOI] [PubMed] [Google Scholar]
39. Mörseburg A, et al. 2016. Multi-layered population structure in Island Southeast Asians. Eur J Hum Genet. 24(11):1605–1611. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. O’Connor RA. 1995. Agricultural change and ethnic succession in Southeast Asian states: a case for regional anthropology. J Asian Stud. 54:968–996. [Google Scholar]
41. Patterson N, et al. 2012. Ancient admixture in human history. Genetics 192(3):1065–1093. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Patterson N, Price AL, Reich D. 2006. Population structure and eigenanalysis. PLoS Genet. 2 (12):e190. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Pickrell J, Pritchard J. 2012. Inference of population splits and mixtures from genome-wide allele frequency data. PLoS Genet. 8:e1002967. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Pierron D, et al. 2014. Genome-wide evidence of Austronesian–Bantu admixture and cultural reversion in a hunter-gatherer group of Madagascar. Proc Natl Acad Sci U S A. 111(3):936–941. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Pittayaporn P. 2014. Layers of Chinese loanwords in protosouthwestern Tai as evidence for the dating of the spread of southwestern Tai. MANUSYA 17(3):47–68. [Google Scholar]
46. Price AL, et al. 2006. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 38(8):904–909. [DOI] [PubMed] [Google Scholar]
47. Purcell S, et al. 2007. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 81(3):559–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. R Core Team. 2013. R: a language and environment for statistical computing. Vienna (Austria: ): R Foundation for Statistical Computing. [Google Scholar]
49. Radcliffe-Brown AR. 2013. The Andaman Islanders. Cambridge: Cambridge University Press. [Google Scholar]
50. Reich D, et al. 2011. Denisova admixture and the first modern human dispersals into Southeast Asia and Oceania. Am J Hum Genet. 89(4):516–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
51. Schebesta P. 1927. Bei den Urwaldzwergen von Malaya. Leipzig: FA Brockhaus. [Google Scholar]
52. Schlebusch CM, et al. 2017. Southern African ancient genomes estimate modern human divergence to 350,000 to 260,000 years ago. Science 358(6363):652–655. [DOI] [PubMed] [Google Scholar]
53. Skeat WW, Blagden O, Ridley HN. 1907. The Pagan Races of the Malay Peninsula. J Straits Branch R Asiat Soc. 49:1–5. [Google Scholar]
54. Teixeira JC, et al. 2021. Widespread Denisovan ancestry in Island Southeast Asia but no evidence of substantial super-archaic hominin admixture. Nat Ecol Evol. 5(5):616–624. [DOI] [PubMed] [Google Scholar]
55. Teo Y-Y, et al. 2009. Singapore Genome Variation Project: a haplotype map of three Southeast Asian populations. Genome Res. 19(11):2154–2162. [DOI] [PMC free article] [PubMed] [Google Scholar]
56. Wang T, et al. 2021. Human population history at the crossroads of East and Southeast Asia since 11,000 years ago. Cell 184(14):3829–3841.e21. [DOI] [PubMed] [Google Scholar]
57. Zheng X,. et al. 2012. A high-performance computing toolset for relatedness and principal component analysis of SNP data. Bioinformatics. 28(24):3326–3328. [DOI] [PMC free article] [PubMed] [Google Scholar]

Supplementary Materials

Data Availability Statement

The data generated in this study will be made available to researchers after review and approval by the Data Access Committee led by the corresponding authors.