Our study presents the first analysis of HLA diversity in the Maniq, a small, isolated nomadic hunter-gatherer group inhabiting the rainforests of Southern Thailand. Recent mitochondrial and genome-wide studies (Kutanan et al. 2018; Göllner et al. 2022) have shed light on the Maniq’s demographic history, providing important context for interpreting their HLA diversity. Their unique demographic trajectory is characterized by early divergence from other Southeast Asian groups, limited gene flow, and prolonged isolation. Genome-wide data suggest that the Maniq retain substantial ancient (hunter-gatherer) Hòabìnhian-related ancestry, combined with approximately 35% East Asian–related admixture introduced through more recent contact with agriculturalist populations, followed by strong genetic drift and endogamy (Göllner et al. 2022). Such a demographic history has probably contributed to the reduced HLA diversity and the skewed allele-frequency patterns we observe. By identifying 32 alleles from classical HLA genes and 14 from non-classical HLA genes, we reveal patterns of genetic diversity shaped by demographic history, drift, and selection pressures.

We determined the HLA alleles using two recently developed methods, HLA-HD (Kawaguchi et al. 2017) and T1K (Song et al. 2023), applied to WGS data. Both methods yielded highly consistent results, underscoring the robustness of these methods in determining HLA diversity from WGS data. Notably, a few alleles at each HLA locus occur at very high frequencies in the Maniq population. This pattern mirrors findings from the Aché, a small Amerindian population of semi-nomadic hunter-gatherers in Eastern Paraguay (Tsuneto et al. 2003; Single et al. 2020). These parallels suggest that restricted HLA diversity, with a concentration of dominant alleles, may characterize very small, isolated indigenous populations, driven by genetic drift and local adaptation. Our previous research (Göllner et al. 2022) showed that the Maniq have one of the highest levels of genetic drift among living human populations, likely due to their prolonged geographic isolation, small population size, and history of endogamy. Consequently, HLA diversity in the Maniq is also relatively low compared to other Southeast Asian populations, possibly resulting in skewed HLA allele frequencies.

However, despite reduced diversity, the Maniq share specific HLA alleles with other Southeast Asian populations, indicating shared ancestry and potential common adaptive responses to regional pathogens. Some of these shared alleles have been found to be associated with protective immunity in some Asian populations (Supplementary Table 2). The most common class Ia alleles in the Maniq were A*24:07:01, B*13:01:01, and C*03:04:01; common class IIb alleles included DRB1*15:01:01, DQA1*01:02:01, DQB1*05:02:01, DPA1*01:03:01, and DPB1*02:01:02. The presence of rare HLA alleles such as B*27:06:01, B*38:02:01, and C*07:199:01, which are uncommon in the broader Asian region, suggests unique evolutionary pressures on the Maniq or the retention of ancestral alleles possibly lost in other populations. Principal component analysis (PCA) of HLA allele frequencies (Fig. 2) revealed distinct patterns of immunogenetic structure among Southeast Asian, Australian, and Oceanian populations. The first two principal components (PC1 and PC2) explained 18.7% and 14.7% of the total variance, respectively. The hunter-gatherer (Semang) populations Maniq, Jehai, and Kintaq formed a tight cluster, closely aligned with the Tao, an indigenous Austronesian-speaking group of Taiwan, and the indigenous groups from Papua New Guinea (Goroka, Madang). Interestingly, despite linguistic and cultural differences—the Semang being Austroasiatic speakers and the Tao being Austronesian—their clustering may reflect shared immune pressures from similar environments or ancient genetic links across Island and Mainland Southeast Asia (Jinam et al. 2012). In contrast, East Asian populations (Chinese, Japanese, Vietnamese) clustered distinctly, with Aboriginal Australians (Kimberly, Cape York) and Māori forming more differentiated positions along both PCs. This structure mirrors broader continental-scale HLA differentiation patterns, such as those identified by Arrieta-Bolaños et al. (2023), who reported marked HLA discontinuities across Southeast Asia, including along the Wallace Line. These findings suggest that populations, even if geographically isolated, are part of larger immunogenetic ecosystems shaped by migration, drift, and region-specific pathogen pressures.

Some of the HLA alleles such as C*07:199:01 are fairly common (~ 18%) across Semang groups, suggesting not only shared ancestry but potentially similar selection pressure maintaining specific HLA alleles at high frequency in the hunter-gatherers on the Thai-Malay Peninsula (Fig. 1). Notably, HLA-C*07:199:01 is nearly identical in sequence to C*07:04:01 (sequence data from the IPD-IMGT/HLA database (Robinson et al. 2020)), differing only at codon 95 in exon 3, where a phenylalanine to leucine substitution occurs. Due to this subtle difference, earlier studies based on lower-resolution HLA typing may have misclassified or failed to report the allele C*07:199 in Southeast Asian populations.

Although specific health data for the Maniq are lacking, studies on related Orang Asli groups in Malaysia indicate high infectious disease burdens, including infections with soil-transmitted helminths, protozoan parasites, and viral pathogens such as hepatitis B virus (HBV) (Sahlan et al. 2019; Mahmud et al. 2022). Notably, recent research has shown that HBV infection rates in some Semang populations in Malaysia are almost three times higher than the national average (Sahlan et al. 2019). Given their geographic proximity and similar subsistence practices, it is plausible that the Maniq experiences comparable pathogen pressures, which may have influenced the selection of specific HLA alleles associated with immunity to these diseases. Several HLA alleles detected in the Maniq are associated with either protective effects or increased susceptibility to specific pathogens. For instance, HLA-DPB1*02:01 is linked to protection against chronic HBV infection (Nishida et al. 2014; Ou et al. 2021), while HLA-DPB1*05:01 and DQB1*05:02 are associated with increased susceptibility (Zhu et al. 2007; Ou et al. 2021). The role of HLA genes in HBV infection is further supported by the prevalence of the common allele HLA-B*13:01 in the Maniq population, which has been associated with enhanced clearance of hepatitis B surface antigen in Asian populations (Miao et al. 2013). Furthermore, a study revealed that DQB1*03:03, which is a common allele in the Maniq population, is associated with protection against Helicobacter pylori (Hp) infection in Asian populations (Wang et al. 2015). Moreover, high prevalence of amoebiasis, caused by Entamoeba histolytica infection, has been recorded among Orang Asli (Anuar et al. 2012), and in a study on Bangladeshi children, it has been found that the heterozygous haplotype DQB1*06:01–DRB1*15:01 had protective effects against this infection (Duggal et al. 2004). Although DQB1*06:01 was not detected in our study, DRB1*15:01 has the highest frequency among DRB1 alleles in the Maniq. This allele has also been found to be associated with a protective role against leishmaniasis (Blackwell et al. 2020). Several of the HLA alleles commonly found in the Maniq, including HLA-C*03:04, DRB1*09:01, DRB1*15:01, and DQB1*05:02, are among globally frequent alleles, reinforcing their potential long-term adaptive value (Sanchez-Mazas et al. 2024). Moreover, Arrieta-Bolaños et al. (2023) identified strong genetic barriers in HLA diversity across Southeast Asia, notably along the Wallace Line, suggesting that even isolated populations such as the Maniq are embedded within larger immunogenetic ecosystems shaped by shared histories of migration, drift, and exposure to regional pathogen pressures. These observations show that the HLA profile of the hunter-gatherer groups on the Thai-Malay Peninsula (see PCA in Fig. 2), while unique, reflects broader patterns of selection acting on human populations across time and geography.

We did not find any significant deviation from HWE at the HLA loci (Supplementary Table 3). The absence of HWE deviations despite pathogen-associated alleles may reflect a long-standing equilibrium shaped by past selection events, suggesting we may be observing the genetic legacy of ancient host–pathogen interactions. However, the EW tests of selective neutrality revealed significant (p > 0.975) deviations at DQA1, DMA, and DMB, with higher observed homozygosity than expected under neutrality, indicating potential directional selection or the effects of strong genetic drift at these loci (Supplementary Table 4). For other HLA loci, no significant deviation from neutrality was detected, suggesting more neutral patterns of allele frequency distribution.

The MHC SNP-based analyses revealed variants linked to different HLA genes under both balancing selection and positive selection (Figs. 3 and 4; Supplementary Tables 5,6 and 7). Balancing selection plays a crucial role in maintaining genetic diversity at immune-related loci, allowing populations to respond to a diverse array of pathogens. Our analysis identified classical HLA class Ia and all classical class IIa loci as candidates under balancing selection. Notably, DPA1 and DPB1 exhibited the highest Beta2_std scores (Supplementary Table 7), indicating strong selective pressure to maintain diversity at these loci. Interestingly, the non-classical HLA class IIb gene DOA also emerged as a candidate under balancing selection. Beyond that, we observed signals of recent positive selection at multiple classical HLA loci, and xp-EHH pinpointed the non-classical locus DOB as under positive selection (Fig. 3). The fact that both classical and non-classical HLA loci show different selective signatures underscores that the entire MHC region may experience varied evolutionary pressures, reflecting the broad pathogen landscape confronting the Maniq. Moreover, the overlapping evidence for long-term balancing and recent positive selection at certain HLA loci highlights a complex interplay wherein populations retain genetic diversity to combat numerous pathogens while also adapting to specific, high-prevalence threats. Notably, DPA1 and DPB1 under positive selection in the Maniq have also been reported as positively selected in indigenous Peruvian and Mesoamerican populations (Caro-Consuegra et al. 2022; Garcia et al. 2023). The lead SNPs at DPA1 and DPB1 are in complete LD with functional 3′UTR variants (rs3077 and rs9277535) associated with HBV infection outcomes in Asian populations (Kamatani et al. 2009; An et al. 2011; Nishida et al. 2014; Mardian et al. 2017). A recent study reported higher levels of HBV diversity in Eastern Eurasia compared to Western Eurasia between 5000 and 3000 years ago, as well as a possible transition from non-recombinant HBV sub-genotypes to recombinant sub-genotypes (Sun et al. 2024). These historical patterns support the hypothesis that HBV may have exerted strong selective pressure favoring protective HLA-DP variants in the Maniq, evidenced by their elevated frequencies and high FST at these SNPs (Supplementary Fig. 2) relative to East Asian populations. These results indicate a central role of HLA genes in host–pathogen co-evolution and the adaptive immune response to viral pathogens like HBV.

In conclusion, HLA diversity in the Maniq reflects a dynamic interplay of genetic drift, balancing selection, and recent positive selection—shaped by unique demographic history and pathogen pressures. These insights underscore the adaptive importance of HLA alleles in small, isolated populations that face diverse pathogenic challenges. The high concordance between FASTQ- and SNP-based genotypes lends confidence to our findings, yet the small sample and the mapping pitfalls of short-read data caution against overinterpretation of locus-specific signals. Future work should combine (i) larger Maniq and neighboring Semang cohorts, (ii) long-read or graph-based assemblies to resolve the complex MHC, and (iii) matched immunological phenotypes. Such integrative studies will clarify how drift, migration, and region-specific pathogens have jointly sculpted HLA evolution in Southeast Asia’s remaining hunter-gatherer populations.