Scoping Review Finds Forensic DNA Phenotyping Fails Southeast Asian Populations

A scoping review published in the Egyptian Journal of Forensic Sciences has mapped the state of forensic DNA phenotyping (FDP) research across Southeast Asian populations, and its conclusions point to a substantial genomic gap. Following PRISMA-ScR guidelines, the review synthesized 25 eligible studies published between 2020 and 2025, drawn from genome-wide association studies (GWAS), whole-genome sequencing investigations, and forensic validation datasets. The central finding is that most FDP models are calibrated on European or East Asian datasets, limiting their reliability in one of the most genetically diverse regions on Earth.

Forensic DNA phenotyping uses single nucleotide polymorphisms (SNPs) to predict externally visible characteristics (EVCs), such as hair, eye, and skin pigmentation, as well as biogeographic ancestry. Conventional STR-based DNA profiling depends on matching against reference databases and offers no investigative lead when no match exists. FDP is meant to fill that void by narrowing suspect pools probabilistically when traditional profiling proves inconclusive. But the review emphasizes that these predictions are only as good as the population data behind them, and Southeast Asian populations remain markedly underrepresented in global genomic datasets, accounting for less than 1% of genomic data according to the review’s comparative analysis.

The review’s most striking evidence concerns allele frequency stratification within a single country. The SLC24A5 rs1426654 variant, often called a light skin allele, shows a sharp drop in frequency across Indonesia: it occurs in roughly 35% of Javanese individuals but only 15% of Papuan groups. This intra-national contrast sits against near-fixation of the allele in European (99%) and Han Chinese (97%) populations and near-absence in Sub-Saharan African populations (4%). The authors argue that such heterogeneity makes uniform “Asian” reference assumptions a source of predictive bias, since allele frequencies and linkage patterns vary substantially across populations.

Beyond classical pigmentation loci, the review catalogued novel associations identified in Southeast Asian cohorts that have not yet entered standard forensic panels. A multi-country analysis of Southeast Asian women identified ADGRL2 as associated with cheek coloration and LINGO2 with pore size, while TANC1 was linked to sebum production. In Vietnamese cohorts, polygenic analyses implicated BNC2 and MC1R pathways in skin-related traits including pigmentation and acne predisposition. These skin texture and surface traits contribute to visible facial appearance, yet panels such as HIrisPlex-S and VISAGE do not routinely incorporate them.

Craniofacial morphology, one of the most recognizable but genetically complex human traits, remains poorly covered by direct Southeast Asian research. Only a limited number of studies directly investigated facial shape in SEA populations. Most loci cited in the review, including PAX3, DCHS2, PRDM16, and EDAR, originate from East Asian comparative studies or multi-ancestry GWAS conducted in Korean, Japanese, and Chinese cohorts. Facial morphology is highly polygenic, with hundreds of loci contributing modest effects and heritability estimates for certain midfacial parameters exceeding 0.80, but the modest proportion of variance explained by currently identified loci constrains high-resolution facial reconstruction from DNA alone.

The evaluation of commercial forensic tools revealed concrete performance limitations. When researchers assessed the Illumina ForenSeq kit in Singaporean cohorts, they reported clustering overlap between Malay and Indonesian individuals and limited differentiation for Filipino and Burmese populations. Ancestry-informative markers accounted for only a modest proportion of genetic variance, indicating restricted discriminatory capacity at fine-scale regional levels. The EDAR V370A variant illustrates the regional divergence: while nearly fixed in East Asian (94%) and Native American (85%) cohorts, its frequency falls to 75% in Thailand and 55% in Indonesia, contrasting sharply with negligible occurrences in Europe (8%) and Africa (1%).

Pigmentation prediction fared better than ancestry inference in Southeast Asian-related studies, but intermediate phenotypes common in the region were inconsistently classified in admixed populations. Proportional ancestry analysis showed Vietnam with the highest East Asian ancestry component at approximately 85%, while Malay, Javanese, and Filipino cohorts displayed moderate East Asian proportions ranging from 55% to 65% alongside substantial mixed ancestry components. HIrisPlex-S and the VISAGE Enhanced Tool demonstrated strong performance for pigmentation traits in European contexts, but they were predominantly calibrated on European population data and exhibit reduced sensitivity when applied to genetically diverse and admixed Southeast Asian groups.

The review also exposed uneven sampling within the region itself. Research clustered primarily in Thailand and Indonesia, while the Philippines, Myanmar, Laos, and Cambodia showed lower published research volumes. Within the compiled dataset, Javanese, Sundanese, Madurese, and Bugis groups constituted approximately 67.5% of reviewed data, followed by Malay, Batak, and Minangkabau groups at 20.0%, with Ambonese, Papuan, and Moluccan groups at 8.0% and Balinese, Sasak, and Timorese groups at just 4.5%. Several cohorts relied on urban or convenience sampling, potentially underrepresenting indigenous populations, and most association studies were exploratory rather than designed for predictive deployment.

Methodologically, the review searched PubMed, Scopus, and Web of Science for studies published between January 2020 and December 2025. From 1,551 database records and 195 citation records, 356 duplicates were removed, 1,390 records screened, and 129 reports assessed for eligibility before 104 were excluded, leaving 25 studies. All 25 were rated as low risk of bias using the Joanna Briggs Institute’s assessment for analytical cross-sectional studies. Eleven studies used GWAS, six employed whole-genome sequencing, and eight applied targeted SNP panels or massively parallel sequencing for forensic validation. Because of heterogeneity in phenotype definition, sequencing depth, and statistical modeling, findings were synthesized descriptively rather than quantitatively.

The authors conclude that region-specific genomic databases are essential to improve the accuracy and equity of DNA-based facial prediction in Southeast Asia. They call for large-scale whole-genome sequencing initiatives in underrepresented regions such as the Philippines and Eastern Indonesia, validation of existing FDP tools against local datasets to quantify error rates, and integration of fine-scale morphological data into prediction models. The review also flags ethical and legal concerns: the region lacks FDP-specific statutory frameworks, and probabilistic pigmentation predictions could inadvertently reinforce colorism and racial profiling in ethnically diverse contexts. Until regionally calibrated resources exist, the authors stress that FDP should be treated as a probabilistic investigative tool rather than a deterministic identification method.

Subject of Research: Forensic DNA phenotyping research and population genetic representation in Southeast Asia

Article Title: Forensic DNA phenotyping in Southeast Asian populations: a scoping review

Article References: Setiawardani, A., Zainuddin, Z., Hadi, H., Furqoni, A. H., & Hanis Zainal Abidin, N. (2026). Forensic DNA phenotyping in Southeast Asian populations: a scoping review. Egyptian Journal of Forensic Sciences, 16(1), Article 84. https://doi.org/10.1186/s41935-026-00579-w

Image Credits: AI Generated

DOI: 10.1186/s41935-026-00579-w

Keywords: forensic DNA phenotyping, Southeast Asia, scoping review, SLC24A5, genomic gap, population genetics, HIrisPlex-S, ForenSeq, craniofacial morphology, pigmentation, allele frequency, forensic genetics