Genetic architecture of dental traits: mapping and pathway enrichment of genome-wide associations

INTRODUCTION

Genome-wide association studies have transformed the understanding of common dental traits by identifying reproducible genetic loci for caries, periodontitis, tooth agenesis, and tooth structure, yet the field has grown in a way that makes synthesis difficult, with each trait studied in a separate literature and few attempts to map them together.(1,2)

The earliest dental genome-wide association studies, conducted in the late 2000s and early 2010s, focused on caries and periodontitis in relatively small samples and yielded few replicable loci, but the subsequent decade brought large meta-analyses and whole-phenome studies that expanded the catalog of associated loci by orders of magnitude, and the field now spans the full range from common inflammatory traits to rare developmental phenotypes.

The mapped genes reported by these studies span developmental transcription factors such as PAX9 and MSX1, which are mutated in the Mendelian forms of tooth agenesis, structural enamel proteins such as amelogenin and enamelin, which are mutated in amelogenesis imperfecta, and immune regulators that predispose to periodontitis, and their biological interpretation depends on knowing which pathways they converge on, information that is scattered across individual reports rather than assembled in one place.(3,4)

The fact that the same genes recur in both rare Mendelian disorders and common complex traits is a well-known feature of dental genetics, and it means that a pathway-level synthesis of the common-variant literature can draw on the rich functional annotation that the Mendelian literature has accumulated.

The classic analyses of Shungin et al.( 5) established that caries and periodontitis have partly shared and partly distinct genetic architectures, and the more recent whole-oral-disease analyses of Karppinen et al.( 6) extended this to dozens of oral phenotypes, but a synthesis of the mapped genes and their pathways across the major dental trait groups has not been assembled.

The question of how diverse the studied populations are is equally pressing, because most genome-wide association studies have been conducted in European ancestry populations, and the extent to which dental trait genetics has repeated this pattern or departed from it has not been quantified.(7,8)

The epidemiology of dental disease is, if anything, the inverse of the ancestry distribution of its genetic studies, since caries and periodontitis burden is highest in low- and middle-income settings and in minority populations, which means that a genetic literature skewed toward European ancestry is least informative precisely for the populations that need it most.

This mismatch between where the disease is and where the genetics is done is not unique to dentistry, but it is particularly consequential for a field whose ultimate goal is prevention, because the genetic findings are only useful to the extent that they can be applied to the populations in which the disease burden is concentrated. A single analysis that maps the genes, pathways, and ancestry of dental trait genetics therefore serves both a scientific and an equity purpose, by showing what has been learned and for whom.

The value of a unified mapping lies in the questions it makes answerable. With a single corpus of dental trait associations, one can ask how the burden of associations and mapped genes is distributed across traits, which biological pathways the mapped genes converge on, and whether the ancestry composition of the studied cohorts reflects the global burden of dental disease.

Each of these questions has a practical consequence: the trait distribution tells funders where the genetic evidence is thin, the pathway convergence tells biologists where to look for mechanism, and the ancestry composition tells the field whether its polygenic tools will travel to the populations that need them.

Previous summaries have addressed individual traits, such as the systematic review of caries genome-wide association studies by Zhou,(9) the genetic reviews of tooth agenesis by Duke,(10) Modafferi,(11) and Fallea,(12) and the genomics of periodontal disease reviewed by Morelli,(13) but none has assembled the five major dental trait groups together with gene mapping, pathway enrichment, and ancestry in a single analysis.

The trait-specific reviews are valuable within their scope, but they cannot answer the cross-trait questions that a unified mapping addresses, such as whether the same developmental pathways recur across caries, tooth agenesis, and enamel structure, or whether the ancestry diversity of the field differs between the inflammatory and the developmental trait groups.

These are not merely descriptive questions, because the answer determines whether a finding in one trait can inform another, whether a pathway discovered in tooth agenesis is a candidate for caries, and whether the genetic tools developed for one population are likely to work in another, all of which bear on the efficient translation of the accumulated genetic evidence.

These cross-trait questions require a common analytical frame applied to all traits simultaneously, which is what the present study provides, and the answers to these questions are what distinguish this synthesis from a mere concatenation of the existing trait-specific reviews.

The objective of this study was to characterize the genetic architecture of dental traits by mapping genome-wide associations to genes and pathways and by describing the ancestry diversity of the studied populations, through three research questions. P1: how many associations and mapped genes are captured for dental traits, and how are they distributed across traits? P2: which biological pathways are enriched among the mapped genes? P3: how diverse is the ancestry composition of the studied cohorts?

METHODS

RESULTS

DISCUSSION

Regarding P1, the distribution of associations and mapped genes across dental traits is highly uneven, with periodontitis and gingival disease carrying more than half of all associations and mapped genes, followed by enamel and dentin structure, whereas tooth agenesis, caries, and malocclusion contribute far fewer. This concentration reflects the large multi-cohort meta-analyses that dominate periodontitis and tooth structure genetics rather than a greater intrinsic genetic complexity of those traits, and it means that the genetic architecture assembled here is, in effect, the architecture of a few large studies.

The contrast with caries, which has only 417 mapped genes despite its high global prevalence, is notable, and it echoes the systematic review of caries genome-wide association studies by Zhou,( 9) which found fewer loci than expected for a trait of such prevalence, and the polygenic score work of Fries et al.,( 17) which showed that even the existing loci explain only a modest fraction of caries liability.

The pulpal and apical disease genome-wide study of Salminen et al.( 18) and the meta-analysis of Nogawa et al.( 19) further illustrate the uneven investment across dental phenotypes, and the integrative analysis of 21 orofacial diseases by Nam et al.( 1) shows that the same unevenness extends across the oral disease spectrum more broadly.

The pattern is not simply a matter of scientific priority, because the scale of the periodontitis and tooth structure studies reflects the availability of large, harmonized cohorts and standardized phenotyping, which those traits have achieved through international consortia, whereas caries has proven more difficult to phenotype consistently across cohorts and has therefore lagged despite its prevalence.

Caries is also unusual among the dental traits in that its heritability is lower and its environmental component larger than that of the developmental traits, which dilutes the genetic signal and makes the discovery of loci more demanding even with equivalent sample sizes, and this biological reality, combined with the phenotyping difficulty, explains why caries genetics has not kept pace with periodontitis genetics.

The contrast between the compact, high-effect architecture of tooth agenesis and the diffuse, polygenic architecture of periodontitis is itself a substantive finding, because it means that the translational prospects of the two trait groups differ: tooth agenesis genes are candidates for mechanistic study and counseling, whereas periodontitis genes point toward polygenic risk stratification rather than single-gene intervention.

This contrast is a direct consequence of the different genetic architectures of developmental and inflammatory traits, and it is visible in the gene counts themselves, with the developmental traits yielding a small number of high-confidence genes and the inflammatory traits yielding a large number of small-effect genes.

Regarding P2, the convergence of the mapped genes on developmental and anatomical structure processes is consistent with the biological expectation that dental traits are, in large part, traits of organ development and morphogenesis, and the presence of a response-to-stimulus term reflects the immune and inflammatory biology that is prominent in periodontitis. The pathways are broad rather than specific, which is a consequence of the large and heterogeneous gene set, and a more focused enrichment within individual traits, particularly the smaller caries and tooth agenesis sets, would be expected to yield more specific terms.

The finding nonetheless confirms that the genomic architecture of dental traits is organized around developmental biology, and it situates the mapped genes within the same developmental program implicated in Mendelian tooth disorders, including the EDA, MSX1, and PAX9 genes that dominate the tooth agenesis literature.(3,4,10,20)

The separation between the developmental terms, which are driven by tooth agenesis and enamel structure, and the response-to-stimulus term, which is driven by periodontitis, is itself a biologically meaningful result, because it shows that the genetic architecture of dental traits partitions into a developmental axis and an inflammatory axis rather than a single shared program.

The periodontitis component, by contrast, points to immune and inflammatory pathways, and the GLT6D1 association reported by Karaoulas et al.,( 21) the combined periodontitis genome-wide association study of Richter et al.,( 22) and the severity association study of Teles et al.,( 23) together with the host genetics reviewed by Nibali( 24) and by Modafferi et al.,( 25) locate this component within immune-mediated disease processes. The two axes are connected, because the same developmental genes that shape the tooth and its supporting structures also determine the architecture of the tissue that periodontitis destroys, but the pathway enrichment suggests that the trait genetics is organized around these two distinct biological poles.

This two-axis structure has practical consequences for the interpretation of genetic findings, because a gene implicated in tooth structure is likely to act through developmental and morphogenetic mechanisms, whereas a gene implicated in periodontitis is more likely to act through immune and inflammatory mechanisms, and the pathway assignment therefore provides a first-pass functional interpretation for the mapped genes.

The developmental axis also connects the common-variant findings to the Mendelian literature, because the same pathways that are implicated by the common variants are disrupted by the rare mutations that cause syndromic and non-syndromic tooth agenesis, and this convergence strengthens the case that the common variants act through the developmental genes in which they are located rather than through distant regulatory effects.

The broadness of the top terms, with hundreds of genes per term, also reflects the fact that the union of all dental trait genes spans much of the genome, and it underscores the value of trait-specific enrichment for any future effort to prioritize individual genes for functional follow-up.

Regarding P3, the ancestry composition is predominantly European at 58.6%, which is lower than the roughly 80% European representation often reported for genome-wide association studies as a whole, and it indicates a comparatively broad, though still incomplete, inclusion of non-European populations in dental trait genetics.

The substantial African American and Afro-Caribbean representation, contributed largely by United States cohort studies of periodontitis, and the East and South Asian component reflect the specific epidemiology and research infrastructure of periodontitis and tooth structure studies, and they distinguish dental trait genetics from fields whose cohorts are even more narrowly European.

Conversely, the European predominance remains a limitation for the transferability of polygenic scores to the populations that bear the greatest dental disease burden, a limitation that the cross-population transferability analyses of Saunders et al.( 26) and the diversity assessments of Molck et al.( 8) have quantified for other traits, and that the polygenic prediction methods of Ruan et al.( 27) and the Hispanic-specific risk models of Maldonado et al.( 28) have sought to address. The inclusion of non-European populations recommended by Ju et al.( 7) is therefore as applicable to dental traits as it is to the wider genome-wide association study enterprise, and the present ancestry profile provides a baseline against which future diversification can be measured.

The 41.4% non-European representation, while far from parity, is a meaningful departure from the most homogeneous fields, and it suggests that dental trait genetics has benefited from the epidemiological interest in periodontitis among African American populations and in tooth traits among East Asian populations, which have partially counterbalanced the default European focus of the field.

Whether this partial diversification is sufficient for the transferability of polygenic scores is a separate question, and the answer is likely to be trait-specific, since a score trained on a trait whose architecture differs across populations will transfer more poorly than one trained on a trait whose architecture is shared.

The developmental traits, whose causal genes are largely shared across populations, may transfer more readily than the inflammatory traits, whose architecture can be shaped by population-specific environmental exposures and allele frequency differences, and this distinction is one that the present ancestry profile can inform but not resolve, since the profile describes who was studied rather than how well the findings travel.

The study has limitations that follow from its design. The mapped genes are those reported by the catalog from genomic intervals rather than from fine-mapped causal genes, so the gene sets are inflated by genes in linkage disequilibrium with the causal variant, and the pathway enrichment reflects this inflation, which is a general limitation of catalog-based gene mapping rather than a defect of the present analysis.

The magnitude of this inflation is largest for the traits with the largest association counts, so the periodontitis and enamel gene counts in particular should be read as upper bounds on the number of genuinely implicated genes, and the pathway enrichment of the developmental terms should be interpreted with the understanding that a developmental locus can implicate many nearby genes that share a genomic interval. The trait grouping relies on the catalog's disease and trait labels, which are heterogeneous and may assign some studies to broad or overlapping groups, and a more granular trait ontology would refine the groupings at the cost of smaller per-trait samples.

The grouping also places biologically distinct phenotypes within the same group, such as gingivitis and periodontitis under the inflammatory heading and amelogenesis imperfecta and dentin defects under the enamel and dentin heading, and a finer-grained mapping would separate these, although the present five-group resolution is sufficient to reveal the major architectural patterns that are the study's focus.

The ancestry table counts each contributing cohort separately, so the large meta-analyses dominate the ancestry composition, and the proportions reflect cohorts rather than participants, which overstates the contribution of studies with many small cohorts. Finally, the pathway enrichment used the full gene set rather than trait-specific sets, which produced broad terms, and a trait-specific enrichment would yield more precise biological interpretation. The multi-omics integration of Cao et al.( 29) and the shared-architecture analyses of Jin et al.( 30) and Ma et al.( 2) point toward the trait-specific analyses that would address this limitation.

The practical implication is that dental trait genetics has reached a scale at which its architecture can be mapped and its gaps identified, and the mapping assembled here points to three priorities: the genetic dissection of caries, which lags behind its disease burden; the functional validation of the developmental genes that dominate the mapped set; and the diversification of study populations beyond European ancestry.

These priorities are mutually reinforcing, because the developmental biology uncovered by the pathway analysis is most clinically relevant when it is translated to the populations in which dental disease is most prevalent, and because the uneven trait coverage and ancestry representation together mean that the current architecture is a partial map of a larger genetic landscape.

The mapping also has immediate uses: it provides a gene and pathway reference for researchers designing functional studies, a baseline against which future ancestry diversification can be measured, and a reminder that the trait distribution of a field is a reflection of its funding and infrastructure history as much as of its biology.

The gene and pathway reference is especially useful because it consolidates information that would otherwise require searching the trait-specific literature one trait at a time, and it provides a starting point for the functional annotation of the many mapped genes whose dental role is not yet understood.

A field that wishes to translate its genetic findings into prevention and treatment must therefore attend not only to the genes it has found but to the traits and populations it has not yet studied, because the gaps in the map are as informative as the map itself when it comes to planning the next experiments and the next cohorts.

CONCLUSIONS

The dental trait genome-wide association literature is concentrated in periodontitis and in enamel and dentin structure, and its mapped genes converge on developmental and anatomical structure pathways, with an immune and inflammatory component contributed by periodontitis, so that the overall genetic architecture partitions cleanly into a developmental axis and an inflammatory axis across the five trait groups studied. The studied cohorts remain predominantly European, with a substantial but still minority non-European representation that limits the transferability of the findings to the populations that bear the greatest dental disease burden worldwide, a limitation that should be acknowledged explicitly. The uneven trait coverage and the ancestry representation together define the priorities for the next stage of dental trait genetics, namely the genetic dissection of caries, the functional validation of the developmental genes, and the further diversification of the studied populations toward the groups that carry the greatest burden of dental disease worldwide.

References

1. 1 Nam K, Eom BS, Kim JY, Lee MG, Lee JM, Lee K, et al. Integrative genomic analysis of 21 orofacial diseases identifies shared genetic architecture with systemic diseases. Nat Commun. 2026;17(1):7186. doi: 10.1038/s41467-026-73925-0
2. 2 Ma X, Zheng X. Multivariate GWAS reveals shared genetic basis of common oral diseases. Front Genet. 2026;17:1807175. doi: 10.3389/fgene.2026.1807175
3. 3 González-Pérez NA, Herrera-Atoche JR, López-González P, Pacheco-Arjona R, Rangel-Méndez JA, Canul-May JE, et al. Genetic Variants of MSX1, PAX9, and AXIN2 in Mayan Probands with Dental Agenesis from Yucatan, Mexico. Odovtos - Int J Dent Sc. 2024;26(3):266-282. doi: 10.15517/ijds.2024.60223
4. 4 Su L, Lin B, Yu M, Liu Y, Sun S, Feng H, et al. EDA Variants Are Responsible for Approximately 90% of Deciduous Tooth Agenesis. IJMS. 2024;25(19):10451. doi: 10.3390/ijms251910451
5. 5 Shungin D, Haworth S, Divaris K, Agler CS, Kamatani Y, Keun Lee M, et al. Genome-wide analysis of dental caries and periodontitis combining clinical and self-reported data. Nat Commun. 2019;10(1):2773. doi: 10.1038/s41467-019-10630-1
6. 6 Karppinen K, Ollila HM, Batool K, Abner E, Rice DP, Palotie A, et al. Genetic architecture of 67 oral diseases and their links to systemic diseases. Human Genetics and Genomics Advances. 2026;7(3):100633. doi: 10.1016/j.xhgg.2026.100633
7. 7 Ju D, Hui D, Hammond DA, Wonkam A, Tishkoff SA. Importance of Including Non-European Populations in Large Human Genetic Studies to Enhance Precision Medicine. Annu Rev Biomed Data Sci. 2022;5(1):321-339. doi: 10.1146/annurev-biodatasci-122220-112550
8. 8 Molck MC, Santos LHBD, Oliveira TCD, Lopes-Cendes I. Exploring genetic diversity in genome-wide association studies and polygenic risk scores in the admixed Brazilian population. Genet Mol Biol. 2026;49(suppl 1):e20250235. doi: 10.1590/1678-4685-gmb-2025-0235
9. 9 Zhou K, Gao C, Wu J, Banerjee A, Ide M, Kang J. Genome-Wide Association Studies on Dental Caries: A Systematic Review. Caries Res. 2025:1-20. doi: 10.1159/000548693
10. 10 Duke A, Paterson M, P. Ashley M, MacNab L. The genetic basis of hypodontia in dental development. Br Dent J. 2023;235(7):525-528. doi: 10.1038/s41415-023-6384-6
11. 11 Modafferi C, Tucci I, Bogliardi FM, Gimondo E, Chiurazzi P, Tabolacci E, et al. Genetic Aspects of Tooth Agenesis. Genes. 2025;16(5):582. doi: 10.3390/genes16050582
12. 12 Fallea A, Vinci M, L’Episcopo S, Bartolone M, Musumeci A, Ragalmuto A, et al. Dissecting the Genetic Contribution of Tooth Agenesis. IJMS. 2025;26(21):10485. doi: 10.3390/ijms262110485
13. 13 Morelli T, Agler CS, Divaris K. Genomics of periodontal disease and tooth morbidity. Periodontology 2000. 2020;82(1):143-156. doi: 10.1111/prd.12320
14. 14 Kastenbom L, Esberg A, Haworth S. Introduction to genome-wide association studies: Exploring their utility in caries and periodontitis. Community Dental Health. 2026;43(1):25-32. doi: 10.1177/0265539x251381496
15. 15 Agler CS, Moss K, Philips KH, Marchesan JT, Simancas-Pallares M, Beck JD, et al. Biologically Defined or Biologically Informed Traits Are More Heritable Than Clinically Defined Ones: The Case of Oral and Dental Phenotypes. In: Advances in Experimental Medicine and Biology. Springer International Publishing; 2019. p. 179-189. doi: 10.1007/978-3-030-28524-1_13
16. 16 Divaris K, Haworth S, Shaffer JR, Anttonen V, Beck JD, Furuichi Y, et al. Phenotype Harmonization in the GLIDE2 Oral Health Genomics Consortium. J Dent Res. 2022;101(11):1408-1416. doi: 10.1177/00220345221109775
17. 17 Fries N, Haworth S, Shaffer JR, Esberg A, Divaris K, Marazita ML, et al. A Polygenic Score Predicts Caries Experience in Elderly Swedish Adults. J Dent Res. 2024;103(5):502-508. doi: 10.1177/00220345241232330
18. 18 Salminen A, Hyvärinen K, Ritari J, Leppilahti JM, Palotie U, Vuollo V, et al. Genome-wide association study of pulpal and apical diseases. Nat Commun. 2025;16(1):6774. doi: 10.1038/s41467-025-61721-1
19. 19 Nogawa S, Morishita S, Saito K, Kato H. Genome-wide association meta-analysis identifies two novel loci associated with dental caries. BMC Oral Health. 2024;24(1):1003. doi: 10.1186/s12903-024-04799-1
20. 20 Yu K, Dou J, Huang W, Wang F, Wu Y. Expanding the genetic spectrum of tooth agenesis using whole‐exome sequencing. Clinical Genetics. 2022;102(6):503-516. doi: 10.1111/cge.14225
21. 21 Karaoulas TA, Xirouchakis C, Neophytou C, Fragkioudakis I. GLT6D1 rs1537415 G Allele and Advanced Periodontitis: A Meta-Analysis. Cureus. 2026;18(8):e114871. doi: 10.7759/cureus.114871
22. 22 Richter GM, Akinloye OM, Kühnlenz T, Weiner J3, Holtfreter B, de Coo A, et al. Combined periodontitis GWAS identifies LINC01541 as a regulator of innate immunity in the oral mucosa. Hum Mol Genet. 2026;35(9):ddag039. doi: 10.1093/hmg/ddag039
23. 23 Teles F, Chandrasekaran G, Martin L, Shrestha P, Moss K, Patel M, et al. Genome‐wide association study of periodontitis severity and progression. Journal of Periodontology. 2026;97(2):247-258. doi: 10.1002/jper.70017
24. 24 Nibali L, Di Iorio A, Tu Y, Vieira AR. Host genetics role in the pathogenesis of periodontal disease and caries. J Clinic Periodontology. 2017;44(S18):S52-S78. doi: 10.1111/jcpe.12639
25. 25 Modafferi C, Grippaudo C, Corvaglia A, Cristi V, Amato M, Rigotti P, et al. Genetic Testing in Periodontitis: A Narrative Review on Current Applications, Limitations, and Future Perspectives. Genes. 2025;16(11):1308. doi: 10.3390/genes16111308
26. 26 Saunders GRB, Vrieze S, Liu DJ. Insights into the Cross-Population Transferability of Polygenic Scores for Substance Use. Behav Genet. 2026;56(3):119-130. doi: 10.1007/s10519-026-10265-1
27. 27 Ruan Y, Lin YF, Feng YCA, Chen CY, Lam M, Guo Z, et al. Improving polygenic prediction in ancestrally diverse populations. Nat Genet. 2022;54(5):573-580. doi: 10.1038/s41588-022-01054-7
28. 28 Maldonado BL, Piqué DG, Kaplan RC, Claw KG, Gignoux CR. Genetic risk prediction in Hispanics/Latinos: milestones, challenges, and social-ethical considerations. J Community Genet. 2023;14(6):543-553. doi: 10.1007/s12687-023-00686-4
29. 29 Cao R, Xu S, Xiang J, Qu B, Zhuang Z, Chu T. Multi-Omics Integration Reveals the Genetic Mechanisms of Periodontitis and Predicts Therapeutic Drugs. International Dental Journal. 2026;76(4):109678. doi: 10.1016/j.identj.2026.109678
30. 30 Jin T, Lin J, Zhang P, Lu Y, Chen S, Lin H. Exploring the shared genetic architecture between periodontitis and cardiovascular disease. BDJ Open. 2026;12(1):28. doi: 10.1038/s41405-026-00421-4