Predictive Capacity of Genetic Variants Associated with Eye Color (Forensic DNA phenotyping) in Yucatan, México
TRABAJO LIBRE SOMETIDO AL III CONGRESO INTERNACIONAL DE CIENCIAS FORENSES DE HONDURAS 8-10 DE OCTUBRE 2025, TEGUCIGALPA, HONDURAS
Keywords:
Forensic DNA phenotyping, Eyes color, Genetic variantAbstract
Introduction: Forensic DNA phenotyping allows the inference of physical traits from DNA in the absence of database matches. Although validated in European populations, its application in admixed groups is limited. This study evaluates the predictive capacity of six genetic variants (rs12913832, rs1800407, rs12896399, rs16891982, rs1393350, and rs12203592) associated with iris color in a mestizo sample from Mexico.
Methodology: The association between genetic variants and iris color was analyzed using qPCR with TaqMan probes, image analysis through PCA (RGB-HSV model), and predictive evaluation with AUC curves. A multinomial regression model was built with four color categories (brown, hazel, green, and blue), estimating model fit with Nagelkerke’s R².
Results: The association between genetic variants explained 75.9% of the variability. Variant rs12913832 showed association with more than one eye color, especially genotype GG, discriminating between brown and blue tones. rs12896399, rs16891982, rs1393350, and rs12203592 showed a p value <0.05 for brown, blue, and green (Table 1). Iris pigmentation analysis revealed that the first PCA component explained 51.38% of the variance, differentiating light and dark tones, suggesting genetic influence. For each color, the AUC was: 91.5% for brown, 88.5% for green, 92.7% for blue, and 80.3% for honey.
Discussion and conclusions: Eye color can be predicted using genetic variants, but current models are based on European groups. In populations such as Yucatán, where there is an asymmetric admixture pattern, some variants behave differently, highlighting the importance of adapting it in each region. Although the IrisPlex system performs well for colors like blue and brown, it faces challenges with intermediate tones such as green or honey. These studies help improve accuracy in forensic genetics and make it more inclusive. Variants rs12913832, rs12896399, rs16891982, rs1393350, and rs12203592 were associated with iris color in the Yucatán population, making them potential predictors of brown, green, or blue eyes.
Table 1. Statistically significant associations between SNVs and iris color in volunteers residing in Yucatán.
Iris color
Gene / SNV / Genotype or allele
OR (95% IC) p
Brown
HERC2 (rs12913832) AG
0.076 (0.02 - 0.25) <0.0001
SLC45A2 (rs16891982) GG
0.46 (0.12 – 0.70) 0.006
SLC45A2 (rs16891982) G
0.29 (0.27 – 0.73) 0.001
TYR (rs1393350) A
2.12 (1.16 – 3.86) 0.01
IRF4 (rs12203592) T
0.09 (0.02 – 0.44) 0.0003
Green
HERC2 (rs12913832) AG
11.9 (3.17 – 44.61) 0.004
HERC2 (rs12913832) GG
16 (5.16 – 49.58) < 0.0001
HERC2 (rs12913832) G
9.3 (4.85 – 18.12) < 0.0001
SLC24A4 (rs12896399) T
2.2 (1.19 – 4.14) 0.01
IRF4 (rs12203592) T
3.25 (1.15 – 9.16) 0.02
Blue
HERC2 (rs12913832) GG
28.6 (5.9 – 136.9) <0.0001
HERC2 (rs12913832) G
16.2 (6.38 – 41.12) <0.0001
SLC45A2 (rs16891982) G
5.2 (2.09 – 13.07) <0.0001
TYR (rs1393350) AA
3.9 (1.23 – 12.43) 0.02
Downloads
9
References
Aguilar J, Llamas-de-Dios B, Córdova-Mercado M, Coronado-Ávila C, Salas-Salas O, López-Quintero A. Accuracy of Eye and Hair Color Prediction in Mexican Mestizos from Monterrey City Based on ForenSeqTM DNA Signature Prep. Genes; 2023. 14 (1120). https://doi.org/10.3390/genes14051120
Palmal S, Adhikari K, Mendoza-Revilla J, Fuentes-Guajardo M, Silva C, Bonfante B, et al. Prediction of eye, hair and skin colour in Latin America. Forensic Sci. Int. Genet. 2021; 53: 1-12. https://doi.org/10.1016/j.fsigen.2021.102517.
Paparazzo E, Gozalishvili A, Lagani V, Geracitano S, Bauleo A, Falcone E. A new approach to broaden the range of eye colour identifiable by IrisPlex in DNA phenotyping. Sci. Rep. 2022; 12:12803. Doi: 10.1038/s41598-022-17208-w.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Regina Bass Coli, Lizbeth Josefina González Herrera, Rodrigo Rubí Castellanos

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
El autor conserva los derechos de autor bajo los terminos de una licencia CC NC 4.0
