1 Department of Applied Mathematics and Computer Science, Technical University of Denmark2 Image Analysis & Computer Graphics, Department of Applied Mathematics and Computer Science, Technical University of Denmark3 University of Copenhagen4 Aalborg University5 Copenhagen Center for Health Technology, Center, Technical University of Denmark
In two recent studies of Spanish individuals [1,2], gender was suggested as a factor that contributes to human eye colour variation. However, gender did not improve the predictive accuracy on blue, intermediate and brown eye colours when gender was included in the IrisPlex model . In this study, we investigate the role of gender as a factor that contributes to eye colour variation and suggest that the gender effect on eye colour is population specific. A total of 230 Italian individuals were typed for the six IrisPlex SNPs (rs12913832, rs1800407, rs12896399, rs1393350, rs16891982 and rs12203592). A quantitative eye colour score (Pixel Index of the Eye: PIE-score) was calculated based on digital eye images using the custom made DIAT software. The results were compared with those of Danish and Swedish population samples. As expected, we found HERC2 rs12913832 as the main predictor of human eye colour independently of ancestry. Furthermore, we found gender to be significantly associated with quantitative eye colour measurements in the Italian population sample. We found that the association was statistically significant only among Italian individuals typed as heterozygote GA for HERC2 rs12913832. Interestingly, we did not observe the same association in the Danish and Swedish population. This indicated that the gender effect on eye colour is population specific. We estimated the effect of gender on quantitative eye colour in the Italian population sample to be 4.9%. Among gender and the IrisPlex SNPs, gender ranked as the second most important predictor of human eye colour variation in Italians after HERC2 rs12913832. We, furthermore, tested the five lower ranked IrisPlex predictors, and evaluated all possible 36 (729) genotype combinations of the IrisPlex assay and their corresponding predictive values using the IrisPlex prediction model . The results suggested that maximum three (rs12913832, rs1800407, rs16891982) of the six IrisPlex SNPs are useful in practical forensic genetic casework.
Forensic Science International: Genetics, 2014, Vol 11, p. 1-6
Forensic genetics; Forensic DNA phenotype; Pigmentation; Eye colour; Gender