Background: Phenotype prediction for eye, hair and skin color is used in a variety of forensic applications, such as trace analysis, the identification of unknown individuals, and analysis of historical DNA traces. The aim of this study was to evaluate the predictive accuracy of the HIrisPlex-S system in a homogeneous North German population. Methods: A cohort of 155 individuals from this population was sampled, and the 41 HIrisPlex-S SNPs were genotyped using the SNaPshot workflow. In addition, the participants assessed their own eye, hair, and skin color using a standardized questionnaire. The statistical analysis included the calculation of diagnostic indicators such as sensitivity (Sens), specificity (Spec), positive and negative predictive values, and accuracy (Acc). In addition, ROC analyses were performed. Results: The results indicated that predictions of skin and hair color were less accurate, whereas eye color could be determined more reliably. Brown and blue eye colors in particular were predicted accurately (brown: Sens = 94.7%, Spec = 87.7%, Acc = 89.5%; blue: Sens = 98.5%, Spec = 57.7%, Acc = 75.7%), while intermediate eye color (Sens = 0.0%, Spec = 100.0%, Acc = 69.1%), hair color and skin color were difficult to differentiate (e.g., blond hair color: Sens = 80.8%, Spec = 56.0%, Acc = 68.2% and pale skin color: Sens = 73.8%, Spec = 44.8%, Acc = 57.2%). Conclusions: In our study, the HIrisPlex-S system primarily provided rough directional information and could distinguish between very different phenotypes but reached its limits when it comes to similar characteristics.
Caliebe et al. (Mon,) studied this question.