DNA-based eye colour prediction across Europe with the IrisPlex system

Area of Science:

• Forensic Genetics
• Human Genetics
• Biotechnology

Background:

• Forensic DNA Phenotyping (FDP) offers a novel approach for identifying unknown individuals when standard methods fail.
• The IrisPlex system was previously developed for DNA-based eye color prediction in Dutch Europeans.
• Existing validated tools for forensic eye color prediction are limited.

Purpose of the Study:

• To validate the IrisPlex prediction model across diverse European populations.
• To assess the accuracy of eye color prediction using extended genotype and phenotype data.
• To confirm the utility of the IrisPlex system in broader forensic applications.

Main Methods:

• IrisPlex analysis was performed on 3840 individuals from seven European sites (EUREYE study).
• Genotype and phenotype data, including high-resolution eye images, were collected.
• Prediction accuracy was evaluated using a probability threshold of 0.7 and Area Under the Curve (AUC).

Main Results:

• The IrisPlex system achieved an average accuracy of 94% for predicting blue or brown eye color across European populations (range 91-98%).
• Area Under the Curve (AUC) values were 0.96 for both blue and brown eye predictions, exceeding previous benchmarks.
• Prediction model parameters derived from the multi-population dataset were highly consistent with earlier findings.

Conclusions:

• The IrisPlex prediction model is accurate and reliable for estimating eye color across various European populations.
• The system demonstrates significant utility for forensic casework, aiding in the identification of unknown individuals.
• Expanded data on European eye color distributions further support the IrisPlex system's forensic applicability.