Further development of forensic eye color predictive tests

Area of Science:

• Forensic genetics
• Human genetics
• Molecular biology

Background:

• Predictive testing for external visible characteristics (EVCs) aids criminal investigations.
• Previous studies identified key single nucleotide polymorphisms (SNPs) for eye color prediction, but differed on additional markers.
• Intermediate eye colors present challenges for accurate prediction due to phenotypic variation.

Purpose of the Study:

• To develop and evaluate improved SNP-based assays for predicting eye, skin, and hair color.
• To assess the impact of an expanded SNP set on eye color prediction accuracy.
• To create a flexible online Bayesian classifier for forensic eye color prediction.

Main Methods:

• Developed two single base extension assays targeting 37 SNPs in pigmentation genes.
• Tested assays on 416 individuals from diverse European populations.
• Utilized AUC analysis and Bayesian classification, including a modified online classifier for SNP profiles.

Main Results:

• Confirmed rs12913832 as a key predictor, with rs1129038 identified as the second most associated SNP.
• Demonstrated that an expanded set of SNPs improves eye color prediction accuracy.
• The modified online classifier successfully handles incomplete SNP profiles and linked SNPs.

Conclusions:

• Expanded SNP combinations enhance the predictive power of forensic eye color tests.
• The developed online classifier offers a flexible and robust tool for forensic geneticists.
• This approach balances predictive value with the practicalities of typing challenging DNA samples.