A continuous model for interpreting microhaplotype profiles of forensic DNA mixtures

Area of Science:

• Forensic genetics
• Population genetics
• Statistical genetics

Background:

• Microhaplotypes (MHs) are valuable forensic DNA markers.
• Analyzing DNA mixtures is a common challenge in forensic investigations.
• Existing methods may struggle with complex mixture interpretation.

Purpose of the Study:

• Develop and evaluate a continuous model for interpreting mixed genotype data from MH markers.
• Assess the model's accuracy in identifying true contributors and non-contributors.
• Improve mixture deconvolution for forensic applications.

Main Methods:

• Characterized MH profile features.
• Modeled allele read counts using a truncated Gaussian distribution.
• Tested the model on 90 DNA mixtures from nine individuals across various proportions.

Main Results:

• High accuracy and specificity in interpreting 2- to 3-person MH DNA mixtures.
• True contributors achieved LR > 1 in 190/200 calculations.
• Low false positive rates for non-contributors in simulated tests (0.0051% for 2-person, 4.68% for 3-person mixtures).
• Average deconvolution accuracy for major contributors was 0.9145.

Conclusions:

• The developed continuous model accurately interprets mixed MH genotype data.
• The model shows high specificity and accuracy for forensic applications.
• Noise and mixture complexity impact genotyping accuracy, highlighting areas for future refinement.