An Investigation into Compound Likelihood Ratios for Forensic DNA Mixtures

Area of Science:

• Forensic Science
• Genetics
• Biotechnology

Background:

• Understanding the performance of different proposition types is crucial in forensic DNA mixture interpretation.
• Probabilistic genotyping software like STRmix™ is widely used for analyzing complex DNA profiles.

Purpose of the Study:

• To evaluate the performance of simple, conditional, and compound propositions in DNA mixture analysis.
• To compare the differentiating power of these proposition types using STRmix™.

Main Methods:

• Analysis of thirty-two DNA samples across two laboratories, four National DNA Index System (NDIS) Core Loci (NOCs), and four mixtures.
• Mixtures included varying numbers of contributors (N = 2, 3, 4, and 5).
• Utilized STRmix™ probabilistic genotyping software for performance evaluation.

Main Results:

• Conditional propositions demonstrated a significantly higher ability to differentiate true donors from false donors compared to simple propositions.
• Compound propositions were found to potentially misstate the weight of evidence, strongly favoring one proposition over the other.

Conclusions:

• Conditional propositions offer superior discriminatory power in forensic DNA mixture interpretation.
• Careful consideration of proposition formulation is essential to avoid misinterpretation of evidence strength with compound propositions.