Compound and Conditioned Likelihood Ratio Behavior within a Probabilistic Genotyping Context.

Area of Science:

• Forensic Science
• Genetics
• Biostatistics

Background:

• Forensic DNA analysis often involves complex mixtures from multiple contributors.
• Evaluating the statistical weight of evidence for joint contributors is crucial in casework.
• Existing guidance suggests comparing individual likelihood ratios (LR) to a joint LR for mixed DNA profiles.

Purpose of the Study:

• To assess the consistency of compound likelihood ratios (LR) with additive behavior compared to simple LR products.
• To aid casework analysts in evaluating statistical weights for mixed DNA profiles.
• To understand the impact of template level and mixture composition on LR calculations.

Main Methods:

• Interpreted two-, three-, and four-person DNA mixture data using STRmix® version 2.8 Probabilistic Genotyping Software.
• Compared compound LRs with simple LR products derived from individual likelihood ratios.
• Analyzed the distribution of log(LR) differences between compound and simple LR comparisons.

Main Results:

• Relative magnitudes of LR increases depended on template level and mixture composition.
• The distribution of log(LR) differences between compound and simple LR comparisons ranged from approximately -2.7 to 28.3.
• Information gain from constrained genotype combinations had a comparable impact whether applied before or after interpretation, peaking around 0.5.

Conclusions:

• Compound LRs generally conform to additive behavior when compared to simple LR products.
• The findings provide insights for casework analysts on evaluating statistical weights in mixed DNA profile interpretation.
• The study confirms that probabilistic genotyping software aids in consistent statistical interpretation of complex DNA mixtures.