A mixed DNA profile controversy revisited

Area of Science:

• Forensic Science
• Genetics
• Biostatistics

Background:

• Previous DNA mixture analysis software (LRmix Studio) produced inflated likelihood ratios (LRs) due to incorrect proposition entry and failure to use peak height data.
• Manual exclusion using peak heights identified several individuals who were incorrectly not excluded by the software.

Purpose of the Study:

• To extend previous findings by simulating two-donor DNA mixtures and analyzing them with STRmix™.
• To evaluate the impact of using peak heights and conditioning on the accuracy of DNA mixture analysis.

Main Methods:

• Simulated three two-donor DNA mixtures (1:1, 3:1, 4:1 ratios) using two known contributor genotypes.
• Analyzed simulated mixtures using STRmix™, with and without conditioning on potential contributors, and assessed the role of peak height data.

Main Results:

• Without conditioning, STRmix™ produced ambiguous results, including support for non-donors in low-template mixtures.
• Using conditioning profiles significantly improved discriminatory power, correctly eliminating non-donors in all mixture ratios.
• Peak height information, when utilized, enhances the specificity and sensitivity of DNA mixture analysis.

Conclusions:

• STRmix™ analysis benefits from the incorporation of peak height data and the use of conditioning profiles.
• Conditioning is crucial for accurate interpretation of complex DNA mixtures, especially in low-template or high-inconclusive allele scenarios.
• The recommended approach improves the reliability of forensic DNA mixture interpretation, reducing the risk of false inclusions.