EFMex: Using EuroForMix to evaluate DNA mixtures with multiple persons of interest.

Area of Science:

• Forensic Science
• Computational Biology
• Genetics

Background:

• DNA mixture interpretation is crucial in forensic science.
• Complex mixtures with related individuals pose interpretation challenges.
• Existing methods may struggle with high allele sharing among contributors.

Purpose of the Study:

• To introduce EFMex (EuroForMix-Exhaustive), an R package for DNA mixture analysis.
• To implement and validate an exhaustive statistical framework for likelihood ratio (LR) calculation.
• To improve the accuracy of identifying contributors in complex forensic DNA mixtures.

Main Methods:

• Developed EFMex, an R package utilizing an exhaustive method for LR calculation.
• Conducted simulation experiments with three- and four-person DNA mixtures.
• Included scenarios with closely related individuals exhibiting high allele sharing.
• Employed a generalized likelihood ratio as an alternative approach for comparison.

Main Results:

• The exhaustive method effectively distinguished true contributors from related non-contributors in simulations.
• EFMex demonstrated robustness even in challenging casework scenarios with high allele sharing.
• Recalculating LRs for candidates with LR>1 further enhanced discrimination.
• The generalized likelihood ratio yielded results comparable to the exhaustive method.

Conclusions:

• The exhaustive method implemented in EFMex is a valuable tool for complex DNA mixture interpretation.
• EFMex provides a robust framework for addressing challenges posed by related individuals in forensic genetics.
• The software, including a user-friendly Shiny app, facilitates practical application in casework.