The variability in likelihood ratios due to different mechanisms.

Area of Science:

• Forensic Science
• Genetics
• Computational Biology

Background:

• Standardization in forensic DNA interpretation has led to the adoption of probabilistic genotyping software.
• Some probabilistic software employs Markov chain Monte Carlo (MCMC) techniques, introducing a novel source of variability due to the inherent randomness of the Monte Carlo method.
• Repeat interpretations using MCMC may yield slightly different results for the same DNA profile.

Purpose of the Study:

• To quantify and compare the variability introduced by MCMC methods in STRmix™ software against other established sources of variability in forensic DNA profiling.
• To assess the impact of MCMC variability within the context of the overall forensic DNA analysis workflow.

Main Methods:

• Comparative analysis of variability sources in forensic DNA profiling.
• Evaluation of Markov chain Monte Carlo (MCMC) variability within STRmix™ software.
• Comparison with variability from Polymerase Chain Reaction (PCR), capillary electrophoresis (CE) parameters (load and injection), and allele frequency database composition.

Main Results:

• Markov chain Monte Carlo (MCMC) variability within the STRmix™ interpretation software was quantified.
• This MCMC variability was compared against other known sources of variability in forensic DNA profiling.
• The MCMC variability was found to be the smallest source of variability among those investigated.

Conclusions:

• The variability introduced by MCMC methods in STRmix™ is minimal compared to other factors in forensic DNA analysis.
• This finding supports the reliability of probabilistic genotyping software utilizing MCMC techniques for DNA interpretation.
• The study highlights the importance of understanding and quantifying all sources of variability in forensic DNA profiling for accurate interpretation.