Value of DNA mixture-to-mixture comparisons within an operational context

Area of Science:

• Forensic Science
• Genetics
• Criminal Justice

Background:

• National forensic DNA databases historically exclude complex DNA profiles (mixtures) with >2 contributors or insufficient loci to prevent false leads.
• Interpretable DNA mixtures, comprising ~10% of analyzed traces in western Switzerland, are often not compared nationally due to database limitations.
• Current practices rarely utilize DNA mixtures for series detection, despite their potential investigative value.

Purpose of the Study:

• To evaluate the utility of probabilistic genotyping software (STRmix v2.7) for comparing DNA mixtures not suitable for national database searches.
• To assess the potential of using mixture-to-mixture comparisons to identify common contributors and potential criminal series.
• To determine the investigative value of non-CODIS loadable DNA mixtures for law enforcement.

Main Methods:

• Utilized STRmix v2.7's mixture-to-mixture tool to compare 235 DNA profiles from potential series, collected by six police services.
• Performed 27,495 pairwise comparisons, analyzing results based on likelihood ratios (LR) with a threshold of 1000.
• Correlated identified associations with police-defined investigative expectations (expected, possible, unexpected).

Main Results:

• 88 pairs (0.3%) of DNA profiles showed potential common contributors (LR ≥ 1000).
• Of these, 68.2% were classified as 'expected' associations by police, indicating valid investigative leads.
• The study demonstrated the feasibility of using complex DNA mixtures to uncover potential series, even when they cannot be uploaded to national databases.

Conclusions:

• Non-CODIS loadable DNA mixtures, when analyzed with advanced software, can provide valuable intelligence for identifying potential criminal series.
• This approach enhances investigative capabilities at both local and national levels by leveraging previously underutilized DNA evidence.
• Further consideration of limitations, such as adventitious associations and cost-benefit, is necessary for full integration into investigative processes.