Evaluation of low-template DNA profiles using peak heights.

Area of Science:

• Forensic Science
• Statistical Genetics
• Computational Biology

Background:

• Statistical models for DNA profile analysis have evolved from presence/absence data to quantitative peak height analysis.
• Peak height data in electropherograms is highly variable due to numerous influencing factors.
• Existing models often struggle with complex DNA profiles, including low-template and mixed samples.

Purpose of the Study:

• To present a novel statistical peak-height model for complex DNA profile analysis.
• To incorporate advanced features like over/double-stutter and improved drop-in handling.
• To provide open-source R code (likeLTD) for practical application and validation.

Main Methods:

• Development of a new peak-height model with novel features (over/double-stutter, drop-in).
• Implementation of the model in open-source R code (likeLTD).
• Application to 108 laboratory-generated crime-scene profiles and novel model validation techniques.

Main Results:

• Demonstrated the application of the new model to real-world forensic samples.
• Explored the benefits and limitations of peak height modeling in complex DNA analysis.
• Assessed the utility of pre-extraction replication and introduced computational approximations for multiple low-level contributors and locus linkage.

Conclusions:

• The developed peak-height model offers advanced capabilities for complex DNA profile analysis.
• Peak height modeling is not universally advantageous and requires careful consideration.
• The study provides valuable tools and insights for forensic DNA interpretation, including approximations for computational efficiency and linkage handling.