Inferring the number of contributors to mixed DNA profiles

Area of Science:

• Forensic Science
• Genetics
• Statistical Analysis

Background:

• Forensic DNA analysis often encounters complex mixtures from multiple individuals.
• Accurately determining the number of contributors is crucial for reliable interpretation.
• Uncertainty in contributor number diminishes the statistical significance of DNA evidence.

Purpose of the Study:

• To introduce a probabilistic mixture algorithm for analyzing forensic DNA samples.
• To enhance the interpretation of DNA mixtures with an unknown number of contributors.
• To improve the statistical weight of evidentiary samples in forensic investigations.

Main Methods:

• Developed a probabilistic mixture algorithm considering allele number, magnitude, and frequency.
• The algorithm computes probabilities of allele presence for various contributor hypotheses (e.g., 2-person, 3-person mixtures).
• Likelihood ratios are derived from the ratio of computed probabilities to assess contributor hypotheses.

Main Results:

• The algorithm provides a quantitative method to assess the likelihood of different contributor numbers.
• It allows for the calculation of likelihood ratios for hypotheses regarding the number of contributors.
• This approach offers a more robust statistical framework for interpreting complex DNA mixtures.

Conclusions:

• The probabilistic mixture algorithm significantly aids in interpreting forensic DNA mixtures.
• It addresses the challenge of determining the number of contributors, enhancing evidence reliability.
• This method strengthens the statistical foundation for forensic DNA mixture analysis.