Internal validation of STRmix™ - A multi laboratory response to PCAST.

Area of Science:

• Forensic Science
• Genetics
• Biotechnology

Background:

• Probabilistic genotyping software like STRmix™ is essential for complex DNA mixture interpretation in forensic science.
• Internal validation studies are critical for assessing software performance and reliability across diverse forensic casework scenarios.

Purpose of the Study:

• To compile and analyze internal validation data for STRmix™ probabilistic genotyping software from multiple forensic laboratories.
• To confirm previously observed trends in likelihood ratio (LR) performance and investigate potential confounding factors.

Main Methods:

• Compilation of 2825 DNA mixtures (3-6 donors) from 31 independent laboratories.
• Analysis of data encompassing various multiplex systems, equipment, mixture proportions, and template DNA amounts.
• Evaluation of likelihood ratio (LR) trends associated with low template DNA and high contributor numbers.

Main Results:

• Confirmed previously reported trends of reduced discriminatory power (lower LRs) for both donors and non-donors under low template DNA conditions and with increased contributor numbers.
• No isolated effect of allelic sharing on LR discrimination was identified.
• Apparent effects of allelic sharing were found to be confounded with the number of contributors.

Conclusions:

• The large-scale validation reinforces the reliability of STRmix™ in forensic DNA mixture analysis.
• The findings highlight the importance of considering template DNA quantity and contributor number when interpreting STRmix™ results.
• Further investigation is needed to fully understand the influence of allelic sharing, potentially confounded by contributor number.