SMART: STR Mixture Analysis and Resolution Tools

Xianchao Ji,Lianjiang Chi,Lan Wu,Jianchao Chen,Anxin Yan,Yongjiu Li,Zheng Tu,Jian Ye,Hua Chen
DOI: https://doi.org/10.1016/j.fsigen.2024.103148
Abstract:The analysis of STR mixture profiles derived from mixed DNA samples plays a critical role in criminal investigations and legal proceedings. In this article, we present SMART, a novel software developed within the fully continuous model framework to analyze STR mixture profiles. SMART incorporates the peak height model, stutter model, drop-in/drop-out model, and population genetics model, offering various functionalities such as calculating likelihood ratios (LR), resolving genotypes of individual contributors, and performing direct database searches using mixed DNA profiles. The performance of SMART was evaluated using laboratory-generated samples and the PROVEDIt dataset following the SWGDAM guidelines. The results demonstrate that SMART achieves high sensitivity, specificity, and precision. Furthermore, the software is computationally efficient, allowing for quick analysis on a desktop computer. Overall, we anticipate that SMART will serve as an invaluable tool for forensic investigations, enhancing the accuracy and reliability of criminal justice outcomes.
What problem does this paper attempt to address?