Cotton, Robin W.Wade, Elizabeth Ann2023-02-282023-02-282022https://hdl.handle.net/2144/45684As the sensitivity of DNA testing improves, analysts are seeing more and more mixtures that are difficult to handle and time-consuming to interpret. The development of probabilistic genotyping softwares, such as STRmix™ (STRmix Ltd, Auckland NZ) have improved the mixture interpretation process by reducing time input and completing a thorough examination of all data with higher accuracy than that of a human being. However, a significant limitation of STRmix™ is its requirement to estimate the number of contributors (NOC) in the mixture before mixture deconvolution can proceed. This process can be labor-intensive and time-consuming. This process is also complicated by factors such as allele sharing, allele dropout, and stutter. An incorrect estimation of the NOC can have negative effects downstream, such as misrepresenting the likelihood ratio (LR). Thus, it is essential that the field of forensic DNA analysis implements an accurate approach for estimating the NOC in a DNA mixture. The software systems FaSTR™ (STRmix Ltd, Auckland NZ) and NOCIt™ (LFTDI, Camden NJ) were introduced into forensic literature as a tool for estimating the NOC. FaSTR™’s performance was assessed with 70 mixture samples, which included two-person, three-person, and four-person mixtures over a range of contribution ratios. NOCIt™’s performance was first assessed with 70 mixture samples, then further evaluated with 36 additional mixture samples. All the mixture samples assessed with NOCIt™ consisted of two-person, three-person, and four-person mixtures over a range of contribution ratios. At the conclusion of the evaluation, it was determined that FaSTR™ and NOCIt™ performed similarly to one another. The overall accuracy and mixture-specific accuracies were comparable: there was higher accuracy in the two-person mixtures and lower accuracy in the four-person mixtures. Thus, both software systems have the potential to improve current laboratory workflow by significantly reducing analyst intervention in the NOC estimation process.en-USAttribution 4.0 Internationalhttp://creativecommons.org/licenses/by/4.0/BiologyDNA mixturesFaSTRForensic DNANOCItNumber of contributorsProbabilistic genotypingThesis/Dissertation2023-02-22