Back to search

Boston University

Analysis of likelihood ratios for 3, 4, and 5-person mixtures from the PROVEDIt database using DNA-View Mixture Solution

Abstract

dc:description.abstract

Through advancements in the sensitivity of DNA typing methodology, increasingly smaller amounts of DNA as well as damaged DNA have been able to be detected. As a result, there has been an increase in complex DNA mixtures. Many of these mixtures are considered indistinguishable by traditional manual interpretation methods as a result of their number of contributors or contributor ratios. The interpretation of these mixtures becomes further complicated by stochastic effects such as stutter, drop-out, drop-in, peak height imbalance, and further by the effect of allele sharing which becomes increasingly prevalent as the number of contributors to a mixture increases. Therefore, over the last several years probabilistic genotyping softwares which combat and target these issues have been rising in popularity. These systems use models and algorithms to analyze DNA profiles and produce likelihood ratios which have become attractive for their ability to approach to mixture interpretation with much less subjectivity. One such software is DNA-View Mixture Solution, which makes use of stochastic modeling to analyze mixtures and directly compute likelihood ratios. Many laboratories choose not to analyze complex mixtures with more than four contributors due to the challenges that these mixtures present. However, these mixtures may contain useful information and produce LR’s with substantial evidentiary weight. For this reason, this study has analyzed and produced likelihood ratios for 3, 4, and 5-person mixtures from the PROVEDIt database using DNA-View Mixture Solution. The goal being to relate variable effects on LR by using a population database in which, at each locus, the frequency of all alleles are equal. This study additionally considered estimates of number of contributors as well as proportions, and how these aspects may be benefited by defining a known contributor within the hypothesis. All hypotheses considering contributors as the sole POI with no assumed contributors produced LRs with strong support for the prosecution hypotheses. It was determined that as the NoC to the mixture increased, LRs generally decreased and estimates for NoC as well as proportion estimates became less accurate. However, when the program was assisted, especially with a known minor contributor, estimates for NoC and proportions improved and LRs increased drastically. Lastly, as the ratio of major to minor contributors became larger, LRs of major contributors generally increased, while LRs of minor contributors stayed consistent or decreased.

Degree

thesis:*
Name thesis:degree_name
Master of Science
Level thesis:degree_level
masters
Discipline thesis:degree_discipline
Biomedical Forensic Sciences
Grantor
Boston University
Year dc:date.issued
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Roling, Madison Mackenzie
Advisor dc:contributor.advisor
  • Cotton, Robin W.

Subjects

dc:subject × 1

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/51831

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
related terms
citation

Roling, Madison Mackenzie. Analysis of likelihood ratios for 3, 4, and 5-person mixtures from the PROVEDIt database using DNA-View Mixture Solution. masters thesis, Boston University, 2024. https://hdl.handle.net/2144/51831