Due to copyright issues, an electronic copy of the thesis must be ordered from the faculty. For the faculty to have time to process the order, the order must be received by the faculty at the latest 2 days before the public defence. Orders received later than 2 days before the defence will not be processed. After the public defence, please address any inquiries regarding the thesis to the candidate.
Trial Lecture – time and place
See Trial Lecture.
Adjudication committee
- First opponent: Associate Professor Andreas Tillmar, National Board of Forensic Medicine and Department of Forensic Genetics and Forensic Toxicology, Sweden
- Second opponent: Senior Lecturer David Ballard, King’s College London, UK
- Third member and chair of the evaluation committee: Professor Bendedicte Alexandra Lie, University of Oslo
Chair of the Defence
Professor Lars Eide, University of Oslo
Principal Supervisor
Researcher Øyvind Bleka, Oslo University Hospital
Summary
Forensic marker genotyping has improved over the last two decades following developments of multiplex DNA amplification panels. In parallel, probabilistic genotyping software has evolved, advancing the interpretation of DNA evidence. Capillary electrophoresis (CE) is currently used for forensic DNA typing, but massively parallel sequencing (MPS) has emerged as an alternative. MPS analyses of forensic samples generate more complex typing data compared to CE, providing allele length and sequence. Data require proper characterisation so that meaningful likelihood ratio (LR) values can be provided. Analysis of artefacts like stutters is crucial, since incorporating this information as part of the probabilistic modelling provides more robust LR results.
Through this PhD, we have laid the foundations for a new probabilistic software tool called MPSproto. Stutter sequences from MPS data were characterized using a beta regression model, which explained stutter proportion variation by the parental uninterrupted stretch (PTUS). Regression parameters were used in MPSproto to calculate expected stutter sizes in evidence profiles. LR results from complex mixtures obtained from surface DNA were compared between MPSproto (MPS data) and EuroForMix (CE data), and we found that CE data showed higher values. Performance of MPSproto using gamma and negative binomial models was examined and the latter yielded larger LRs. Allele frequencies for the Southern Norway dataset were submitted to the global database STRidER, showing no significant differences from Northern Norway and Denmark datasets. This indicated that a single database could be used for all areas. Genetic diversity studied using FST showed no population differences.
To conclude, this thesis provides insights into important features of stutters in MPS data. The allele frequency for the Norwegian population is publicly available through STRidER. We have demonstrated that MPSproto is a solid tool for calculating LRs with MPS data.
Additional information
Contact the research support staff.