Abstract
The field of forensic science is a unique collaboration between science and the court of law, providing scientific findings that can be applied in a legal context, often in the pursuit of criminal investigations. Implementation of novel scientific methodologies to the field of forensic science can be hindered by this, with legal requirements of reliability and reproducibility necessary for forensic evidence to be deemed admissible. CRISPR typing analysis, applied in a forensic context, has the potential to provide additional relevant information that may aid investigations associated with microbial evidence. Current literature suggests a lack of research into the potential applications of CRISPR typing analysis in a forensic context. The purpose of this thesis is to determine if CRISPR typing analysis can produce scientific findings that are both reproducible and reliable enough for inclusion in a forensic investigation, and if such findings provide rationale to support future investigation into the application of CRISPR typing analysis in a forensic context. Current and emerging applications of CRISPR typing analysis are also considered, with a focus on where literature is lacking and could benefit from future research.
In order to determine the reproducibility and reliability of CRISPR typing analysis, in silico investigation of the diversity of two forensically relevant genus, Yersinia and Salmonella, was undertaken. This investigation included the comparison of two established subtyping methodologies, 16S rRNA and rMLST subtyping, to CRISPR typing analysis. Significance of findings is determined by the discriminatory power of the methodology to provide unique identifications of genomes with similar characteristics. 16S rRNA subtyping was unable to provide significant diversity to differentiate between genomes of the same genera in both Yersinia and Salmonella. rMLST subtyping was able to provide significant enough diversity to
differentiate between Yersinia spp. and between Salmonella enterica subsp. enterica serovars in most cases, with most genomes grouped into large singular subtype groups and then identified as individually unique. Overall, CRISPR typing analysis was able to differentiate these genomes to higher degree of discrimination, although at an individual level variation between Yersinia spp. and between Salmonella enterica subsp. enterica serovars is noted. In particular, the absence of a detectable CRISPR-Cas system in Yersinia enterocolitica makes the methodology impossible to undertake, however both Yersinia pestis and Yersinia pseudotuberculosis can be identified to a significant degree of discrimination, with all 80 genomes identified at an individual level in comparison to the 23 genomes identified at an individual level by rMLST subtyping. In the Salmonella dataset, Salmonella enterica subsp. enterica serovars Dublin, Enteritidis, Paratyphi A and Typhi are shown to have little diversity in their CRISPR characteristics, with rMLST subtyping providing greater discriminatory power than CRISPR typing analysis. The opposite is true for Salmonella enterica subsp. enterica serovars Heidelberg, Newport and Typhimurium, however.
Considerations of the diversity of CRISPR characteristics, and then inherent repetitive nature of CRISPR arrays, is lacking in current literature. This lack of consideration has been addressed, with some observations made surrounding potential cases such as the limitations of CRISPR typing analysis due to the inconsistent presence of CRISPR-Cas systems, and the presence of currently established methodologies. Additionally, experimental data from this thesis is used to support the concept that any future application of CRISPR typing analysis in a forensic context is likely to be more complementary, to reduce the potential risks associated with the lack of reproducible discriminatory power across and within all genera.