Abstract
Advances in forensic science over the last 25 years have enabled trace amounts of material to be analysed in ever more complex ways. In the field of forensic genetics, this has led to the routine production of DNA profiles from only a few cells of damaged human DNA, providing a direct link between an individual and a crime scene. The human DNA present within any such sample may however only represent a fraction of the total DNA present, and metagenomics is the field of study that focusses on understanding the diversity and implications of this other world of environmental DNA that coexist in the sample. Investigating the non-human DNA (bacteria, archaea, protists, plants, fungi, and animals) within such samples could provide useful intelligence to progress stalled investigations, including narrowing down the suspect pool. Depending on the sample type and the question at hand, this could include activity level information such as where someone might have been and what they may have come in to contact with before, during, and even after a crime event has taken place.While multiple studies have investigated the application of microbial and metagenomic analyses in a range of different fields, many have only focused on single taxonomic groups of interest, using targeted sequencing approaches. Additionally, there has been limited quality control to ensure that these methods are fit-for-purpose for forensic application, with taxonomic classification errors common in published literature.
The primary aim of this doctoral research was to develop a method that was able to elicit the full taxonomic composition from forensically relevant samples while ensuring that accurate and reproducible results suitable for use within a criminal setting were obtained.
To achieve this, a novel massively parallel sequencing method incorporating a sequence independent random priming strategy was developed. In addition, in depth studies into the background taxa within the laboratory setting and the laboratory processing stages were conducted. The purpose of this was to establish the effect background taxa had on results, and to construct robust procedures with the aim of eliminating these sources of contamination and where this was not possible, minimising their presence and constructing a corrective strategy to manage this.
To enable thorough investigation of the above, bacteria within two sample types; touch deposits and saliva were studied. To further investigate the suitability of this method, it is recommended that future work investigates taxa from all taxonomic Kingdoms, and that the method is applied to other forensically relevant sample types.
One key finding from this research is that ubiquitous taxa is observed within negative controls and samples, originating from background contamination, and those samples with low biomass (low total quantity or weight of organisms within a given area or volume) are more susceptible. Additionally, the work has highlighted the many factors that can influence the results obtained; this poses a significant risk to the consistency and accuracy of the results produced and consequently any downstream comparative analysis that might be conducted. Key areas include changes to the sample processing method and the selection of data analysis platforms, including bioinformatic pipeline settings and the databases used.
Recommendations from this research include the need for negative controls to be run alongside true samples and for these negative controls and samples to be run in (at least) triplicate to enable comparison and to account for intra-sample variability. Overall, this work has highlighted the need for due care and strategic planning when processing samples and analysing the results obtained, to include the use of controls, consistent standard operating procedures, and databases to ensure that accurate, comparable, and reliable results suitable for use within a criminal setting are produced.
The method developed within this doctoral research is ready for validation to commence and once complete it can be added to the forensic toolkit with the aim of being implemented into criminal casework.
Date of Award | 1 Dec 2022 |
---|---|
Original language | English |
Awarding Institution |
|
Supervisor | Denise Syndercombe-Court (Supervisor) & David Ballard (Supervisor) |