The evaluation of metagenomic analysis software, using in-silico and in-vitro mock community datasets, for the accurate study of bio-aerosol samples.

Abstract

The bio-aerosol is an important medium for the potential dispersal of biological warfare agents within the battlefield space. In order to better protect the military personnel who work within this environment it is imperative that we increase our understanding of this matrix, especially the naturally occurring variation and its causes. Understanding the naturally occurring variation within the bio-aerosol will enable future and current biological detection platforms to be put through better test and evaluation processes, thus reducing the potential for false alarms and false negatives. Analysing bio-aerosol samples collected across a temporal gradient through a metagenomics approach will enable the natural variation to be better understood. However, metagenomic analysis tools have been shown to have contradictory reviews within the literature, it is therefore essential to identify the most suitable analysis approach. Here I developed a metagenomic analysis pipeline which delivers high confidence taxonomic identification to species level, as well as accurate measures of diversity and homogeneity. The analysis pipeline that was developed takes the output from multiple tools thus reducing the number of false positives, delivering high confidence taxonomic identification. The analysis pipeline also gives a more accurate measure of diversity and homogeneity compared to any of the tools being used individually. This improved accuracy will deliver superior results when measuring the change in abundance of species identified within the bio-aerosol in sampling regimes carried out at Dstl. These improvements will lead to more accurate test bio-aerosols being developed for biological detection platform evaluation. Fundamentally this will improve the UK military’s capability to detect biological warfare releases within the battlespace.