Developing strain-level resolution metagenomic methods to profile the microbiome

Microbes exist all around us and take part in shaping the world as we know it. Invisible to the naked eye, they co-inhabit all types of environmental niches and create vast and complex communities, termed microbiomes. They are essential for life on earth, where they play a central role in shaping the ecosystems and have a great impact on human health. The interplay of the microbes and our health is directly linked to the specific composition of the microbiome. To understand their impact it is crucial to be able to identify the finest-possible granularity, moving from identification at species-level to strain-level resolution.
By applying metagenomics this becomes possible. Metagenomics is the study of DNA extracted directly from the environment, bypassing the need for cultivation of the microbes. With this approach the entire genetic content of the microbes are analysed, enabling strain-level analysis. However, due to the complexity and variability found within the microbial world, this is a task that remains unsolved.
In this thesis efforts have been made to develop strain-level resolution metagenomic methods for accurately profiling the microbiome. In the first published work we proposed a method for selecting a set of signature genes, which can be used for accurate identification and abundance estimates of the bacteria found within the microbiomes. As the signature genes are unique for each biological entity, they can be used to profile the microbes even at very low abundance.
For the second project in this thesis, we use the signature genes in single nucleotide variant analysis, which facilitates sub-species level identification. Through this project we created the bioinformatic tool MAGinator, which enables de novo quantification and taxonomic annotation of the microbes found within the metagenomics sample. Through a combination of both gene- and contig-based techniques it offers insights into the genetic and functional content along with the bacterial origin.
Subsequently we explored the antimicrobial resistance gene (ARG) profiles of young adults and infants, to determine differences and identify the specific bacteria harbouring them. The analysis revealed that bacterial composition, especially Escherichia coli, critically influences the ARG profile. Specific ARG clusters were identified and linked with certain strains of Escherichia and Bifidobacterium, highlighting the importance of strain-level identification.
The final project reported in this thesis investigated the spread and diversity of the opportunistic pathogen Pseudomonas aeruginosa across the globe. The results revealed no evolutionary differences in the genomes across different environmental niches. The metabolites produced by the microbes varied between the environments, however it remains to explore if this can also be found for the metabolites specific to Pseudomonas aeruginosa.
As a whole, the presented work has covered methods for strain-level analysis of the microbiome. Being able to identify strains opens a door to understand the interplay between the microbes, and also the effects that they have on the environment they occupy.