ETD

• human metagenomics
• Oxford Nanopore Technologies

Oxford Nanopore Technologies (ONT) enables long-read sequencing of DNA, without the need for PCR during library preparation, avoiding amplification bias, an important advantage in metagenomics. Additionally, Adaptive Sampling (AS) allows real-time enrichment of microbial DNA by depleting host DNA during sequencing. However, the lack of standardization may limit the broader use of this technology. The aim of this thesis was to develop and optimize, through the comparison of different wet-lab and bioinformatics approaches, a workflow for ONT-based human metagenomics, from DNA extraction to taxonomic assignment of the oral microbiome species, followed by a preliminary diversity analysis. DNA was extracted from saliva samples of individuals with pancreatic diseases. Singleplex sequencing (one sample per sequencing run) was initially tested; then, sample multiplexing was implemented. Raw sequencing data were processed in the computational phase. Taxonomic assignment and abundance estimation of the species were carried out respectively with Kraken2 and Bracken, comparing different approaches. Following the generation of an abundance table, a preliminary diversity analysis was performed. AS increased the number of microbial reads from 298,451 (without AS) to 720,029 (with AS) by sequencing the same sample, corresponding to a yield increase of 141.3%. The optimizations introduced throughout the study reduced DNA loss during library preparation, enabling the reduction of DNA input for libraries. The best taxonomic assignment performance was achieved using standalone Kraken2 and Bracken, and a database including NCBI collections of Bacteria, Archaea, Fungi, Protozoa, Virus and plasmids, obtaining an average of 94.1% of classified reads per-sample. The resulting workflow can be applied to a wide range of ONT-based human metagenomics studies. Notably, the bioinformatic pipeline is largely automated and modular, allowing its application to any dataset generated using ONT sequencing, by adjusting a limited number of parameters.