A computational pipeline for the analysis of the role of proteins from unknown species in shotgun metagenomics data: application to the study of gut microbiota in a mouse model of Alzheimer's disease.

The aim of this computational project is to investigate from a functional point of view the protein repertoire of genomes deriving from shotgun metagenomics experiments. We will specifically focus on the study of microbiota from a mouse model of Alzheimer's disease, with the ultimate goal of elucidating the role of proteins inferred by reconstructed genomes of unknown species. Nowadays, shotgun metagenomics allows the comprehensive reconstruction of the set of genomes present in a certain biological sample. Most of the currently available bioinformatics tools are mainly focused on the characterization of reconstructed genomes in terms of the gene repertoire, taxa identification and classification. However, relatively little was developed so far for the characterization of the set of proteins annotated on reconstructed genomes, especially in the context of neurodegenerative diseases such as Parkinson's and Alzheimer's disease. This project relies on a set of considerations and preliminary results, among which: 1) we previously investigated the role of the gut microbiota in a mouse model of Alzheimer disease, through a metabolo-genomic approach involving a 16S sequencing (Favero et a, 2022). On this mouse model, we recently performed a complete metagenomic shotgun sequencing. 2) we previously developed a preliminary version of a bioinformatic pipeline able to characterize bacterial taxa reconstructed from shotgun sequencing experiments with a meta-proteomic functional characterization of their potentially expressed proteins. This pipeline was tested investigating a large publicly available metagenomic dataset from Parkinson patients and controls (Favero et al., BITS 2022 Conference, Verona, Italy). We showed that, at least in this case study, functional investigation of proteome associated to newly reconstructed genomes potentially lead to the identification of new pathways involved in the considered disease, comparing patients and controls. Stemming from these considerations, the purpose is this proposal is to extend our previous expertize as follows: 1) Rebuild and automate a bioinformatic pipeline for the annotation of the protein repertoire on reconstructed genomes from metagenomic shotgun data. 2) Apply the pipeline to investigate the functional annotation of proteome from metagenomics shotgun data already produced, relative to the microbiota characterization for a mouse model of Alzheimer's disease. The final aim is to analyze the differences in terms of protein domains and relative functions of proteins associate specifically with genome of known taxa with respect to the ones associate to genomes of unknown species, comparing control mices and disease-modeling mices. Overall, our hypothesis is that genomes of unknown species from metagenomics data are carriers of specific sets of proteins and relative metabolic pathways, not enriched in known species, that could have implications for therapeutic interventions.