eudysbiome User Manual

Large amounts of data for metagenomics, especially the earliest studies on 16S ribosomal RNA gene, are produced by high-throughput screening methods. These are processed in the form of quantitative comparisons (between two microbiomes’ conditions) of reads’ counts. Reads’ counts are interpreted as a taxon’s abundance in a microbial community under given conditions, such as a medical treatments or environmental changes. The comparative analysis of such microbiomes with a baseline condition permits to identify a list of microbes (classified in species, genus or higher-order taxa) that are differential in abundance among the conditions. The taxonomic classification of 16S rRNA sequences is generally dependent on two strategies to assign sequences into populations: one is the phylotype-based method that assigns sequences to taxa based on the similarity to a reference database; the other is the operational taxonomic unit (OTU)-based method which does not rely on the association to known taxa, but consists of the definition of clusters of sequences sharing high similarity among them (typically a 97% identity threshold). These clusters are identified by an OTU ID, and where possible associated to a species. The phylotype-based method limits the taxonomic classification of novel sequences from previously unknown taxa and is less sensitive to the sequencing errors[1]. The OTU-based method overcomes the limitations of phylotype-based method and permits the classification of OTUs down to the species level, as typically an OTU is thought of as representing a species, and for this the OTU-based method necessarily relies on a reference data set including species. eudysbiome applies the OTU-based classification by mapping OTUs’ unclassified representative sequences (produced via microbiome analysis pipelines such as emphMothur[2] or QIIME [3] during the generation of the OTUs) to a selection of classified representative sequences of OTUs, obtained from clustering the truncated SILVA[4] Small subunit (16S/18S, SSU) ribosomal RNA (rRNA) reference dataset at 97% similarity. This representative dataset is built within QIIME and included in our package. In this mapping, each entry in the representative dataset is associated to a taxonomic path containing the taxonomy rank designations from kingdom to species. In order to translate the GI microbiome variations into potential clinical interpretation, it is helpful to assess whether the variation leads towards a microbiome that is more or less synergistic with the host, i.e. more or less pathogenic. To achieve this goal, in addition to the standard taxonomic annotation (classification, from now on) described above, this package annotates species as harmful/harmless based on their ability to contribute to mammals’ host diseases (as indicated in literature) and hence based on their pathogenic potential (annotation from now on). Several assumptions are needed to achieve this goal, due to the quickly growing,