SHAMAN: a user-friendly website for meta-taxonomic analysis from raw reads to statistical analysis

Profiling differences in microbiome composition regarding a healthy or a disease phenotype is central in microbiome research. It typically involves the sequencing, the data processing, the statistical analysis and the graphical representation of the detected signatures. This process is currently addressed through several separated applications that requires expertise for installation, data pre-processing and in some case, programming skills. We present SHAMAN, which is an interactive web application that facilitates the use of a workflow for OTU calculation, the robust statistical modelling proposed by the DESeq2 package and provides more than ten interactive visualisations. It also implements new normalization methods which account for the sparsity of metagenomic data. %which is relevant for the DESeq2 approach.SHAMAN is primary designed for biologist with no experience in metagenomics, but can also meet the needs of specialist who can easily access to the configuration of processed tools and packages. It is freely accessible at http://shaman.pasteur.fr/, but also work as a standalone application with docker (aghozlane/shaman) or in R with packrat. The source code is written in R and is available at https://github.com/aghozlane/shaman. Using a mock community sequencing and a published 16S metagenomic dataset, we show here that SHAMAN allows to perform a complete meta-taxonomic analysis.