Microbiome analysis workflow
Karelle Rheault 
Davide Barsotti
Johan P.B. Jørgensen
This workflow was created for microbiome analyses within the Silva Nova project, to ensure that all publications related to the Silva Nova project use the same data processing, and avoid inconsistencies between sub-projects. This workflow should also make microbial analyses more accessible and streamlined.
Decisions for each step of the workflow are supported by literature and R scripts, which aim to answer the following questions:
- How to handle negative and positive controls?
- Should rare and scarce species be removed? If so, what are the thresholds?
- After taxonomy assignation, which taxa should be kept or removed? What are the quality threshold for species identification with BLAST?
- When should we rarefy? What other normalization methods should be used?
- Which diversity metrics (alpha and beta) can be used with microbiome data?
Structure
In the navigation menu of this webpage, you will find each step of the workflow:
Get started
- Install R and required packages: get started with RStudio and install the packages required for the analyses in this workflow
- Data preparation: prepare OTU tables, manage metadata, and convert raw files to .RData
Handling OTU tables
- Controls: deal with OTUs found in negative and positive controls
- Filtering: filter out OTUs based on abundance and prevalence, or based on taxonomy and BLAST quality
- Rarefaction and normalization: rarefy or normalize data to control for variable sequencing depth
Statistical analyzes
- Alpha diversity: choose the best alpha diversity metrics to answer your hypothesis
- Beta diversity: choose the right transformation and the right distance matrix for beta diversity analyses
- Differential abundance: perform statistical tests to find differential abundance of taxa
Download all scripts and data
Click on the buttons below to download all pre-made scripts and example raw data sets used with the examples on this website.