Exploratory Microbiome Data Analysis and Visualization • microEDA

Overview

microEDA is an R package for exploratory data analysis of microbiome data, designed to answer the fundamental question: “Who is in my samples?”

In addition to core visualization functions, the package provides some utility functions to detect common data issues.

Installation

The development version of microEDA can be installed from GitHub.

Tip: The Introduction to microEDA vignette is available online and can be viewed without installing the package.

Option 1: Using `pak` (Recommended)

This method is fastest and handles Bioconductor dependencies (like phyloseq) automatically. Note that vignettes are not included with this method.

if (!requireNamespace("pak", quietly = TRUE)) {
  install.packages("pak")
}
pak::pak("jrotzetter/microEDA")

Option 2: Using `remotes` (Include Vignettes)

Select this method to include package vignettes for offline viewing.

Important: Because microEDA depends on phyloseq (hosted on Bioconductor), please install phyloseq first to avoid dependency errors when using remotes.

# 1. Install Bioconductor dependency
if (!requireNamespace("BiocManager", quietly = TRUE)) {
  install.packages("BiocManager")
}
BiocManager::install("phyloseq")

# 2. Install microEDA with vignettes
if (!requireNamespace("remotes", quietly = TRUE)) {
  install.packages("remotes")
}
remotes::install_github("jrotzetter/microEDA", build_vignettes = TRUE)

Currently Implemented Features

microEDA-class: Extends the phyloseq-class to hold additional information.
microEDA-class summarisation: show(), summary() and show_filter_history() print overviews of microEDA objects.
MetaPhlAn support: Load and merge MetaPhlAn profiles using load_metaphlan() and join_mpa_profiles().
Feature filtering: filter_features() allows filtering by abundance and prevalence, either globally or within groups defined by a metadata variable (stratified filtering).
Taxonomic agglomeration: agglomerate_taxa() aggregates taxa at specified taxonomic ranks. This implementation is significantly faster than phyloseq::tax_glom().
Taxonomy table utilities: Functions to trim or add taxonomic prefixes (e.g., k__, p__) in a taxonomyTable.
Conversion to phyloseq: to_phyloseq() converts metaphlanProfile and microEDA objects into standard phyloseq objects.
Taxonomic consistency check: check_taxonomic_consistency() identifies inconsistencies within a taxonomyTable.
Presence lists: get_presence_list() returns unique taxa present in each group defined by a sample metadata variable.
Taxa overlap analysis: get_taxa_overlaps() computes overlaps and unique sets of taxa across sample groups, serving as an alternative to UpSet plots.

Visualization Functions

microEDA provides a collection of plotting functions, built on ggplot2, and designed for exploratory data analysis of microbiome data.

1. Taxonomic Composition Barplot

plot_taxa_barchart() visualizes abundance of taxa across samples.

Aggregates data at any taxonomic rank (e.g., Phylum, Genus).
Optionally Groups samples by metadata variables (e.g., treatment, disease state).

2. Mean Abundance & Prevalence Heatmap

plot_taxa_heatmap() displays a dual-metric view where each cell represents:

Mean abundance: Average taxon abundance within a group.
Prevalence: Proportion of samples in the group where the taxon is detected.

This dual metric helps distinguish consistently abundant taxa from those that are sporadically present.

3. Taxonomic Intersection UpSet Plot

plot_taxa_upset() visualizes shared and unique taxa across multiple sample groups.

Complemented by get_taxa_overlaps() for programmatic access to intersection data.

4. Taxonomic Flow Sankey Plot

plot_taxa_sankey() illustrates hierarchical taxonomic relationships from higher (e.g., Phylum) to lower (e.g., Species) taxonomic ranks.

Supports display of abundance flow for a single sample or mean abundance across a group of samples.

Example Usage

All functions are compatible with phyloseq objects, though the package internal microEDA-class provides some additional information and functionalities.

A microEDA object can be created by calling:

library(microEDA)
data(GlobalPatterns, package = "phyloseq")

me <- microEDA(GlobalPatterns)

In turn, a microEDA or metaphlanProfile object can be converted to phyloseq with:

ps <- to_phyloseq(me)

identical(GlobalPatterns, ps)
#> [1] TRUE

For more details please see vignette("microEDA") or the help pages in the documentation. Both are also available online at https://jrotzetter.github.io/microEDA/.

Planned features

Custom color palette to increase available distinct colors is planned

Citation

microEDA is a free, open-source project licensed under GPLv3. If you find these functions useful for your publications, presentations, or commercial workflows, please consider citing the package. This helps track the project’s impact and ensures proper credit for the development of these exploratory data analysis utilities.

BibTeX:

@Manual{microEDA,
  title = {{microEDA}: Exploratory Microbiome Data Analysis and Visualization},
  author = {Jérémy Rotzetter},
  year = {2026},
  note = {R package version 1.0.1},
  url = {https://github.com/jrotzetter/microEDA},
}

Acknowledgments

microEDA is built on top of the phyloseq infrastructure and leverages the visualization capabilities of ggplot2, ComplexUpset, and ggsankeyfier. Development of this package was made possible by the foundational work of the authors and maintainers of these packages.

Note: If you use microEDA in your research, please also consider citing the underlying packages you utilize, particularly phyloseq.

microEDA