Skip to content

mia

Microbiome analysis — tools for microbiome analysis based on the SummarizedExperiment, SingleCellExperiment, and TreeSummarizedExperiment infrastructure; focuses on data handling and analysis for taxonomic information including community indices and data summarization.

Workshop at EuroBioC2025

  • Presenter: Tuomas Borman (University of Turku)
  • Day / session: Day 2 (Thu Sep 18, 2025) — afternoon hands-on workshop (parallel with OSTA, msqrob2, Workshop Platform, iSEE)
  • Workshop title: "Orchestrating Microbiome Analysis with Bioconductor"
  • Format: hands-on tutorial, attendees follow vignettes in their own R session
  • Workshop materials: to verify — typically a per-event GitHub repo from the microbiome group
  • Video: workshops not recorded; the Bioconductor YouTube playlist covers plenaries only
  • Abstract / schedule: eurobioc2025 schedule

What it does

mia wraps microbiome count and taxonomy data into a TreeSummarizedExperiment so that taxonomic hierarchy, sample metadata, and feature counts live in one S4 object compatible with the rest of Bioconductor. On top of that substrate it provides community indices (alpha/beta diversity, dominance, divergence), data transformation utilities, and per-rank aggregation — the layer most downstream microbiome analyses (miaViz, miaTime, miaDash) sit on.

How it works

Core idea. Use TreeSummarizedExperiment as the canonical container: feature counts, sample metadata, taxonomic hierarchy, and a phylogenetic tree all live in one S4 object, so every downstream operation (diversity, agglomeration, transformation, ordination) is just a method on that object. This collapses the historical fragmentation across phyloseq / QIIME2 / BIOM into one substrate compatible with the rest of Bioconductor.

Inputs / outputs. Input: count tables + taxonomy + sample metadata, imported from phyloseq objects, QIIME2 outputs, DADA2 results, or BIOM files via dedicated convert* functions. Output: a TreeSummarizedExperiment carrying the data plus any computed indices (alpha/beta diversity, dominance, evenness) stored in colData or alternative experiments.

Key innovation. agglomerateByRank / agglomerateByRanks produces per-rank aggregated assays (Kingdom → Species) and stores them as alternative experiments, so subsetting samples stays synchronized across ranks. transformAssay and rarefyAssay apply normalizations (log + pseudocount, CLR, rarefaction) without leaving the container.

Parameters worth knowing. Rank choice for agglomeration (Genus and Species are the most common). Diversity index family (addAlpha for Shannon and friends; getDissimilarity for Bray-Curtis; getJSD, getUniFrac, getDPCoA for tree-aware beta-diversity). Transformation choice in transformAssay (log, relative, CLR). Rarefaction depth in rarefyAssay.

Canonical example. The vignette uses the GlobalPatterns dataset — 19,216 taxa across 26 samples with seven taxonomic ranks — as the running example, supplemented by the enterotype and esophagus datasets for specific operations. Diversity, agglomeration, and transformation are demonstrated on this object end-to-end.

Where it fits in the corpus

  • AACR 2026: no direct dossier; microbiome analyses surface in tumor-microbiome and gut-microbiome / immunotherapy talks
  • miaTime (entry) — extends mia for longitudinal microbiome studies
  • ISMB 2026: scaffold
  • GBCC 2025: queued — mia workshops are recurring fixtures
  • Nextflow Summit 2026: not mentioned (different ecosystem)

Notes

mia is the canonical Bioconductor microbiome data class. Workshop substrate is the OMA (Orchestrating Microbiome Analysis) book also from this group, which is the de-facto entry point for new users.