Skip to content

MIRit

Integrate microRNA and gene expression to decipher pathway complexity — a Bioconductor framework that investigates relationships between miRNAs and genes across biological conditions, supporting differential expression analysis and network characterization between miRNA and mRNA expression.

  • Maintainer: Jacopo Ronchi (University of Milano-Bicocca, UniMiB) — jacopo.ronchi@unimib.it
  • Bioconductor: MIRit v1.8.0
  • Source: git.bioconductor.org/packages/MIRit
  • Vignettes: MIRit
  • License: GPL (>=3)
  • Status at EuroBioC2025: methods talk on miRNA-mRNA network integration

Talk at EuroBioC2025

  • Speaker: Jacopo Ronchi (University of Milano-Bicocca)
  • Day / session: Day 1 (Wed Sep 17, 2025) — afternoon short-talk session, after the Holmes keynote
  • Talk title: "MIRit framework for miRNA-mRNA network identification"
  • Slides (Zenodo DOI): to verify after publish
  • Video: not recorded — short talks at EuroBioC2025 were not part of the plenary YouTube playlist
  • Abstract / schedule: eurobioc2025 schedule

What it does

MIRit jointly models miRNA and mRNA expression to recover regulatory relationships across biological conditions. It performs differential expression analysis on both modalities and characterizes miRNA-mRNA interaction networks, so that pathway-level signal is interpreted through the regulatory layer rather than from gene expression alone.

How it works

Core idea. A MultiAssayExperiment-based container (MirnaExperiment) holds matched miRNA and mRNA expression matrices; MIRit runs differential expression on each modality, retrieves predicted and validated miRNA-target interactions, and integrates the two sides — via correlation analysis for paired samples and via association tests / rotation gene-set tests for unpaired designs — to flag miRNA-target pairs whose joint dysregulation is consistent with regulatory action.

Inputs / outputs. Input: a gene-expression matrix (HGNC gene symbols × samples), a miRNA-expression matrix (miRBase miRNA names × samples), and a sample metadata data.frame keyed by subject ID with miRNA/gene column mappings and a condition variable. Output: a MirnaExperiment annotated with per-modality DE results, retrieved target interactions, integrative significance per miRNA-target pair, and enrichment results (ORA, GSEA, CAMERA).

Key innovation. Wraps the miRNA-target integration workflow — DE on both layers, target lookup, paired-vs-unpaired integration, enrichment, and SNP-disease association at miRNA loci — into a single Bioconductor container, rather than leaving it as a series of bespoke scripts. The specific DE algorithm and target databases are not explicitly named in the introduction vignette — not specified in vignette excerpt; needs talk slides.

Parameters worth knowing. pairedSamples in the MirnaExperiment constructor — switches the integrative test between correlation (paired) and association/rotation gene-set tests (unpaired). primary / mirnaCol / geneCol columns in the metadata data.frame — the join keys for the two matrices.

Canonical example. Vignette uses an RNA-Seq thyroid-cancer dataset from Riesco-Eizaguirre et al. (2015): 8 papillary thyroid carcinoma tumors with matched contralateral normal thyroid tissue from the same patients. Workflow: build the MirnaExperiment with pairedSamples = TRUE, visualize with plotDimensions() (MDS by condition), run DE on each modality, retrieve targets, then integrate to recover miRNA-gene pairs whose paired dysregulation is consistent with target repression.

Where it fits in the corpus

  • AACR 2026: no current dossier; relevant axis is bioinfo / AI methods — network-based miRNA-mRNA integration is conceptually adjacent to MOSClip's pathway-based multi-omics
  • MOSClip (entry) — sibling pathway/network method on the same Day 1 short-talk session
  • GBCC 2025: queued
  • Nextflow Summit 2026: not mentioned

Notes

miRNA-mRNA integration sits in the network-based pathway analysis tradition; complement to MOSClip (entry) on the same session, with MIRit emphasizing the regulatory-network framing rather than survival.