IsoNetAnalyzeR

IsoNetAnalyzeR compares the predicted protein-interaction neighbourhoods of transcript isoforms. Given isoforms whose usage goes up and isoforms whose usage goes down, for example from a differential transcript usage analysis, it expands each set in an isoform-resolved predicted interaction network and compares the resulting neighbourhoods per gene and across genes. It also provides connected-component analysis, network diagnostics, plots and enrichment analysis.

Installation

IsoNetAnalyzeR requires R 4.6 or later. Until it is available from Bioconductor, install it from GitHub; build_vignettes = TRUE also builds the tutorial:

install.packages(c("BiocManager", "remotes"))
BiocManager::install("IsoformAnalysisGroup/IsoNetAnalyzeR",
                     dependencies = TRUE, build_vignettes = TRUE)

Once it is available from Bioconductor:

BiocManager::install("IsoNetAnalyzeR")

Getting started

The vignette walks through a complete analysis of isoform switches between obese type 2 diabetic and lean adipose tissue (GEO GSE141432), using example data shipped with the package:

library(IsoNetAnalyzeR)
vignette("IsoNetAnalyzeR", package = "IsoNetAnalyzeR")

For your own data, prepare a table with gene_id, isoform_id and direction ("up" or "down") and follow the user instructions. Statistical seed selection happens upstream; network differences are predictions to interpret alongside independent biological evidence.

Documentation

• Vignette source
• User instructions
• Runnable synthetic example
• Network eligibility rationale
• Data provenance and data scripts
• Version history

Function reference: help(package = "IsoNetAnalyzeR").

Licence

MIT; see LICENSE.md. The bundled interaction network is derived from STRING physical interactions, which STRING distributes under CC BY 4.0; see inst/PROVENANCE.md for the attribution.