Skip to content

footprintR

Tools for working with single-molecule footprinting data — reads base modifications from BAM files or modkit output, represents single-molecule footprinting data efficiently as R objects, and provides functions to manipulate and visualize them. Output is a SummarizedExperiment with positions × samples and assays for modified/total base counts.

  • Authors: Charlotte Soneson, Panagiotis Papasaikas, Sebastien Smallwood, Michael Stadler — FMI Basel (Friedrich Miescher Institute)
  • Source / install: github.com/fmicompbio/footprintR · BiocManager::install("fmicompbio/footprintR") (not in Bioconductor release as of 3.23)
  • Pkgdown: fmicompbio.github.io/footprintR
  • Status at EuroBioC2025: external / GitHub-distributed (Bioconductor submission likely pending)

Talk at EuroBioC2025

  • Speaker: Charlotte Soneson (FMI Basel)
  • Day / session: Day 2 (Thu Sep 18, 2025) — morning short-talk session, after the Sharpe keynote
  • Talk title: "footprintR for single molecule footprinting data"
  • 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

footprintR ingests base-modification calls (from BAM files or from modkit) and lays them out as a SummarizedExperiment keyed by genomic position × sample, with assays for modified and total base counts. From there it provides manipulation and visualization functions tailored to single-molecule footprinting (SMF) experiments — where exogenous methyltransferase footprints reveal TF binding at single-DNA-molecule resolution.

How it works

Core idea. Read per-base modification calls (typically from Oxford Nanopore data processed by modkit) into a SummarizedExperiment whose rows are genomic positions and columns are samples, with two key assays: Nmod (count of modified bases observed at that position in that sample) and Nvalid (count of total valid base calls at the same position). Methylation rate per position per sample is then simply Nmod / Nvalid, but keeping the two counts separate preserves the coverage information needed for downstream statistics.

Inputs / outputs. Input: BAM files with MM/ML modification tags, or a bedMethyl-format file produced by modkit pileup. Output: a SummarizedExperiment with rowRanges of positions, colData of samples, and Nmod / Nvalid assays.

Key innovation. Modern Bioconductor-style data class for SMF data — SummarizedExperiment rather than bespoke list/array structures — which makes SMF analyses composable with the rest of the Bioconductor ecosystem (e.g. GenomicRanges subsetting, plyranges operations). Successor in spirit to the older SingleMoleculeFootprinting package from the same FMI group, with more modern data structures.

Parameters worth knowing. readBedMethyl() arguments controlling which modification type to ingest (e.g. 5mC vs 6mA from a multi-mod modkit output). Standard SummarizedExperiment subsetting on rowRanges to restrict to regions of interest before plotting.

Canonical example. The package ships an example modkit_pileup_1.bed.gz (accessible via system.file("extdata", ...)). Vignette workflow: readBedMethyl() loads the bedMethyl file into a SummarizedExperiment, plotRegion() then visualizes modification rates across a genomic region of interest — the SMF readout that maps TF footprints at single-molecule resolution.

Where it fits in the corpus

  • AACR 2026: axis = epigenomics / regulatory genomics; SMF is a niche but powerful TF-binding readout
  • ISMB 2026: scaffold; epigenomics methods
  • GBCC 2025: queued
  • Nextflow Summit 2026: not mentioned — modkit upstream is CLI-based but the analysis side is R

Notes

Single-molecule footprinting (SMF) reads out TF binding at single-DNA-molecule resolution by exposing chromatin to an exogenous methyltransferase whose footprint marks bound vs. unbound DNA. footprintR sits downstream of modkit calls and is complementary to the older SingleMoleculeFootprinting package on Bioconductor — same FMI lineage, more modern data structures.