multiomeR is a targets-based workflow for processing and analyzing single-nucleus 10x Genomics Multiome data. It is designed as a lean, readable framework that users can adapt to their own studies rather than as a black-box command-line pipeline.
The active workflow is a single root targets project driven by _targets.R, the settings in configuration/. To select a separate settings directory for your project, use the ignored root configuration.local file.
multiomeR has been stable since 1.0. The manuscript describing it is under peer review (link to come).
Releases tag stable, validated commits; main can carry development work that is not yet validated. Update your clone to a release tag rather than pulling main. Commit or stash local changes first; resolve any merge conflicts as usual. With the newest release's tag in place of vX.Y.Z:
cd multiomeR # your clone
git fetch origin tag vX.Y.Z
git merge --ff --no-edit vX.Y.Z
pixi install --locked --run-post-link-scripts
pixi run --use-environment-activation-cache install-r-github-packagesEach release's notes repeat these commands with its tag and list any further steps. Major versions change configuration, target names or output schemas incompatibly; minor versions add features or change scientific methods; patch versions fix bugs. Any release can rebuild targets, and its notes say which.
The user manual is built from the Quarto book in website/. It includes a quickstart guide, an output gallery and full implementation details: https://koefoeden.github.io/multiomeR/
- Linux system with at least 32 GB of RAM, preferably equipped with a job-scheduler supported by the crew.cluster package: SLURM, PBS, SGE or LSF.
- The most memory-intensive targets need about 7 GB plus 2.5 GB per 100,000 nuclei. The optional
Seurat_SCTGEX backend needs more, about 0.25 GB per 1,000 nuclei to normalize gene expression.
If you use multiomeR, please cite the release you used: GitHub's Cite this repository gives its authors from CITATION.cff, and the release's Zenodo record gives its DOI. Once the manuscript describing multiomeR is published, this section will cite it.
multiomeR is built on targets and BPCells, so please also cite them:
- Landau WM. The targets R package: a dynamic Make-like function-oriented pipeline toolkit for reproducibility and high-performance computing. Journal of Open Source Software 2021;6:2959. https://doi.org/10.21105/joss.02959
- Parks B, Greenleaf W. Scalable high-performance single cell data analysis with BPCells. bioRxiv 2025. https://doi.org/10.1101/2025.03.27.645853
Please also cite the methods behind the results you report, such as weighted nearest-neighbor integration or SCAVENGE. The In depth pages of the user manual cite each method where it is used.
Bug reports and broadly useful feature requests are welcome, especially when they affect users analyzing 10x Multiome data. The project prioritizes lean, inspectable workflow changes over broad abstractions or site-specific convenience layers. See .github/CONTRIBUTING.md.