Software
A method that nobody can run is not a method. Everything we publish ships as documented, maintained open-source code.
ANCOMBC
R / Bioconductor
Differential abundance analysis for microbiome and other compositional data. Implements ANCOM-BC (two-group, bias-corrected) and ANCOM-BC2 (multi-group, covariate adjustment, repeated measures), plus SECOM for linear and nonlinear taxa–taxa correlation.
BiocManager::install("ANCOMBC")
MetVAE
Python / PyPI
A variational autoencoder for metabolomics data that corrects for confounding while estimating molecular co-occurrence, supporting correlation-based molecular networks built from large-scale untargeted metabolomics.
pip install metvae
q2-composition
Contributor
The QIIME 2 plugin for compositional data analysis, which brings ANCOM-BC to the QIIME 2 ecosystem so microbiome researchers can run bias-corrected differential abundance inside their existing pipeline.
More coming
In development
New tools for multi-omics integration, causal mediation with compositional mediators, and explainable prediction models are in development and will be released here. A scalable Python implementation of ANCOM-BC is also in preprint.
How to cite
If our software supports your work, please cite the corresponding method paper:
- ANCOM-BC: Lin H, Peddada SD. Nature Communications 11:3514 (2020).
- ANCOM-BC2: Lin H, Peddada SD. Nature Methods 21:83–91 (2024).
- SECOM: Lin H, Eggesbø M, Peddada SD. Nature Communications 13:4946 (2022).
- MetVAE: Lin H, Zhang L, Lotfi A, Jarmusch A, Lee I, Kim A, Morton JT, Aksenov A. bioRxiv (2026). doi:10.1101/2025.04.26.649581 Preprint, under revision at STAR Protocols.