| name | maaslin2 |
| description | MaAsLin2 / MaAsLin3 — multivariable association discovery between metadata and meta-omic features (taxonomy, pathways, genes) with filtering, normalization, transforms, and FDR control. Use for differential abundance with covariates or longitudinal designs. Upstream: https://github.com/biobakery/Maaslin2 and https://github.com/biobakery/MaAsLin3. Route via research-analysis. Classic LDA biomarkers → lefse; figure workflows → microeco / visualization.
|
| license | MIT |
| category | evaluation |
| tags | ["MaAsLin2","MaAsLin3","differential-abundance","statistics","bioBakery"] |
| upstream | https://github.com/biobakery/Maaslin2 |
| stage | report |
MaAsLin2 / MaAsLin3
Upstream: biobakery/Maaslin2 ·
biobakery/MaAsLin3
Citation
Mallick, H. et al. Multivariable association discovery in population-scale
meta-omics studies. PLoS Comput. Biol. 17, e1009442 (2021).
https://doi.org/10.1371/journal.pcbi.1009442
MaAsLin 3: Nickols, W. A. et al. MaAsLin 3: refining and extending generalized
multivariable linear models for meta-omic association discovery. Nat. Methods
(2025). https://doi.org/10.1038/s41592-025-02923-9
See also docs/references.md.
Analytical thinking
Default for covariate-adjusted associations on compositional meta-omics
tables (MetaPhlAn, HUMAnN, MAG abundances, …). Pin normalization/transform,
fixed/random effects, and q-value threshold. MaAsLin3 adds prevalence vs
abundance association framing — cite the version you run.
| vs | Prefer MaAsLin when |
|---|
lefse | Multiple covariates / mixed models / modern FDR workflow |
microeco | Need model coefficients vs exploratory diversity suite |
How to run
Do not invent formula syntax — follow the installed package vignette.
Decision tree
Differential / association analysis?
├─ Covariates / longitudinal → maaslin2 (2 or 3)
├─ Classic two-class LDA biomarkers → lefse
├─ Explore diversity first → microeco
└─ Plot checklist → visualization
Related skills
research-analysis · ancombc · lefse · stamp · microeco ·
phyloseq · visualization · metaphlan · humann · study-design ·
tool-selection