Skip to contents

microbiomedataset

R-CMD-check pkgdown Lifecycle experimental badge

microbiomedataset is part of tidymicrobiome.

Why this package

microbiomedataset is a data infrastructure package for microbiome analysis. It provides a standardized object system for:

  • abundance matrices
  • sample metadata
  • taxonomy metadata
  • phylogenetic trees
  • reference sequences
  • process tracking

The package is designed around the same data-engineering philosophy as massdataset, so microbiome and metabolome data can be analyzed with aligned workflows and linked in cross-omics studies.

What is new

This package is built around three manuscript-facing ideas.

  1. A tree-aware microbiome data model. microbiome_dataset extends the massdataset object model with taxonomy, explicit tree-link metadata, and reference sequences.

  2. A unified analysis and visualization layer. The package standardizes preprocessing, diversity, ordination, differential abundance, network analysis, and publication-style plotting around a single input object.

  3. A microbiome-metabolome integration layer. microbiomedataset can work directly with massdataset objects for paired sample alignment, correlation analysis, multiblock integration, and mechanism-oriented taxon-pathway-metabolite linking.

Installation

Install the development version from GitHub:

if (!requireNamespace("remotes", quietly = TRUE)) {
  install.packages("remotes")
}

remotes::install_github("tidymicrobiome/microbiomedataset")

Core workflows

1. Microbiome data infrastructure 

library(microbiomedataset)

data("global_patterns", package = "microbiomedataset")

global_patterns

plot_composition(global_patterns, taxonomic_rank = "Phylum", top_n = 8)

2. Differential abundance and ordination 

library(microbiomedataset)

data("demo_crossomics", package = "microbiomedataset")

microbiome_object <- summarise_taxa(
  demo_crossomics$microbiome_data,
  taxonomic_rank = "Genus"
)

ordination_result <- run_ordination(microbiome_object, method = "PCoA")
plot_ordination(ordination_result, color_by = "study_group")

3. Microbiome-metabolome integration 

library(microbiomedataset)

data("demo_crossomics", package = "microbiomedataset")

correlation_result <- calculate_correlation(
  microbiome_data = demo_crossomics$microbiome_data,
  metabolome_data = demo_crossomics$metabolome_data,
  sample_link = demo_crossomics$sample_link,
  microbiome_rank = "Genus",
  method = "spearman",
  metabolome_transform = "none"
)

network_object <- build_correlation_network(
  correlation_result,
  min_abs_correlation = 0.2,
  max_q_value = 1,
  top_n = 25
)

plot_correlation_network(network_object)

Tutorials

Documentation site: https://tidymicrobiome.github.io/microbiomedataset/

Key tutorials:

  • Get started
  • Import and preprocess
  • Tree and sequence handling
  • Microbiome visualization
  • Cross-omics workflow
  • Advanced visualization

Citation

If you use microbiomedataset, please cite the package and the associated paper when available.

Contact

Xiaotao Shen