MACARRoN

Metabolome Analysis and Compound Annotation with Ranked Relevance on Networks

What is MACARRoN?

MACARRoN (Metabolome Analysis and Compound Annotation with Ranked Relevance on Networks) is a workflow for systematic prioritization of metabolites from untargeted metabolomics studies. It integrates metabolomics data with microbial community profiles (e.g., from MetaPhlAn or HUMAnN) to identify metabolites that are:

  1. Chemically distinct (novel or understudied)
  2. Differentially abundant between groups
  3. Microbiome-associated

MACARRoN answers the question: โ€œWhich metabolites in this dataset are the most biologically relevant and worth following up on?โ€

When to Use MACARRoN

Use MACARRoN when you have untargeted metabolomics data (LC-MS) and you want to:

  • Identify novel microbiome-linked metabolites
  • Prioritize metabolites for follow-up biochemical characterization
  • Integrate metabolomics with metagenomics data
  • Reduce the โ€œunknown compoundโ€ problem in metabolomics

Installation

R / Bioconductor

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("MACARRoN")

From GitHub

devtools::install_github("biobakery/macarron")

Basic Usage

MACARRoN is an R package. A typical workflow:

library(MACARRoN)

# Load input data
metabolomics_data <- read.csv("metabolomics.csv", row.names = 1)
metadata          <- read.csv("metadata.csv", row.names = 1)
annotations       <- read.csv("annotations.csv", row.names = 1)

# Run full MACARRoN pipeline
mac_output <- MACARRoN::MACARRoN(
  met_dat  = metabolomics_data,
  metadata = metadata,
  annotation_table = annotations,
  output   = "macarron_output/"
)

Input Requirements

Input Format Description
Metabolomics data CSV/TSV Feature ร— sample abundance matrix
Sample metadata CSV/TSV Sample covariates and group labels
Annotation table CSV/TSV Chemical annotations (from MS libraries)

Output

MACARRoN outputs a ranked table of metabolite features with the following scores:

Column Description
MAC_score Composite prioritization score
Avg_Abundance Mean abundance across samples
Prevalence Fraction of samples with detection
Qval Adjusted p-value from MaAsLin2
Effect_size Effect size of differential abundance
Annotation_cluster Chemical cluster identity

Tips & Gotchas

Tip

Pair with MaAsLin2 โ€” MACARRoN internally calls MaAsLin2 for the differential abundance step. Ensure MaAsLin2 is installed.

Warning

Missing annotations โ€” Many features in untargeted metabolomics are unannotated. MACARRoN can still prioritize these based on abundance and association patterns.

Tip

Hierarchical clustering of metabolite features is a key MACARRoN step. Features that cluster together chemically are scored together.

Further Reading