MicrobiomeProfiler

MicrobiomeProfiler is an R package and Shiny application for functional enrichment analysis in microbiome studies. It builds on the clusterProfiler ecosystem and provides both interactive and programmatic workflows for common microbiome annotation tasks.

The package now supports a hybrid data-delivery model:

• Stable built-in resources remain available in the package.
• Larger or fast-moving annotation resources are distributed from GitHub Pages and downloaded on demand.
• Remote datasets can be cached locally and refreshed independently from the package release cycle.

What It Supports

MicrobiomeProfiler currently supports:

• KEGG enrichment and GSEA for microbiome gene profiles
• COG enrichment and GSEA
• eggNOG-based KEGG pathway enrichment and GSEA
• Microbe-disease enrichment with Disbiome
• Microbial signature enrichment with BugSigDB
• Metabolite pathway enrichment with HMDB and SMPDB
• Interactive Shiny exploration with tables, dotplots, and barplots

Installation

Install the release version from Bioconductor:

if (!requireNamespace("BiocManager", quietly = TRUE)) {
    install.packages("BiocManager")
}
BiocManager::install("MicrobiomeProfiler")

Or install the development version from GitHub:

if (!requireNamespace("remotes", quietly = TRUE)) {
    install.packages("remotes")
}
remotes::install_github("YuLab-SMU/MicrobiomeProfiler")

Launch The Shiny App

library(MicrobiomeProfiler)
run_MicrobiomeProfiler()

The Shiny interface currently provides:

• Gene enrichment analysis
  • KEGG
  • COG
  • eggNOG
  • eggNOG ORA and GSEA modes in the same page
• Microbe disease/signature enrichment
  • Disbiome
  • BugSigDB
• Metabo-Pathway analysis

For eggNOG GSEA in the app, provide one ranked item per line, for example:

OG0001 2.5
OG0002 1.5
OG0003 -0.8

Programmatic Usage

KEGG And COG

library(MicrobiomeProfiler)

data(Rat_data)
ko_res <- enrichKO(Rat_data)

data(Psoriasis_data)
cog_res <- enrichCOG(Psoriasis_data, dtype = "pathway")

eggNOG

library(MicrobiomeProfiler)

og <- c(
    "Collectrin@131567|A-1*",
    "Collectrin@7711|C-2",
    "Collectrin@75365|II-17"
)

ora_res <- enrichEggNOG(og, minGSSize = 1, maxGSSize = 500)

# GSEA expects a named ranked numeric vector of real eggNOG OG IDs.
geneList <- c(2.5, 1.5, -0.8)
names(geneList) <- og

Microbe-Disease And Signature Enrichment

library(MicrobiomeProfiler)

data("microbiota_taxlist", package = "MicrobiomeProfiler")

bugsigdb_res <- enrichBugSigDB(microbiota_taxlist)
mda_res <- enrichMDA(microbiota_taxlist)

External Data Delivery

Some annotation resources are distributed from the package GitHub Pages site instead of being bundled directly in the source tarball. This keeps package size under control and allows annotation artifacts to be refreshed independently.

Current external datasets include:

• bugsigdb
• disbiome
• eggnog

Inspect configured datasets:

library(MicrobiomeProfiler)

available_datasets()
available_datasets(include_remote = TRUE)

Pre-download data into the local cache:

download_dataset("bugsigdb")
download_dataset("disbiome")
download_dataset("eggnog")

Refresh or inspect cache state:

download_dataset("eggnog", refresh = TRUE)

dataset_cache_info()
dataset_cache_info("eggnog")

clear_dataset_cache("eggnog")

Runtime behavior:

• enrichBugSigDB() downloads BugSigDB artifacts on demand.
• enrichEggNOG() and gseEggNOG() download eggNOG artifacts on demand.
• enrichMDA() and gseMDA() use remote-first Disbiome data and fall back to bundled internal data if the remote source is unavailable.

Documentation

See the package vignette for a broader introduction:

• Introduction to MicrobiomeProfiler

Reference

• T Wu^#, E Hu^#, S Xu, M Chen, P Guo, Z Dai, T Feng, L Zhou, W Tang, L Zhan, X Fu, S Liu, X Bo, G Yu. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. The Innovation. 2021, 2(3):100141. doi: 10.1016/j.xinn.2021.100141