Metabonaut presents a series of workflows based on a small LC-MS/MS dataset, utilizing R and Bioconductor packages. These workflows demonstrate how to adapt various algorithms to specific datasets and seamlessly integrate R packages for efficient, reproducible data processing.
Louail P, De Graeve M, Mahmoud A, Marques de SÑ E Silva D, Nishida K, Suksi V, Tagliaferri A, Tomè G, Verri Hernandes V, Rainer J. Metabonaut: an open educational resource for learning reproducible, script-based untargeted LC-MS/MS metabolomics. Metabolomics. 2026 Oct 3;22(5):165. doi: 10.1007/s11306-026-02538-x. PMID: 42829392.
There are three ways to run the workflows, from the least to the most setup:
rformassspectrometry/metabonaut:v1.6.2@sha256:e17ab1dd7b397798d70c1bf03c923dfa4ea6f7b96263d79d92eb748f47e38ef2
and run it locally. See the installation
instructions.This primary workflow guides you through each step of the analysis, from preprocessing raw data to statistical analysis and metabolite annotation. π View Full R Code (Quarto)
Before diving into the analysis, learn about key aspects to examine in your dataset to ensure smooth processing and prevent troubleshooting issues later in the pipeline.
Explore notame as a robust alternative for normalization and feature selection. This vignette covers essential steps for quality control, normalization, and feature selection to ensure your data is clean and ready for analysis.
While xcms is known for its scalability, this guide shows you how to practically handle large-scale dataset processing (>4,000 files) on standard hardware.
Learn how to create and utilize custom metabolomics data annotation resources in R. This vignette guides you through building your own annotation databases and integrating them into your analysis workflow. We demonstrate by downloading data from GNPS and MassBank libraries.
Explore the SpectriPy package for LC-MS/MS data annotation. This tutorial demonstrates how to combine the strengths of Python and R MS libraries for comprehensive annotation.
Use the RuSirius package to annotate features with SIRIUS, including formula identification, structure database search, and de novo structure generation using MSNovelist for features with low-confidence database matches.
Discover how to use a flexible alignment algorithm to integrate new datasets with previously processed ones based on features of interest.
Use RmzTabM to import and export Summarized Experiment object to mzTab-M 2.1 files.
For a full list of all available vignettes, visit the Metabonaut website.
We strive for reproducibility. These workflows are designed to remain stable over time, allowing you to run all vignettes together as one comprehensive super-vignette. An archived Docker image can be used to fully replicate the results. The Metabonaut Docker image can also be re-build based on the repository's Dockerfile, which will however not create an identical image due to eventually different versions of CRAN packages available at build time.
The archived Metabonaut Docker image for version 1.6.2 is
rformassspectrometry/metabonaut:v1.6.2@sha256:b963d18e1496742af4e75a7992ba386de6c7b00fd3cc9c8100e1dac04d329c20
The tutorials assume basic knowledge of R and RMarkdown / Quarto. If you're new to these, we recommend starting with a short tutorial before running the vignettes.
This is just the beginning of our Metabonaut journey, and we're actively refining the website. If you're experiencing any issues:
β Ensure you have the latest versions of all required packages. π If the issue persists, report it with a reproducible example on GitHub Issues.
Currently, all vignettes run without problems.
Interested in contributing? Please check out the RforMassSpectrometry Contributions Guide.
We follow the RforMassSpectrometry Code of Conduct to maintain an inclusive and respectful community.
This work is funded by the European Union under the HORIZON-MSCA-2021 project 101073062: HUMAN β Harmonising and Unifying Blood Metabolic Analysis Networks.
π Learn more: HUMAN Project Website