Skip to content

FBMN with XCMS3

Introduction

The Feature-Based Molecular Networking (FBMN) is a computational method that bridges popular mass spectrometry data processing tools for LC-MS/MS and molecular networking analysis on GNPS. The tools supported are: MZmine2, OpenMS, MS-DIAL, MetaboScape, XCMS, and Progenesis QI.

The main documentation for Feature-Based Molecular Networking can be accessed here:

Below follows a description on how to use xcms version >= 3 (XCMS3) with the FBMN workflow in GNPS.

Mass spectrometry processing with XCMS3

Installation

Install the latest version of XCMS3 from Bioconductor (version >= 3.4) in R with:

install("BiocManager")
BiocManager::install("xcms")

See also the [xcms Bioconductor package] (https://www.bioconductor.org/packages/release/bioc/html/xcms.html).

Clone also the GitHub repository https://github.com/jorainer/xcms-gnps-tools for utility functions (XCMS-GNPS-TOOLS) specific to this workflow.

Citations and development

This work builds on the efforts of our many colleagues, please cite their work:

https://github.com/sneumann/xcms

Tautenhahn R, Boettcher C, Neumann S. Highly sensitive feature detection for high resolution LC/MS BMC Bioinformatics, 9:504 (2008).

Smith, C.A., Want, E.J., O'Maille, G., Abagyan,R., Siuzdak, G. XCMS: Processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching and identification. Analytical Chemistry, 78, 779–787 (2006).

Mass Spectrometry Data Processing with XCMS3

As for all preprocessing software tools, a sequence of steps is performed to process mass spectrometry data in XCMS3.

Here we will present key steps and representative scripts required to process untargeted LC-MS/MS data collected using data dependent acquisition.

SCRIPT AVAILABILITY: Example XCMS3 scripts are accessible as Jupyter notebook and RCommander script on https://github.com/DorresteinLaboratory/XCMS3_FeatureBasedMN.

IMPORTANT: XCMS3 parameters will vary depending on the mass spectrometer, the acquisition parameters, and the samples investigated. The following documentation serves as a basic guideline for using XCMS3 with the Feature-Based Molecular Networking workflow.

Please consult these resources for more details on XCMS3 processing and definition of parameter settings:

1. Convert your LC-MS/MS Data to an Open Format

XCMS3 accepts different input formats. Note that we recommand to first convert your files to the mzML format before using XCMS3 for processing. See the documentation here.

2. Processing Steps with XCMS3

  1. Import data (readMSData)
  2. Peak picking (findChromPeaks)
  3. Retention time alignment (adjustRtime).
  4. Peak grouping (groupChromPeaks).
  5. Gap filling (fillChromPeaks).
  6. Export the data for FBMN on GNPS
    • a feature quantification table with ion intensities (.TXT file format).
    • a MS/MS spectral summary file (.MGF file format). Note that it is recommended to use the maxTIC option for the MGF export.

Perform the Feature-Based Molecular Networking on GNPS

The files exported from XCMS3 can be uploaded to the GNPS web-platform and a Feature-Based Molecular Networking job can be launched.

FBMN with XCMS3 can be performed either using the [Superquick FBMN start page] (http://dorresteinappshub.ucsd.edu:5050/featurebasednetworking) or the standard interface of the FBMN workflow (you need to be logged in GNPS first).

More information on the Feature Based Molecular Networking workflow on GNPS can be obtained at this documentation page.

Note that you can upload a metadata table with your job. See documentation.

Representative files and job

Test files are accessible here:

  1. A feature quantification table - Download here
  2. A MS/MS spectral summary file from above - Download here
  3. A metadata table - Download here

Here is an example FBMN job with XCMS3 from a subset of the American Gut Project.

Tutorials

See our XCMS-FBMN tutorials using python and R with a subset of the American Gut Project samples on this repository DorresteinLaboratory/XCMS3_FeatureBasedMN/.

Page contributors

Johannes Reiner (Eurac Research), Madeleine Ernst (UCSD), Ricardo da Silva (UCSD), Michael Wittig (Munich University), Louis Felix Nothias (UCSD).

Contribute to the Documentation