patRoon aims to provide comprehensive mass spectrometry based non-target analysis (NTA) workflows for environmental analysis. The name is derived from a Dutch word that means pattern and may also be an acronym for hyPhenated mAss specTROmetry nOn-target aNalysis.
Project news
December 2023 patRoon 2.3.1 is now released. This is a small maintenance release with several bug fixes. The most important fix is a correction for the prediction of concentrations from SIRIUS fingerprints. Please see the Project NEWS for more details and important notes on updating patRoon.
November 2023 patRoon 2.3 is now released. The most significant changes include improved installation strategies and integration of MS2Tox and MS2Quant to predict feature toxicities and concentrations. Please see the Project NEWS for more details and important notes on updating patRoon.
May 2023 patRoon 2.2 is now released. The most significant change is the addition of a new reporting interface, which brings a much improved HTML interface, many optimizations and other important new functionality. Furthermore, patRoon 2.2 introduces improved SIRIUS 5 support, a new TP screening algorithm using formula libraries and many other improvements, of which many thanks to the great user feedback. Please see the Project NEWS for details.
March 2023 The Docker images moved to a new host. Please see the see installation details in the handbook to obtain the latest images.
May 2022 patRoon 2.1 is now available. This new release integrates prediction of transformation products with CTS, adds several feature intensity normalization methods, adds new functionality and improvements for reporting TP data and supports loading, processing and annotation with MS libraries such as MassBank. Please see the Project NEWS for details.
Introduction
Mass spectrometry based non-target analysis is used to screen large numbers of chemicals simultaneously. For this purpose, high resolution mass spectrometry instruments are used which are typically coupled (or hyphenated) with chromatography (e.g. LC or GC). The size and complexity of resulting data makes manual processing impractical. Many software tools were/are developed to facilitate a more automated approach. However, these tools are generally not optimized for environmental workflows and/or only implement parts of the functionality required.
patRoon combines established software tools with novel functionality in order to provide comprehensive NTA workflows. The different algorithms are provided through a consistent interface, which removes the need to know all the details of each individual software tool and performing tedious data conversions during the workflow. The table below outlines the major functionality of patRoon.
| Functionality | Description | Algorithms |
|---|---|---|
| Raw data pre-treatment | MS format conversion (e.g. vendor to mzML) and calibration. |
ProteoWizard, OpenMS, DataAnalysis |
| Feature extraction | Finding features and grouping them across analyses. | XCMS, OpenMS, enviPick, DataAnalysis, KPIC2, SIRIUS, SAFD |
| Suspect screening | Finding features with suspected presence by MS and chromatographic data. Estimation of identification confidence levels. | Native |
| MS data extraction | Automatic extraction and averaging of feature MS(/MS) peak lists. | Native, mzR, DataAnalysis |
| Formula annotation | Automatic calculation of formula candidates for features. | GenForm, SIRIUS, DataAnalysis |
| Compound annotation | Automatic (in silico) compound annotation of features. | MetFrag, SIRIUS, Native |
| Componentization & adduct annotation | Grouping of related features based on chemistry (e.g. isotopes, adducts and homologs), hierarchical clustering or MS/MS similarity into components. Using adduct and isotope annotations for prioritizing features and improving formula/compound annotations. | RAMClustR, CAMERA, nontarget R package, OpenMS, cliqueMS, Native |
| Combining algorithms | Combine data from different algorithms (e.g. features, annotations) and generate a consensus. | Native |
| Sets workflows | Simultaneous processing and combining +/- MS ionization data | Native |
| Transformation product (TP) screening | Automatic screening of TPs using library/in-silico data, MS similarities and classifications. Tools to improve compound TP annotation. | BioTransformer, PubChemLite, Native |
| Reporting | Automatic reporting of all important workflow data. An example HTML report can be viewed here. | Native |
| Data clean-up & prioritization | Filters for blanks, replicates, intensity thresholds, neutral losses, annotation scores, identification levels and many more. | Native |
| Data curation | Several graphical interactive tools and functions to inspect and remove unwanted data. | Native |
The workflow of non-target analysis typically depends on the aims and requirements of the study and the instrumentation and methodology used for sample analysis. For this reason, patRoon does not enforce a certain workflow. Instead, most workflow steps are optional, fully configurable and algorithms can easily be mixed or even combined.
Implementation details
-
patRoonis implemented as an R package, which allows easy interfacing with the many otherRbased MS tools and other data processing functionality fromR. - Fully open-source (GPLv3).
- Developed on Windows, Linux and macOS
- S4 classes and generics are used to implement a consistent interface to all supported algorithms.
- Continuous integration is used to automatically perform unit tests, update the Website and documentation, and maintaining installation resources such as the patRoonDeps repository, Docker image and
patRoonbundle (see the handbook for more details). - Supports all major instrument vendor input formats (through usage of ProteoWizard and DataAnalysis).
- Optimizations
-
data.tableis used internally as a generally much more efficient alternative todata.frame. - The processx and future
Rpackages are used for parallelization. - Results from workflow steps are cached within a SQLite database to avoid repeated computations.
- Code for loading MS and EIC data, MS similarity calculations and others were implemented in
C++to reduce computational times.
-
- The RDCOMClient
Rpackage is used to interface with Bruker DataAnalysis algorithms. - The Shiny
Rpackage was used to implement several GUI tools. - The reporting functionality relies on the excellent R markdown and related packages such as flexdashboard, bslib and reactable.
Installation
patRoon and its dependencies can be installed in various ways. Please see the installation section in the handbook for more information.
Getting started
For a very quick start:
The newProject() function will pop-up a dialog screen (requires R Studio), which will allow you to quickly select the analyses and common workflow options to subsequently generate a template R processing script.
However, for a better guide to get started it is recommended to read the tutorial. Afterwards the handbook is a recommended read if you want to know more about advanced usage of patRoon. Finally, the reference outlines all the details of the patRoon package.
Citing
When you use patRoon please cite its publications:
Rick Helmus, Thomas L. ter Laak, Annemarie P. van Wezel, Pim de Voogt and Emma L. Schymanski. patRoon: open source software platform for environmental mass spectrometry based non-target screening. Journal of Cheminformatics 13, 1 (2021)
Rick Helmus, Bas van de Velde, Andrea M. Brunner, Thomas L. ter Laak, Annemarie P. van Wezel and Emma L. Schymanski. patRoon 2.0: Improved non-target analysis workflows including automated transformation product screening. Journal of Open Source Software, 7(71), 4029
patRoon builds on many open-source software tools and open data sources. Therefore, it is important to also cite their work when using these algorithms via patRoon.
Contributing
For bug reports, code contributions (pull requests), questions, suggestions and general feedback please use the GitHub page.