Abstract
A common problem with scientific research software is the lack of support, maintenance and further development. In particular, development by a single researcher can easily result in orphaned and dysfunctional software packages, especially if combined with non-compliance with software development standards.
The RforMassSpectrometry initiative aims to develop an efficient and stable infrastructure for mass spectrometry (MS) data. As part of this initiative, a growing ecosystem of tightly integrated R packages covering various aspects of metabolomics and proteomics data analysis is being developed. Their interoperability enables creation of complete analysis workflows without tedious data conversions. To avoid the aforementioned problems, community contributions are encouraged and emphasis is placed on open development, documentation and long-term support.
At the heart of the package ecosystem are the “Spectra” and “Chromatograms” packages that provide the core infrastructure for processing LC-MS data. Their design allows easy expansion to support additional file types, data representations or storage formats. The “xcms” package for LC-MS data preprocessing was recently updated to leverage this infrastructure enabling preprocessing of very large or remote data sets. Such data sets can also be loaded directly from the MetaboLights repository using the “MsBackendMetaboLights” package allowing seamless reanalysis of public dataset. Annotation of untargeted metabolomics experiments can be performed using the “MetaboAnnotation” package while the required annotation resources are easily accessible through packages such as “MsBackendMassbank”, “MsBackendMsp” or “CompoundDb”, the latter also allowing creation and management of lab-specific compound databases. The “SpectriPy” package enables the direct integration of Python libraries to create even more powerful analysis workflows. Finally, the `MsCoreUtils` and `MetaboCoreUtils` packages provide efficient implementations of commonly used algorithms, designed to be reused in other R packages. Ultimately, and in contrast to a monolithic software design, the RforMassSpectrometry software ecosystem enables implementation of powerful, data set tailored, and reproducible analysis workflows.