Spectra: a Scalable and Flexible Infrastructure for Mass Spectrometry Data Analysis in R

The R statistical environment has become one of the most frequently used analysis platforms in (bio)statistical data analysis. Several software solutions exist also in R to handle, process and analyze metabolomics data or mass spectrometry (MS) data in general, but they mostly don't share data structures hence preventing interoperability. The objective of the R for Mass Spectrometry initiative is to provide a data analysis and development environment that is flexible (it must support user-centered applications such as data exploration and analysis, and developer-centered requirements such as enabling novel method development), efficient and scalable (applicable to small and large data sets on desktops, laptops or HPC systems), reproducible (to support reproducible research practices) and thoroughly tested and documented. As first packages we implemented the MsCoreUtils package that provides core functionality for MS data which is independent of any data structure, and the Spectra package to represent MS spectra data independently of its storage or origin and provide user functionality to visualize and process this data efficiently, including centroiding, smoothing, combining, normalizing, comparing and matching of MS spectra e.g. against MS spectral databases. Adapting existing software packages such as xcms to these new infrastructure will simplify their integration into complete data analysis pipelines. The R for Mass Spectrometry initiative has a strong commitment to open source software and community contributions are highly welcome. All of its packages are, or will be, part of the Bioconductor project