A common source of metabolomics data is chromatography mass spectrometry. Adding support for this type of data to the MassSpecWavelet package will require adapting and expanding the package to consider the chromatographic time dimension.
A separate peak detection algorithm may be required for chromatographic peaks, since their shape differs from that of mass spectral peaks. Chromatographic peaks often overlap, necessitating a deconvolution procedure. Also, the large degree of temporal shift between samples requires aligning them by retention time.
The identification of metabolites is especially challenging in
gas chromatography, where each metabolite is fragmented into multiple ions. The individual ions must be grouped together for identification. The new version of MassSpecWavelet must address all of these issues. To avoid reinventing the wheel, the package should be integrated with existing software for metabolomics, such as the xcms Bioconductor package, where appropriate.

It would also be beneficial to make the MassSpecWavelet package generally more adaptable and more accessible to biologists. This would involve improvements along three axes. First, the design should be based on a modular, extensible data analysis pipeline. This would enable expert users to adapt MassSpecWavelet for specific needs. Second, MassSpecWavelet should visualize results and algorithm diagnostics with interactive graphics that promote open-ended
exploration of the data. Every dataset is unique and requires a flexible means of visualization in order to detect problems and to discover the unexpected. Finally, a graphical user interface (GUI) should be provided so that biologists are able to use MassSpecWavelet without spending valuable research time climbing steep technical learning curves.

Adding metabolomics support to MassSpecWavelet is the priority for this project. General improvements will only receive attention when they directly support the metabolomics efforts.