- Data should probably get filtered before creating an MsExperiment. A sageMsExperiment() constructor is thus arguably not needed. It would be better to document how to create an MsExperiment with the available test data.
- Write unit tests.
- Also add the full dataset?
- Re-generate mzML files including noise (see -n argument) and re-run sage to have SN ration.
- Add the filename corresponding the basename(dataOrigin)), needed for KEY.
- The current subset only contains good spectra. Would be better to also have some decoy and of low score.
- Add cell types: Pellets from 11 human cell lines (RKO, A549, U87_MG, HCT116, HEK293T, HeLa, MCF7, U2OS, SUM159, PANC1, and Jurkat)
- MS data has only MS2 scans.
- Remove contaminants.
- How to link the different slots in an MsExperiment? Could have a KEY
variable as
spectraData(.)$KEY,rowData(.)$KEYand.$KEYto be used as a unique consolidated foreign key. To be discussed with Jo. - We could then have a
ProtGenerics::makeKEY()generic and different methods ifSpectra,QFeaturesandPSMthat add the KEY variables. - Assuming we have it, how to use it? Could be a
subByKEY()function that filter the respective data types. When called on anMsExperiment, it delegates to its data types.
- R interface to manage config files; provide a set of standard configs.