filterPeakSatellites: Filter satellite peaks
In sneumann/RMassBank: Workflow to process tandem MS files and build MassBank records

Description Usage Arguments Details Value Note Author(s) See Also Examples

Filters satellite peaks in FT spectra which arise from FT artifacts and from conversion to stick mode. A very simple rule is used which holds mostly true for MSMS spectra (and shouldn't be applied to MS1 spectra which contain isotope structures...)

1	filterPeakSatellites(peaks, filterSettings = getOption("RMassBank")$filterSettings)

`peaks`	A peak dataframe with at least the columns `mz, int`. Note that `mz` is used even for the recalibrated spectra, i.e. the desatellited spectrum is identical for both the unrecalibrated and the recalibrated spectra.
`filterSettings`	The settings used for filtering. Refer to `analyzeMsMs` documentation for filter settings.

The function cuts off all peaks within 0.5 m/z from every peak, in decreasing intensity order, which are below 5 intensity. E.g. for peaks m/z=100, int=100; m/z=100.2, int=2, m/z=100.3, int=6, m/z 150, int=10: The most intense peak (m/z=100) is selected, all neighborhood peaks below 5 peak) and the next less intense peak is selected. Here this is the m/z=150 peak. All low-intensity neighborhood peaks are removed (nothing). The next less intense peak is selected (m/z=100.3) and again neighborhood peaks are cut away (nothing to cut here. Note that the m/z = 100.2 peak was alredy removed.)

Returns the peak table with satellite peaks removed.

This is a very crude rule, but works remarkably well for our spectra.

Michael Stravs

analyzeMsMs, filterLowaccResults

# From the workflow:
## Not run: 
    # Filter out satellite peaks:
    shot <- filterPeakSatellites(shot)
    shot_satellite_n <- setdiff(row.names(shot_full), row.names(shot))
    shot_satellite <- shot_full[shot_satellite_n,]
    # shot_satellite contains the peaks which were eliminated as satellites.

## End(Not run)