Parameters for averaging MS peak list data

Source: R/utils-mspeaklists.R

Create parameter lists for averaging MS peak list data.

getDefAvgPListParams(..., IMS = getLimIMS())

Arguments

...

Optional named arguments that override defaults.

IMS

A character that specifies for which IMS instrument defaults are returned. Should be "bruker" or "agilent". Defaults to what is specified in limits.

Value

getDefAvgPListParams returns a list with the peak list averaging parameters.

Details

The parameters set used for averaging peak lists are set by the avgFeatParams and avgFGroupParams arguments to generateMSPeakLists and its related algorithm specific functions. The parameters are specified as a named list with the following values:

  • method,clusterMzWindow The cluster method and window (see clustering parameters) used to average mass spectra. clusterMzWindow is defaulted as defaultLim("mz", "medium") (see limits).

  • topMost Only retain this maximum number of MS peaks when generating averaged spectra. Lowering this number may exclude more irrelevant (noisy) MS peaks and decrease processing time, whereas higher values may avoid excluding lower intense MS peaks that may still be of interest.

  • minIntensityPre MS peaks with intensities below this value will be removed (applied prior to selection by topMost and averaging).

  • minIntensityPost MS peaks with intensities below this value will be removed (after averaging).

  • minIntensityIMS MS peaks in spectra of raw IMS frames with intensities below this value will be removed (applied prior to any other treatment steps).

  • absMinAbundance,relMinAbundance Minimum absolute/relative abundance of an MS peak across the spectra that are averaged. If absMinAbundance exceeds the number of spectra then the threshold is automatically lowered to the number of spectra.

  • minRelCumIntensity Minimum relative cumulative intensity of an MS peak in the averaged spectrum.

  • smoothWindowIMS,halfWindowIMS,maxGapIMS Parameters used for centroiding m/z peaks from IMS-HRMS data. See Centroiding IMS data for more details.

  • withPrecursorMS For MS data only: ignore any spectra that do not contain the precursor peak.

    For IMS data this excludes MS spectra within an IMS frame that do not contain the precursor peak, typically due to mobility separation. Hence, setting this option performs some crude cleanup of MS spectra, even for features for which no mobilities were assigned (e.g. non-IMS workflows).

  • pruneMissingPrecursorMS For MS data only: if TRUE then peak lists without a precursor peak are removed. Note that even when this is set to FALSE, functionality that relies on MS (not MS/MS) peak lists (e.g. formulae calculation) will still skip calculation if a precursor is not found.

  • retainPrecursorMSMS For MS/MS data only: if TRUE then always retain the precursor mass peak even if is not amongst the topMost peaks. Note that MS precursor mass peaks are always kept. Furthermore, note that precursor peaks in both MS and MS/MS data may still be removed by intensity thresholds (this is unlike the filter method function).

The getDefAvgPListParams function can be used to generate a default parameter list. The defaults are (with IMS="bruker"):

list(
  method = "distance_mean",
  clusterMzWindow = 0.005,
  topMost = 50,
  minIntensityPre = 500,
  minIntensityPost = 500,
  minIntensityIMS = 25,
  absMinAbundance = 0,
  relMinAbundance = 0,
  minRelCumIntensity = 0,
  smoothWindowIMS = 0,
  halfWindowIMS = 2,
  maxGapIMS = 0.005,
  withPrecursorMS = TRUE,
  pruneMissingPrecursorMS = TRUE,
  retainPrecursorMSMS = TRUE
)

Centroiding IMS data

With IMS-HRMS data the m/z peaks are often not or partially centroided. The following steps are performed to centroid the data:

  1. Sum up mass spectra within an IMS frame. If the feature has mobility data, only spectra within its mobility boundaries are considered.

  2. Use point-distance clustering (see clustering parameters) with a window defined by maxGapIMS to find related mass signals. This is primarily meant for non-continuous data, e.g. due to intensity thresholding. The maxGapIMS parameter should be set to a value that represents the maximum expected distance between two m/z datapoints. For some instruments, such as Agilent IMS-QTOF, this value may be higher than expected. For that reason, if IMS="agilent" then the default is set to 0.01.

  3. Smooth the intensity data using a centered moving average with window size smoothWindowIMS (set to zero to disable smoothing).

  4. Find local maxima within sliding window with +/- halfWindowIMS points and eliminate non-centroids. This algorithm is based on the C_localMaxima function from MALDIquant.

References

Gibb S, Strimmer K (2012). “MALDIquant: a versatile R package for the analysis of mass spectrometry data.” Bioinformatics, 28(17), 2270–2271. doi:10.1093/bioinformatics/bts447 .