This function performs post acquisition signal processing for list of MassSpectrum objects using commonly used methods : transform intensities ("sqrt"), smoothing ("Wavelet"), remove baseline ("SNIP"), calibrate intensities ("TIC") and align spectra. Methods used are selected from the MALDIquant and MALDIrppa R packages.
SignalProcessing(x,
transformIntensity_method = "sqrt",
smoothing_method = "Wavelet",
removeBaseline_method = "SNIP",
removeBaseline_iterations = 25,
calibrateIntensity_method = "TIC",
alignSpectra_NoiseMethod = "MAD",
alignSpectra_method = "lowess",
alignSpectra_halfWs = 11,
alignSpectra_SN = 3,
tolerance_align = 0.002,
referenceSpectra = NULL,
minFrequency= 0.5,
binPeaks_method = "strict",
keepReferenceSpectra = FALSE,
...)A list of modified MassSpectrum objects (see MALDIquant R package) according to chosen arguments. If the argument referenceSpectra is not completed and the argument keepReferenceSpectra is TRUE, a list containing the MassSpectrum objects modified named "spectra" and the created reference spectrum named "RefS" is returned.
a list of MassSpectrum objects (see MALDIquant R package).
a character indicating the method used to transform intensities: "sqrt" by default. This function can be replaced by another mathematical function such as "log".
a character indicating the smoothing methods used. By default, it performs undecimated Wavelet transform (UDWT) for list of MassSpectrum objects. This Smoothing method can be remplaced by "SavitzkyGolay" or "MovingAverage". See wavSmoothing in the MALDIrppa R package for details.
a character indicating the method used to remove baseline. It uses "SNIP" method for list of MassSpectrum objects. This baseline estimation method can be remplaced "TopHat", "ConvexHull" or "median". See removeBaseline-methods of the MALDIquant R package for details.
a numeric value indicting the number of iterations to remove baseline (by default = 25). See removeBaseline-methods of the MALDIquant R package for details.
a character indicating the intensities calibration method used ("TIC" method by default). This calibration method can be remplaced by "PQN" or "median".See calibrateIntensity-methods of the MALDIquant R package for details.
a character indicating the noise estimation method. It uses "MAD" method for list of MassSpectrum objects. This noise estimation method estimation method can be remplaced "SuperSmoother". See estimateNoise-methods of the MALDIquant R package for details.
a character indicating the warping method. It uses "lowess" method for list of MassSpectrum objects. This warping method method can be remplaced "linear", "quadratic" or "cubic" . See determineWarpingFunctions of the MALDIquant R package for details.
a numeric value half window size to detect peaks (by default = 11). See detectPeaks-methods of the MALDIquant R package for details.
a numeric value indicating the signal-to-noise ratio used to detect peaks (by default = 3). See detectPeaks-methods of the MALDIquant R package for details.
a numeric value indicating a maximal relative deviation of a peak position (mass) to be considered as identical in ppm (by default = 0.002). See determineWarpingFunctions of the MALDIquant R package for details.
a MassPeaks reference spectrum for alignment of the sample(s). If this reference spectrum is not provided, a reference spectrum is created using the minFrequency and binPeaks_method arguments. See referencePeaks of the MALDIquant R package for details.
a numeric minimum frequency for each peak over all analyzed spectra (by default = 0.5) for the creation of the reference spectrum. See referencePeaks of the MALDIquant R package for details.
a character indicating the method used to equalize masses for similar peaks for the creation of the reference spectrum. The "strict" method is used by default corresponding to a unique peak per bin from the same sample. This method can be remplaced by "relaxed" corresponding to multiple peaks per bin from the same sample. See binPeaks of the MALDIquant R package for more details.
a logical value indicating if the created reference spectrum is returned by the function (FALSE, default).
other arguments from MALDIrppa packages for the wavSmoothing function such as n.levels (corresponding to the depth of the decomposiion for the wavelet function). See wavSmoothing of the MALDIrppa R package for details.
The SignalProcessing function provides an analysis pipeline for MassSpectrum objects including intensity transformation, smoothing, removing baseline.
The Wavelet method relies on the wavShrink function of the wmtsa package and its dependencies (now archived by CRAN). The original C code by William Constantine and Keith L. Davidson, in turn including copyrighted routines by Insightful Corp., has been revised and included into MALDIrppa for the method to work.
All the methods used for SignalProcessing functions are selected from MALDIquant and MALDIrppa packages.
Godmer et al. (2025) presents a comparison of different pre-processing workflows that can help you to optimize your workflow.For a comprehensive guide, additional applications, and detailed examples of using this package, please visit our GitHub repository: here.
Gibb S, Strimmer K. MALDIquant: a versatile R package for the analysis of mass spectrometry data. Bioinformatics. 2012 Sep 1;28(17):2270-1. tools:::Rd_expr_doi("10.1093/bioinformatics/bts447"). Epub 2012 Jul 12. PMID: 22796955.
Javier Palarea-Albaladejo, Kevin Mclean, Frank Wright, David G E Smith, MALDIrppa: quality control and robust analysis for mass spectrometry data, Bioinformatics, Volume 34, Issue 3, 01 February 2018, Pages 522 - 523, tools:::Rd_expr_doi("https://doi.org/10.1093/bioinformatics/btx628")
Alexandre Godmer, Yahia Benzerara, Emmanuelle Varon, Nicolas Veziris, Karen Druart, Renaud Mozet, Mariette Matondo, Alexandra Aubry, Quentin Giai Gianetto, MSclassifR: An R package for supervised classification of mass spectra with machine learning methods, Expert Systems with Applications, Volume 294, 2025, 128796, ISSN 0957-4174, tools:::Rd_expr_doi("10.1016/j.eswa.2025.128796").
library("MALDIquant")
library("MSclassifR")
## Load mass spectra
data("CitrobacterRKIspectra", package = "MSclassifR")
# plot first unprocessed mass spectrum
PlotSpectra(SpectralData=CitrobacterRKIspectra[[1]], col_spec="blue")
## spectral treatment
spectra <- SignalProcessing(CitrobacterRKIspectra,
transformIntensity_method = "sqrt",
smoothing_method = "Wavelet",
removeBaseline_method = "SNIP",
removeBaseline_iterations = 25,
calibrateIntensity_method = "TIC",
alignSpectra_Method = "MAD",
alignSpectra_halfWs = 11,
alignSpectra_SN = 3,
tolerance_align = 0.002)
# plot first processed mass spectrum
PlotSpectra(SpectralData=spectra[[1]], col_spec="blue")
Run the code above in your browser using DataLab