quantify-methods: Quantifies 'MSnExp' and 'Spectrum' objects

In lgatto/MSnbase: Base Functions and Classes for Mass Spectrometry and Proteomics

Quantifies 'MSnExp' and 'Spectrum' objects

Description

This method quantifies individual "Spectrum" objects or full "MSnExp" experiments. Current, MS2-level isobar tagging using iTRAQ and TMT (or any arbitrary peaks of interest, see "ReporterIons") and MS2-level label-free quantitation (spectral counting, spectral index or spectral abundance factor) are available.

Isobaric tag peaks of single spectra or complete experiments can be quantified using appropriate methods. Label-free quantitation is available only for MSnExp experiments.

Since version 1.13.5, parallel quantitation is supported by the BiocParallel package and controlled by the BPPARAM argument.

Arguments

object	An instance of class "Spectrum" (isobaric tagging only) or "MSnExp".
method	Peak quantitation method. For isobaric tags, one of, possibly abreviated "trapezoidation", "max", or "sum". These methods return respectively the area under the peak(s), the maximum of the peak(s) or the sum of all intensities of the peak(s). For label-free quantitation, one of "SI" (spectral index), "SIgi" (global intensity spectral index), "SIn" (normalised spectral index), "SAF" (spectral abundance factor) or "NSAF" (normalised spectral abundance factor). Finally, the simple "count" method counts the occurrence of the respective spectra (at this stage all 1s) that can then be used as input to combineFeatures to implement spectra counting.
reporters	An instance of class "ReporterIons" that defines the peak(s) to be quantified. For isobaric tagging only.
strict	For isobaric tagging only. If strict is FALSE (default), the quantitation is performed using data points along the entire width of a peak. If strict is set to TRUE, once the apex(es) is/are identified, only data points within apex +/- width of reporter (see "ReporterIons") are used for quantitation.
BPPARAM	Support for parallel processing using the BiocParallel infrastructure. When missing (default), the default registered BiocParallelParam parameters are applied using bpparam(). Alternatively, one can pass a valid BiocParallelParam parameter instance: SnowParam, MulticoreParam, DoparParam, ... see the BiocParallel package for details.
parallel	Deprecated. Please see BPPARAM.
qual	Should the qual slot be populated. Default is TRUE.
pepseq	A character giving the peptide sequence column in the feature data. Default is "sequence".
verbose	Verbose of the output (only for MSnExp objects).
...	Further arguments passed to the quantitation functions.

Details

"ReporterIons" define specific MZ at which peaks are expected and a window around that MZ value. A peak of interest is searched for in that window. Since version 1.1.2, warnings are not thrown anymore in case no data is found in that region or if the peak extends outside the window. This can be checked manually after quantitation, by inspecting the quantitation data (using the exprs accessor) for NA values or by comaring the lowerMz and upperMz columns in the "MSnSet" qual slot against the respective expected mz(reporters) +/- width(reporters).

Once the range of the curve is found, quantification is performed. If no data points are found in the expected region, NA is returned for the reporter peak MZ.

Note that for label-free, spectra that have not been identified (the corresponding fields in the feature data are populated with NA values) or that have been uniquely assigned to a protein (the nprot feature data is greater that 1) are removed prior to quantitation. The latter does not apply for method = "count" but can be applied manually with removeMultipleAssignment.

Methods

signature(object = "MSnExp", method = "character", reporters = "ReporterIons", verbose = "logical", ...)

For isobaric tagging, quantifies peaks defined in reporters using method in all spectra of the MSnExp object. If verbose is set to TRUE, a progress bar will be displayed.

For label-free quantitation, the respective quantitation methods and normalisations are applied to the spectra. These methods require two additional arguments (...), namely the protein accession of identifiers (fcol, with detault value "DatabaseAccess") and the protein lengths (plength, with default value "DBseqLength"). These values are available of the identification data had been collated using addIdentificationData.

An object of class "MSnSet" is returned containing the quantified feature expression and all meta data inherited from the MSnExp object argument.

signature(object = "Spectrum", method = "character", reporters = "ReporterIons")

Quantifies peaks defined in reporters using method in the Spectrum object (isobaric tagging only).

A list of length 2 will be returned. The first element, named peakQuant, is a 'numeric' of length equal to length(reporters) with quantitation of the reporter peaks using method.

The second element, names curveStats, is a 'data.frame' of dimension length(reporters) times 7 giving, for each reporter curve parameters: maximum intensity ('maxInt'), number of maxima ('nMaxInt'), number of data points defined the curve ('baseLength'), lower and upper MZ values for the curve ('lowerMz' and 'upperMz'), reporter ('reporter') and precursor MZ value ('precursor') when available.

Author(s)

Laurent Gatto and Sebastian Gibb

References

For details about the spectral index (SI), see Griffin NM, Yu J, Long F, Oh P, Shore S, Li Y, Koziol JA, Schnitzer JE. Label-free, normalized quantification of complex mass spectrometry data for proteomic analysis. Nat Biotechnol. 2010 Jan;28(1):83-9. doi: 10.1038/nbt.1592. PMID: 20010810; PubMed Central PMCID: PMC2805705.

For details about the spectra abundance factor, see Paoletti AC, Parmely TJ, Tomomori-Sato C, Sato S, Zhu D, Conaway RC, Conaway JW, Florens L, Washburn MP. Quantitative proteomic analysis of distinct mammalian Mediator complexes using normalized spectral abundance factors. PNAS. 2006 Dec 12;103(50):18928-33. PMID: 17138671; PubMed Central PMCID: PMC1672612.

Examples

## Quantifying a full experiment using iTRAQ4-plex tagging
data(itraqdata)
msnset <- quantify(itraqdata, method = "trap", reporters = iTRAQ4)
msnset

## specifying a custom parallel framework
## bp <- MulticoreParam(2L) # on Linux/OSX
## bp <- SnowParam(2L) # on Windows
## quantify(itraqdata[1:10], method = "trap", iTRAQ4, BPPARAM = bp)

## Checking for non-quantified peaks
sum(is.na(exprs(msnset)))

## Quantifying a single spectrum
qty <- quantify(itraqdata[[1]], method = "trap", iTRAQ4[1])
qty$peakQuant
qty$curveStats


## Label-free quantitation
## Raw (mzXML) and identification (mzid) files
quantFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
                 full.name = TRUE, pattern = "mzXML$")
identFile <- dir(system.file(package = "MSnbase", dir = "extdata"),
                 full.name = TRUE, pattern = "dummyiTRAQ.mzid")

msexp <- readMSData(quantFile)
msexp <- addIdentificationData(msexp, identFile)
fData(msexp)$DatabaseAccess

si <- quantify(msexp, method = "SIn")
processingData(si)
exprs(si)

saf <- quantify(msexp, method = "NSAF")
processingData(saf)
exprs(saf)

lgatto/MSnbase documentation built on Nov. 12, 2024, 10:58 a.m.