cosmiq: cosmiq - main wrapper function
In dajofischer/cosmiq: cosmiq - COmbining Single Masses Into Quantities

Description Usage Arguments Details Author(s) Examples

This is the main wrapper function for the package cosmiq. Every processing step of cosmiq will be calculated during this function, including mass spectra combination, detection of relevant masses, generation and combination of extracted ion chromatograms, detection of chromatographic peaks, Localisation and quantification of detected peaks.

combine_spectra imports each raw file using xcmsRaw and assigns each ion to a previously defined mass vector, which is created using bin size parameter mzbin. This process is repeated for each raw file.

    cosmiq (files, RTinfo=TRUE, mzbin=0.003, 
        center=0, linear=FALSE, profStep=1, retcorrect=TRUE, 
        continuum=FALSE, rtcombine=c(0,0), scales=c(1:10), 
        SNR.Th=10, SNR.area=20, RTscales=c(1:10, seq(12,32,2)), 
        RTSNR.Th=20, RTSNR.area=20, mintr=0.5)

`files`	String vector containing exact location of each file
`RTinfo`	Logical indicating whether retention time should be used or not. If FALSE, the quantitative information will be retrieved as a sum of ion intensities whithin a retention time window. This retention time window
`rtcombine`	Numerical, with two entries. If no retention time information is to be used, `rtcombine` provides the retention time window where ion intensities will be summed for quantification. If `rtcombine = c(0,0)`, All scans will be used
`mzbin`	Bin size for the generation of mass vector
`center`	Indicates number of file which will be used for retention time alignment. If zero, file will be automatically selected based on maximum summed ion intensity of the combined mass spectrum.
`linear`	logical. If TRUE, linear vector will be generated with `mzbin` increments. If FALSE, mass vector will be generated using a non-linear function. This option is recommended for TOF-type mass detectors
`profStep`	step size (in m/z) to use for profile generation from the raw data files.
`retcorrect`	Logical, should retention time correction be used? `default = TRUE`
`continuum`	Logical, is continuum data used? `default = FALSE`
`scales`	Peak Detection of Mass Peaks in the combined mass spectrum: scales for continuous wavelet transformation (CWT).
`SNR.Th`	Peak Detection of Mass Peaks in the combined mass spectrum: Signal to noise ratio threshold
`SNR.area`	Peak Detection of Mass Peaks in the combined mass spectrum: Area around the peak for noise determination. Indicates number of surrounding peaks on the first CWT scale. `default = 20`
`RTscales`	Peak Detection of Chromatographic peaks on the combined extracted ion chromatogram: scales for continuous wavelet transformation (CWT), see also the `peakDetectionCWT` function of the `MassSpecWavelet` package.
`RTSNR.Th`	Peak Detection of Chromatographic peaks on the combined extracted ion chromatogram: Signal to noise ratio threshold
`RTSNR.area`	Peak Detection of Chromatographic peaks on the combined extracted ion chromatogram: Area around the peak for noise determination. Indicates number of surrounding peaks on the first CWT scale. `default = 20`
`mintr`	Minimal peak width intensity treshold (in percentage), for which two overlapping peaks are considered as separated. The default calue is 0.5.

Current data processing tools focus on a peak detection strategy where initially each LCMS run is treated independently from each other. Subsequent data processing for the alignment of different samples is then calculated on reduced peak tables. This function involves the merging of all LCMS datasets of a given experiment as a first step of raw data processing. The merged LCMS dataset contains an overlay and the sum of ion intensities from all LCMS runs. Peak detection is then performed only on this merged dataset and quantification of the signals in each sample is guided by the peak location in the merged data.

See also xcms::retcor.obiwarp and cosmiq::create_datamatrix which executes each step of the cosmiq function.

For running examples please read the package vignette or ?cosmiq::create_datamatrix.

Christan Panse, David Fischer 2014

    ## Not run: 
    # the following lines consume 2m17.877s  
    # on a Intel(R) Xeon(R) CPU X5650  @ 2.67GHz
    library(faahKO)

    cdfpath <- file.path(find.package("faahKO"), "cdf")
    my.input.files <- dir(c(paste(cdfpath, "WT", sep='/'), 
        paste(cdfpath, "KO", sep='/')), full.names=TRUE)

     x <- cosmiq(files=my.input.files, mzbin=0.25, SNR.Th=0, linear=TRUE)

     image(t(x$eicmatrix^(1/4)), 
        col=rev(gray(1:20/20)),
        xlab='rt', 
        ylab='m/z', axes=FALSE )

     head(xcms::peakTable(x$xcmsSet))
    
## End(Not run)