msDenoise: Mother Function for Mass Spectra Denoising

In zeehio/msProcess: Protein Mass Spectra Processing

Description

Denoise spectra with various functions.

Usage

msDenoise(x, FUN="wavelet",
    attach.noise=TRUE, event="Denoising", ...)

Arguments

x	An object of class msSet.
...	Additional arguments to FUN. They are passed unchanged to each call of FUN and include their names. See the help documentation of the specified FUN for details.
FUN	Either an object of class "character" or of class "function". character: A character string denoting the method to use in denoising the data. Supported choices are "wavelet" for wavelet shrinkage, "mrd" for partial summation of a wavelet-based multiresolution decomposition, and "smooth" for robust running medians. Default: "wavelet". function: A user-defined function with an argument list of the form (x, ...) where x is a required argument corresponding to a numeric vector (typically these values will be the intensity values of a mass spectrum). In either case, the additional arguments ... will be passed directly to the specified FUN.

attach.noise	A logical indicating if the noise removed should be attached or not. Default: TRUE.
event	A character string denoting the name of the event to register with the (embedded) event history object of the input after processing the input data. Default: "Denoising".

Value

An object of class msSet, optionally, with the estimated noise attached as element "noise".

Note

If FUN="mrd", an mrd object containing meta information regarding the multiresolution decomposition is attached to the msSet output object for subsequent use by other MRD-based function calls such as msPeak(x, FUN="mrd", ....

See Also

msDenoiseSmooth, msDenoiseWavelet, msDenoiseMRD, matchObject.

Examples

if (!exists("qcset")) data("qcset", package="msProcess")

## denoise a spectrum portion via waveshrink and a
## smoothing function
mz   <- (qcset$mz > 3000 & qcset$mz < 5000)
data <- qcset[mz, 1, drop=FALSE]

## add a little Gaussian noise for illustration
noise  <- rnorm(length(data$intensity), sd=stdev(data$intensity)/3)
xnoise <- data
xnoise$intensity <- data$intensity +
    matrix(noise, ncol=1)

## denoise using the supported routines
z1 <- msDenoise(xnoise, FUN="wavelet")
z2 <- msDenoise(xnoise, FUN="smooth")
z3 <- msDenoise(xnoise, FUN="mrd", levels=4:6)

## create a user-defined (albeit naive) denoising
## function
my.fun <- function(x, wavelet="d4"){
    filt <- wavDaubechies(wavelet=wavelet, norm=FALSE)$scaling
    return(filter(x, filt))
}
z4 <- msDenoise(xnoise, FUN=my.fun, wavelet="s12")

## create a stackplot of the results
z <- list(original=data$intensity[,1],
    noisy=xnoise$intensity[,1],
    waveshrink=z1$intensity[,1],
    smooth=z2$intensity[,1],
    mrd=z3$intensity[,1],
    "my function"=z4$intensity[,1])
wavStackPlot(z, col=seq(along=z), same.scale=TRUE)

zeehio/msProcess documentation built on May 4, 2019, 10:15 p.m.