bakedpi: Process raw data to compute density estimate.

In hansenlab/yamss: Tools for high-throughput metabolomics

Description

The bakedpi method stands for bivariate approximate kernel density estimation for peak identification. It performs background correction, retention time correction, and bivariate kernel density estimation.

Usage

bakedpi(cmsRaw, dbandwidth = c(0.005, 10), dgridstep = c(0.005, 1),
  outfileDens = NULL, dortalign = FALSE, mzsubset = NULL, verbose = TRUE)

Arguments

cmsRaw	An object of class CMSraw.
dbandwidth	A length-2 vector indicating the kernel density bandwidth in the M/Z and retention time (scan) directions. Default: c(0.005,10)
dgridstep	A length-2 vector indicating the grid step sizes. Default: c(0.005,1).
outfileDens	Name of a file to save density estimate. If NULL, no output is saved.
dortalign	A logical value. Should retention time correction be performed?
mzsubset	A length-2 vector indicating a subset of the M/Z range to process. NULL otherwise.
verbose	Should the function be verbose?

Details

bakedpi first performs region-specific background correction. An optional retention time correction step follows in which M/Z region-specific shifts are computed to align the raw data. Next the two-dimensional density estimate is computed. The purpose of this function is to take the raw data read in by readMSdata and perform the steps necessary for bivariate kernel density estimation. The output of this function is used by slicepi to detect peaks and provide peak quantifications.

Value

An object of class CMSproc containing background corrected intensities, the bivariate kernel density estimate, and quantiles of the nonzero values in the density estimate.

Examples

## A very small dataset
data(cmsRawExample)
cmsProc1 <- bakedpi(cmsRawExample,
                    dbandwidth = c(0.01, 10), dgridstep = c(0.01, 1),
                    dortalign = TRUE, mzsubset = c(500,510))

## A longer example which takes a few minutes to run.
## This is still a smaller mz-slice of the full data.
## Not run: 
if (require(mtbls2)) {

data(mtbls2)
filepath <- file.path(find.package("mtbls2"), "mzML")
files <- list.files(filepath, pattern = "MSpos-Ex1", recursive = TRUE, full.names = TRUE)
colData <- DataFrame(sampClasses = rep(c("wild-type", "mutant"), each = 4))
cmsRaw <- readMSdata(files = files, colData = colData, verbose = TRUE)
cmsProc2 <- bakedpi(cmsRaw, dbandwidth = c(0.01, 10), dgridstep = c(0.01, 1),
                   outfileDens = NULL, dortalign = TRUE, mzsubset = c(500, 520))

}

## End(Not run)

hansenlab/yamss documentation built on Feb. 8, 2022, 4:28 p.m.