groupFeatures-abundance-correlation {xcms}

R Documentation

Compounding/feature grouping based on similarity of abundances across samples

Description

Features from the same originating compound are expected to have similar intensities across samples. This method this groups features based on similarity of abundances (i.e. feature values) across samples. See also AbundanceSimilarityParam() for additional information and details.

This help page lists parameters specific for xcms result objects (i.e. the XCMSnExp() object). Documentation of the parameters for the similarity calculation is available in the AbundanceSimilarityParam() help page in the MsFeatures package.

Usage

## S4 method for signature 'XCMSnExp,AbundanceSimilarityParam'
groupFeatures(
  object,
  param,
  msLevel = 1L,
  method = c("medret", "maxint", "sum"),
  value = "into",
  intensity = "into",
  filled = TRUE,
  ...
)

Arguments

object	XCMSnExp() object containing also correspondence results.
param	AbudanceSimilarityParam object with the settings for the method. See AbundanceSimilarityParam() for details on the grouping method and its parameters.
msLevel	integer(1) defining the MS level on which the features should be grouped.
method	character(1) passed to the featureValues call. See featureValues() for details. Defaults to method = "medret".
value	character(1) passed to the featureValues call. See featureValues() for details. Defaults to value = "into".
intensity	character(1) passed to the featureValues call. See featureValues() for details. Defaults to intensity = "into".
filled	logical(1) whether filled-in values should be included in the correlation analysis. Defaults to filled = TRUE.
...	additional parameters passed to the groupFeatures method for matrix.

Value

input XCMSnExp with feature group definitions added to a column "feature_group" in its featureDefinitions data frame.

Author(s)

Johannes Rainer

See Also

feature-grouping for a general overview.

Other feature grouping methods: groupFeatures-eic-similarity, groupFeatures-similar-rtime

Examples

library(MsFeatures)
## Load a test data set with detected peaks
data(faahko_sub)
## Update the path to the files for the local system
dirname(faahko_sub) <- system.file("cdf/KO", package = "faahKO")

## Disable parallel processing for this example
register(SerialParam())

## Group chromatographic peaks across samples
xodg <- groupChromPeaks(faahko_sub, param = PeakDensityParam(sampleGroups = rep(1, 3)))

## Group features based on correlation of feature values (integrated
## peak area) across samples. Note that there are many missing values
## in the feature value which influence grouping of features in the present
## data set.
xodg_grp <- groupFeatures(xodg,
    param = AbundanceSimilarityParam(threshold = 0.8))
table(featureDefinitions(xodg_grp)$feature_group)

## Group based on the maximal peak intensity per feature
xodg_grp <- groupFeatures(xodg,
    param = AbundanceSimilarityParam(threshold = 0.8, value = "maxo"))
table(featureDefinitions(xodg_grp)$feature_group)

---