Group features using greedy algorithm

Group features using a greedy algorithm that maximizes group scores based on retention time, m/z, mobility, and intensity similarities.

Usage

groupFeaturesGreedy(feat, ...)

# S4 method for class 'features'
groupFeaturesGreedy(
  feat,
  rtalign = FALSE,
  rtWindow = defaultLim("retention", "medium"),
  mzWindow = defaultLim("mz", "medium"),
  mobWindow = defaultLim("mobility", "medium"),
  scoreWeights = c(retention = 1, mz = 1, mobility = 1, intensity = 1),
  verbose = TRUE
)

Arguments

feat

The features object with the features to be grouped.

...

Further parameters passed to the selected grouping algorithm.

rtalign

Not yet supported. Provided for consistency with other grouping methods.

rtWindow, mzWindow, mobWindow

Numeric tolerances for retention time (seconds), m/z, and mobility, respectively. The scoring terms are normalized to these values. Defaults to defaultLim("retention", "medium"), defaultLim("mz", "medium"), and defaultLim("mobility", "medium"), respectively (see limits).

scoreWeights

Numeric vector specifying the scoring weights. Should contain the following named elements: "retention", "mz", "mobility", and "intensity". Missing elements are defaulted to one.

verbose

if FALSE then no text output will be shown.

Value

An object of a class which is derived from featureGroups.

Details

This function uses greedy to group features. This function is called when calling groupFeatures with algorithm="greedy".

The greedy algorithm is a simple feature grouping algorithm that can work with both HRMS and IMS-HRMS data. The algorithm groups features by iteratively building the best possible groups. Features are processed in order of decreasing intensity. For each feature, candidate groups are formed from all other (ungrouped) features within the specified retention time, m/z and mobility windows. Each candidate group only contains a maximum of one feature per analysis. The candidates are then scored and the group with the lowest overall variations in retention time, m/z, mobility and replicate intensity is then selected. This process is repeated until all features have been assigned to a group. The weights for each of the scoring terms can be configured.

Note

Any links between IMS precursors and IMS features are removed. This can occur e.g. when greedy is used to generate a feature consensus from a post mobility assignment workflow.

See also

groupFeatures for more details and other algorithms.