opts.duplicate: Duplicate Feature Detection Parameters

In hhabra/Combiner: Method for Combining LC-MS Metabolomics Feature Measurements

Duplicate Feature Detection Parameters

Description

Lists the parameters for detection of two or more rows that represent the same entity, based on similar m/z and retention time values.

Usage

opts.duplicate(
  mz = 0.0025,
  rt = 0.05,
  resolve = c("single", "merge"),
  weighted = FALSE
)

Arguments

mz	m/z tolerance for duplicate feature detection
rt	RT tolerance for duplicate feature
resolve	character. Either "single" (default) or "merge".
weighted	logical. Option to weight m/z, RT, Q by mean abundance of each row (TRUE) or take single representative values (FALSE).

Details

The presence of duplicate features has negative consequences for the LC-MS alignment task. The package offers several options for resolving the issue of feature duplication. Pairwise m/z and RT tolerances define which features are to be considered as duplicates within a single data set. Setting mz or rt to 0 skips duplicate feature filtering altogether.

When duplicates are detected, either a single master copy is retained (resolve = "single") or merged into a single row (resolve = "merge").The master copy is the copy with lower proportion of missingness, followed by the most abundant (by median or mean). If missingness and abundance is equivalent for duplicates, the first copy that appears is retained. The "merge" option fuses duplicate feature rows, with quantitative descriptors (m/z, RT) either calculated as a weighted average (weighted = TRUE) or otherwise taken from the top representative row; id and adduct values are concatenated; the maximum feature value is used for each sample; and all 'extra' values are taken from the 'master copy' row, similar to the "single" option.

Examples

data(plasma20)
pars.duplicate <- opts.duplicate(mz = 0.01, rt = 0.05, resolve = "single")
p20 <- metabData(plasma20, samples = "Red", duplicate = pars.duplicate)

#to prevent removal of duplicate features
p20 <- metabData(plasma20, samples = "Red", duplicate = opts.duplicate(0))

##merge option
pars.duplicate <- opts.duplicate(mz = 0.01, rt = 0.05, resolve = "merge")
p20 <- metabData(plasma20, samples = "Red", duplicate = pars.duplicate)

hhabra/Combiner documentation built on June 5, 2024, 10:56 a.m.