Feature quality definitions and utilities

Source: R/main.R, R/utils-exported.R

Functions to define and work with chromatographic peak quality metrics used for feature and feature group quality calculations.

featureQualities(qualities = NULL)

featureGroupQualities(qualities = NULL)

featureQualityNames(feat = TRUE, group = TRUE, scores = FALSE, totScore = TRUE)

Arguments

qualities

A character vector specifying which qualities to return. If NULL, returns all available quality definitions.

feat

If TRUE then names specific to features are returned.

group

If TRUE then names specific to groups are returned.

scores

If TRUE the score names are returned, otherwise the quality names.

totScore

If TRUE (and scores=TRUE) then the name of the total score is included.

Value

For featureQualities and featureGroupQualities: A named list containing quality definitions. Each element contains:

  • func: The MetaClean function to calculate the quality metric

  • HQ: "HV" (high values good) or "LV" (low values good)

  • range: Expected range of values (may be Inf for unbounded ranges)

For featureQualityNames: A character vector with quality and/or score names.

Details

These functions provide access to quality metrics that are used to assess the quality of detected features and feature groups by the calculatePeakQualities function. The quality metrics are calculated using the MetaClean package and are useful for filtering out low-quality features before further analysis.

Feature qualities

The featureQualities function defines quality metrics that are calculated for individual features. The following quality metrics are available:

  • ApexBoundaryRatio - Ratio between apex and boundary intensities

  • FWHM2Base - Full width at half maximum to base ratio

  • Jaggedness - Measure of peak smoothness

  • Modality - Measure of peak multiplicity

  • Symmetry - Measure of peak symmetry (range: -1 to 1)

  • GaussianSimilarity - Similarity to Gaussian distribution

  • Sharpness - Measure of peak sharpness

  • TPASR - Triangle peak area similarity ratio

  • ZigZag - Zig-zag index measure

Feature group qualities

The featureGroupQualities function defines quality metrics that are calculated for feature groups across multiple analyses. The following quality metrics are available:

  • ElutionShift - Measure of retention time consistency across analyses

  • RetentionTimeCorrelation - Correlation of retention times across analyses

References

Chetnik K, Petrick L, Pandey G (2020). “MetaClean: a machine learning-based classifier for reduced false positive peak detection in untargeted LC-MS metabolomics data.” Metabolomics, 16(11). doi:10.1007/s11306-020-01738-3 .