Relative Log Abundance (RLA) Plot — plot_rla

Relative log abundance (RLA) plots show standardized feature abundances across samples. Standardization is done by removing either the within-batch or across-batch median from each feature

RLA plots are useful for visualizing technical effects that impact all features in a similar manner, such as batch effects due to changes in instrument response, pipetting errors, or sample spillage. Unlike plots of raw or normalized abundances, RLA plots are more robust to these types of effects.

Usage

plot_rla_boxplot(
  data = NULL,
  rla_type_batch = c("within", "across"),
  variable = c("intensity", "norm_intensity", "conc", "conc_raw", "area", "height",
    "fwhm"),
  filter_data = FALSE,
  qc_types = NA,
  x_axis_variable = c("run_seq_num"),
  include_feature_filter = "",
  exclude_feature_filter = "",
  plot_range_indices = NA,
  min_feature_intensity = 0,
  y_lim = NA,
  ignore_outliers = FALSE,
  show_batches = TRUE,
  batches_as_shades = FALSE,
  batch_line_color = "#b6f0c5",
  batch_shading_color = "grey93",
  x_gridlines = FALSE,
  linewidth = 0.2,
  base_font_size = 8,
  relative_log_abundances = TRUE
)

Arguments

data: MidarExperiment
rla_type_batch: Character, must be either "within" or "across", defining whether to use within-batch or across-batch RLA
variable: Variable to plot, must be one of "intensity", "norm_intensity", "conc", "conc_raw", "area", "height", "fwhm".
filter_data: Logical, whether to use QC-filtered data
qc_types: QC type(s) to plot. Can be a vector of QC types or NA/NULL to plot all available QC types.
x_axis_variable: Character, variable to use for the x-axis. Must be one of: "run_seq_num", "run_no", "analysis_id", or "timestamp"
include_feature_filter: Regex pattern to select specific features
exclude_feature_filter: Regex pattern to exclude specific features
plot_range_indices: Numeric vector of length 2, specifying the start and end indices of the sequence to be plotted. NA plots all samples.
min_feature_intensity: Numeric, exclude features with overall median signal below this value
y_lim: Numeric vector of length 2, specifying the lower and upper y-axis limits. Default is NA, which uses limits calculated based on ignore_outliers.
ignore_outliers: Logical, whether to exclude outlier values based on 4x MAD (median absolute deviation) fences
show_batches: Logical, whether to show batch separators in the plot
batches_as_shades: Logical, whether to show batches as shaded areas instead of line separators
batch_line_color: Character, color of the batch separator lines
batch_shading_color: Character, color of the batch shaded areas
x_gridlines: Logical, whether to show major x-axis gridlines
linewidth: Numeric, line width used for whiskers of the boxplot
base_font_size: Numeric, base font size for the plot
relative_log_abundances: Logical, whether to use relative log abundances (RLA) or just log-transformed values

Value

A ggplot object representing the RLA plot

References

De Livera et al. (2012) Normalizing and integrating metabolomics data. Analytical Chemistry 10768-10776 DOI: 10.1021/ac302748b De Livera et al. (2015) Statistical Methods for Handling Unwanted Variation in Metabolomics Data. Analytical Chemistry 87(7):3606-3615 DOI: 10.1021/ac502439y