Skip to contents

Pairwise and paired non-parametric tests

Source: R/stats.R
perform_non_parametric.Rd

Performs pairwise and paired non-parametric tests. The default is Mann- Whitney U test, use is_paired for Wilcoxon signed rank tests.

Usage

perform_non_parametric(
  object,
  formula_char,
  is_paired = FALSE,
  id = NULL,
  all_features = FALSE,
  assay.type = NULL,
  ...
)

Arguments

object

a SummarizedExperiment or MetaboSet object

formula_char

character, the formula to be used in the tests

is_paired

logical, use paired test

id

character, name of the subject identification column for paired version

all_features

should all features be included in FDR correction?

assay.type

character, assay to be used in case of multiple assays

...

other parameters passed to test wilcox.test

Value

A data frame with the results.

Details

P-values of each comparison are corrected separately from each other. The model is fit on combined_data(object). Thus, column names in pheno data can be specified. To make the formulas flexible, the word "Feature" must be used to signal the role of the features in the formula. "Feature" will be replaced by the actual features during model fitting. For example, if testing for equality of means in study groups, use "Feature ~ Group".

See also

wilcox.test

Examples

data(example_set)
# Including QCs as a study group for example for pairwise tests
mann_whitney_results <- perform_non_parametric(example_set, 
  formula_char = "Feature ~ Group")
#> INFO [2025-06-23 22:37:32] Starting tests for A & B
#> INFO [2025-06-23 22:37:32] Tests performed
#> INFO [2025-06-23 22:37:33] Starting tests for A & QC
#> INFO [2025-06-23 22:37:33] Tests performed
#> Joining with `by = join_by(Feature_ID)`
#> INFO [2025-06-23 22:37:33] Starting tests for B & QC
#> INFO [2025-06-23 22:37:34] Tests performed
#> Joining with `by = join_by(Feature_ID)`
# Using paired mode (pairs with QC are skipped as there are no common IDs in 
# 'example_set')
wilcoxon_signed_results <- perform_non_parametric(example_set,
  formula_char = "Feature ~ Time",
  is_paired = TRUE,
  id = "Subject_ID")
#> INFO [2025-06-23 22:37:34] Starting paired tests for 1 & 2
#> INFO [2025-06-23 22:37:34] Found 20 complete pairs.
#> INFO [2025-06-23 22:37:34] Paired tests performed.
#> INFO [2025-06-23 22:37:34] Starting paired tests for 1 & QC
#> INFO [2025-06-23 22:37:34] Found 0 complete pairs.
#> Warning: Skipped 1 & QC: no common IDs.
#> Joining with `by = join_by(Feature_ID)`
#> INFO [2025-06-23 22:37:34] Starting paired tests for 2 & QC
#> INFO [2025-06-23 22:37:34] Found 0 complete pairs.
#> Warning: Skipped 2 & QC: no common IDs.
#> Joining with `by = join_by(Feature_ID)`
# Only two groups
mann_whitney_results <- perform_non_parametric(drop_qcs(example_set), 
  formula_char = "Feature ~ Group")
#> INFO [2025-06-23 22:37:34] Starting tests for A & B
#> INFO [2025-06-23 22:37:35] Tests performed