compute_post_hoc_test_with_limma {lipidomeR}

Compute pairwise post-hoc comparisons for a multi-level factor

Description

Use this function to achieve the post-hoc comparisons between the multiple levels of an independent variable. These comparisons follow analysis of variance (ANOVA) or analysis of covariance (ANCOVA). The pairwise comparisons are based on a result of an F-test produced by the function compute_F_test_with_limma. To use this function, first call the functions compute_models_with_limma and compute_F_test_with_limma consecutively.

Usage

compute_post_hoc_test_with_limma(
  x,
  p.adj.threshold = NULL,
  remap.level.names = FALSE
)

Arguments

x

(Required) list of models for which the test will be done. The pairwise comparisons will be computed using the levels of the first independent variable that was specified in the independent.variables argument to the function compute_models_with_limma. The comparisons will be computed for the lipids that had an F-test result of statistical significance from the function compute_F_test_with_limma.

p.adj.threshold

(Optional) numeric value specifying the threshold of statistical significance in the pairwise comparisons after a correction for multiple testing. We recommend to leave this argument unfilled, leading to the same threshold to be used as in the preceding F-test.

remap.level.names

(Optional) TRUE or FALSE: Should the levels of the factor independent variable be re-mapped? This feature can be used to solve a problem with the factor levels, which may occur in some versions of the limma package. We recommend to keep this argument unchanged from the default value.

Value

List x supplemented by the results of the pairwise post-hoc comparisons.

See Also

compute_models_with_limma for the model computation step that is required prior to calling this function.

compute_F_test_with_limma for the F-test step that is required prior to calling this function.