Skip to contents

Perform ANOVA-style differential expression across multiple groups

do_anova.Rd

This function takes a SummarizedExperiment object and uses group labels in colData(object) to test, for each protein, whether mean abundance differs across all levels of condition at once. It uses limma's moderated F-test (empirical Bayes-shrunk variance, as in do_limma_binary) rather than a classical per-protein aov(), which is more robust when group sizes are small. condition must have at least 3 groups; for exactly 2 groups, use do_limma_binary instead.

Usage

do_anova(object, condition, covariates = NULL)

# S4 method for class 'SummarizedExperiment'
do_anova(object, condition, covariates = NULL)

Arguments

object

A SummarizedExperiment object containing protein intensities and metadata.

condition

String: column name in colData(object) that holds group labels (>= 3 groups).

covariates

Optional character vector: column names in colData(object) to use as covariates.

Value

A data frame with metadata, intensities, one log fold change column per non-reference group (versus an arbitrary baseline group), the omnibus F statistic, P.Value, and adj.P.Val.

Functions

  • do_anova(SummarizedExperiment): Method for SummarizedExperiment objects.