Correct for Batch Effects — batch

This function uses the removeBatchEffect method from the limma package to adjust for known batch effects in the data.

Usage

batch_correct(object, batch_variable, bio_variables = NULL)

# S4 method for class 'SummarizedExperiment,character'
batch_correct(object, batch_variable, bio_variables = NULL)

Arguments

object: A SummarizedExperiment object. The first assay will be used.
batch_variable: A single character string specifying the column name in colData(object) that contains the batch information.
bio_variables: An optional character vector of column names in colData(object) that correspond to biological variables of interest. These effects will be preserved during the correction. If NULL (the default), only an intercept is protected.

Value

A SummarizedExperiment object with a new assay named "corrected" containing the batch-corrected data.

Examples

# --- Create a sample SummarizedExperiment object ---
set.seed(123)
counts <- matrix(rnorm(100 * 6, mean = 10, sd = 2), nrow = 100, ncol = 6)
colnames(counts) <- paste0("Sample", 1:6)

sample_info <- data.frame(
  row.names = colnames(counts),
  Condition = rep(c("A", "B"), each = 3),
  Batch = rep(c("Batch1", "Batch2"), times = 3)
)

se <- SummarizedExperiment::SummarizedExperiment(assays = list(log_intensities = counts),
                           colData = sample_info)

# --- Run batch correction, preserving the "Condition" variable ---
corrected_se <- batch_correct(se,
                              batch_variable = "Batch",
                              bio_variables = "Condition")
#> Starting batch correction...
#>   - Preserving biological variables: Condition

# --- Inspect the result ---
SummarizedExperiment::assayNames(corrected_se) # Now includes "corrected"
#> [1] "log_intensities"