remove_batcheffect: Removing Batch Effect from Expression Sets

In IOBR: Immune Oncology Biological Research

Removing Batch Effect from Expression Sets

Description

Removes batch effects from expression datasets using sva::ComBat (for microarray/TPM data) or sva::ComBat_seq (for RNA-seq count data). Generates PCA plots to visualize data before and after correction.

Usage

remove_batcheffect(
  eset1,
  eset2,
  eset3 = NULL,
  id_type = "ensembl",
  data_type = c("array", "count", "tpm"),
  cols = "normal",
  palette = "jama",
  log2 = TRUE,
  check_eset = TRUE,
  adjust_eset = TRUE,
  repel = FALSE,
  path = NULL
)

Arguments

eset1	First expression set (matrix or data frame with genes as rows).
eset2	Second expression set.
eset3	Optional third expression set. Use 'NULL' if not available.
id_type	Type of gene ID in expression sets (e.g., '"ensembl"', '"symbol"'). Required for count data normalization.
data_type	Type of data: '"array"', '"count"', or '"tpm"'. Default is '"array"'.
cols	Color scale for PCA plot. Default is '"normal"'.
palette	Color palette for PCA plot. Default is '"jama"'.
log2	Whether to perform log2 transformation. Default is 'TRUE'. Ignored for count data.
check_eset	Whether to check expression sets for errors. Default is 'TRUE'.
adjust_eset	Whether to adjust expression sets by removing problematic features. Default is 'TRUE'.
repel	Whether to add repelling labels to PCA plot. Default is 'FALSE'.
path	Directory where results should be saved. Default is 'NULL' (display only).

Value

Expression matrix after batch correction.

Author(s)

Dongqiang Zeng

References

Zhang Y, et al. ComBat-seq: batch effect adjustment for RNA-seq count data. NAR Genomics and Bioinformatics. 2020;2(3):lqaa078. doi:10.1093/nargab/lqaa078

Leek JT, et al. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882-883.

Examples

# Simulate data
set.seed(123)
sim_eset1 <- matrix(rnorm(100 * 5, mean = 10, sd = 2), 100, 5)
sim_eset2 <- matrix(rnorm(100 * 5, mean = 12, sd = 2), 100, 5)
rownames(sim_eset1) <- rownames(sim_eset2) <- paste0("Gene", 1:100)
colnames(sim_eset1) <- paste0("S1_", 1:5)
colnames(sim_eset2) <- paste0("S2_", 1:5)

# Run batch correction
if (requireNamespace("sva", quietly = TRUE) && requireNamespace("BiocParallel", quietly = TRUE)) {
  eset_corrected <- remove_batcheffect(sim_eset1, sim_eset2, data_type = "tpm")
  if (!is.null(eset_corrected)) head(eset_corrected)
}

IOBR documentation built on May 30, 2026, 5:07 p.m.