batchCorrection: batchCorrection

In ConesaLab/MultiBaC: Multiomic Batch effect Correction

Description

This function performs the ARSyNbac correction [1] for each omic contained in mulBatchDesign input object.

Usage

batchCorrection(mbac, multiBatchDesign, Interaction = FALSE,
  Variability = 0.9)

Arguments

mbac	mbac object generated by createMbac. PLS models slot must be present.
multiBatchDesign	A list containing the original and predicted omic for each batch. All omics must be present in every batch. Output object of genMissingOmics function
Interaction	Logical. Whether to model the interaction between experimental factors and bacth factor in ARSyN models. By default, FALSE.
Variability	From 0 to 1. Minimum percent of data variability that must be explained for each ARSyN model. By default, 0.90.

Value

Custom mbac object. Elements in a mbac object:

1. ListOfBatches: A list of MultiAssayExperiment objects (one per batch).

2. commonOmic: Name of the common omic between the batches.

3. CorrectedData: Same structure than ListOfBatches but with the corrected data instead of the original.

4. PLSmodels: PLS models created during MultiBaC method performance (one model per non-common omic data type).

5. ARSyNmodels: ARSyN models created during MultiBaC performance (one per omic data type).

6. InnerRelation: Table of class data.frame containing the inner correlation (i.e. correlation between the scores of X (t) and Y (u) matrices) for each PLS model across all components.

References

[1] Nueda MJ, Ferrer A, Conesa A. ARSyN: A method for the identification and removal of systematic noise in multifactorial time course microarray experiments. Biostatistics. 2012;13(3):553-566. doi:10.1093/biostatistics/kxr042

Examples

data('multiyeast')

my_mbac <- createMbac (inputOmics = list(A.rna, A.gro, B.rna, B.ribo, C.rna, C.par),
                       batchFactor = c("A", "A", "B", "B", "C", "C"),
                       experimentalDesign = list("A" =  c("Glu+", "Glu+", "Glu+",
                       "Glu-", "Glu-", "Glu-"),
                       "B" = c("Glu+", "Glu+", "Glu-", "Glu-"),
                       "C" = c("Glu+", "Glu+", "Glu-", "Glu-")),
                       omicNames = c("RNA", "GRO", "RNA", "RIBO", "RNA", "PAR"),
                       commonOmic = "RNA")

my_mbac_2 <- genModelList (my_mbac, test.comp = NULL,
                           scale = FALSE, center = TRUE,
                           crossval = NULL,
                           showinfo = TRUE)
multiBatchDesign <- genMissingOmics(my_mbac_2)
my_finalwise_mbac <- batchCorrection(my_mbac_2,
                                     multiBatchDesign = multiBatchDesign,
                                     Interaction = FALSE,
                                     Variability = 0.9)

ConesaLab/MultiBaC documentation built on Jan. 24, 2022, 5:17 a.m.