Sample PCA Plot for Transformed Data

Wrapper for DESeq2::plotPCA() that improves principal component analysis (PCA) sample coloring and labeling.

# S4 method for bcbioRNASeq
plotPCA(object, transform = "rlog", interestingGroups,
  genes, censorSamples, label = FALSE, shape = FALSE, returnData = FALSE)

Arguments

object

Object.
transform

String specifying rlog (recommended) or vst DESeqTransform slotted inside the bcbioRNASeq object.
interestingGroups

Optional. Interesting groups to use for point appearance. If missing, color defaults to all interestingGroups parameters set in the bcbioRNASeq object.
genes

Optional. Character vector of gene identifiers to use.
censorSamples

Optional. Censors to exclude from PCA plot.
label

Optional. Superimpose sample text labels on the plot.
shape

Optional. Make points easier to inspect with differing shapes. Generally this isn't recommended for PCA and works poorly for more than 6 discrete factors, as defined by the interestingGroups argument. We recommend typically leaving this to FALSE.
returnData

Return PCA loadings data instead of plotting.

Value

ggplot.

See also

DESeq2::plotPCA().

Examples

plotPCA(bcb, label = TRUE)
plotPCA(bcb, label = FALSE)

# Manually set interesting groups
# Note these columns aren't present in our example data

# NOT RUN {
plotPCA(bcb, interestingGroups = c("genotype", "treatment"))
# }