DRomics (version 2.6-2)

PCAdataplot: Performs and plots the results of a PCA on omic data

Description

Provides a two dimensional plot (two first components) of a principal component analysis (PCA) performed on omic data after normalization and/or transformation, to check the promiximity of samples exposed to the same dose and optionally the presence/absence of a potential batch effect.

Usage

PCAdataplot(omicdata, batch, label)

Value

a ggplot object.

Arguments

omicdata

An object of class "microarraydata", "RNAseqdata" or "continuousomicdata" respectively returned by functions microarraydata, RNAseqdata or continuousomicdata.

batch

Optionnally a factor coding for a potential batch effect (factor of length the number of samples in the dataset).

label

Could be FALSE (default choice), TRUE or a character vector defining the sample names. In the two last cases, the points are replaced by labels of samples (so the batch cannot be identified by the shape of points, but may appear in the sample names.

Author

Marie-Laure Delignette-Muller

Examples

# (1) on a microarray dataset
# 
datafilename <- system.file("extdata", "transcripto_very_small_sample.txt", 
  package="DRomics")
o <- microarraydata(datafilename, check = TRUE, norm.method = "cyclicloess")
print(o)
plot(o)
PCAdataplot(o)
PCAdataplot(o, label = TRUE)
samplenames <- paste0("sample", 1:ncol(o$data))
PCAdataplot(o, label = samplenames)


# \donttest{
# (2) an example on an RNAseq dataset with a potential batch effect 
#
data(zebraf)
str(zebraf)
data4DRomics <- formatdata4DRomics(signalmatrix = zebraf$counts, 
                           dose = zebraf$dose)
o <- RNAseqdata(data4DRomics, transfo.method = "vst")
PCAdataplot(o, batch = zebraf$batch)
PCAdataplot(o, label = TRUE)

# }