normalizeWithinArraysByOptimalLowess: Optimal LOWESS Within-Slide Normalization of a MAList

In adele/omicsMA: Variance-Preserving Estimation and Normalization of M-A Values

Description

Applies LOWESS for within-slide normalization of the MA values. The parameters are set by a data-driven parameter selection method, proposed by \insertCiteribeiro2019omicsMAomicsMA, which is parsimonious and considers intrinsic characteristics of microarray data, such as heteroskedasticity. Particularly, the best bandwidth is selected according to the HRCp criterion \insertCiteliu2013heteroscedasticityomicsMA.

Usage

normalizeWithinArraysByOptimalLowess(MA, eva.values = seq(0.2, 1, by =
  0.05), debug = TRUE, save.objs = TRUE, save.plots = TRUE,
  dir.to.save = "./")

Arguments

MA	A MAList object, as in the limma R package, with the non-normalized MA values.
eva.values	A vector with values between 0 and 1 corresponding to the bandwith values to be considered by the LOWESS parameter selection method.
debug	A logical value indication if you want to view logs of the execution.
save.objs	A logical value indicating if you want to save objects with partial results.
save.plots	A logical value indicating if you want to generate the M plots illustrating the bandwidth parameter selection process.
dir.to.save	Path to the folder you want to save the output objects.

Value

A MAList object, as in the limma R package, with the MA values after applying within-slide normalization by LOWESS with optimal parameter settings.

References

\insertAllCited

Examples

data(metaplasia)
MA <- estimateMAValues(metaplasia$R.mean, metaplasia$G.mean,
                       metaplasia$R.var, metaplasia$G.var, metaplasia$RG.cov,
                       metaplasia$R.bckg, metaplasia$G.bckg,
                       array.ids=metaplasia$array.ids, gene.ids=metaplasia$gene.ids)
normMA <- normalizeWithinArraysByOptimalLowess(MA)

adele/omicsMA documentation built on May 24, 2019, 1:32 a.m.