medianScaling: Median centering and scaling

In PYangLab/PhosR: A set of methods and tools for comprehensive analysis of phosphoproteomics data

Median centering and scaling

Description

Median centering and scaling of an input numeric matrix

Usage

medianScaling(mat, scale = FALSE, grps = NULL, reorder = FALSE, assay = NULL)

Arguments

mat	a matrix with rows correspond to phosphosites and columns correspond to samples.
scale	a boolean flag indicating whether to scale the samples.
grps	a string or factor specifying the grouping (replciates).
reorder	To reorder the columns by group (grps). By default (reorder=FALSE), original column order is maintained.
assay	an assay to be selected if mat is a PhosphoExperiment object.

Value

A median scaled matrix

Examples

data('phospho.cells.Ins.sample')
grps = gsub('_[0-9]{1}', '', colnames(phospho.cells.Ins))
phospho.cells.Ins.filtered <- selectGrps(phospho.cells.Ins, grps, 0.5, n=1)

set.seed(123)
phospho.cells.Ins.impute <-
    scImpute(phospho.cells.Ins.filtered,
            0.5,
            grps)[,colnames(phospho.cells.Ins.filtered)]

set.seed(123)
phospho.cells.Ins.impute[,seq(5)] <- ptImpute(
    phospho.cells.Ins.impute[,seq(6,10)],
    phospho.cells.Ins.impute[,seq(5)], percent1 = 0.6,
    percent2 = 0, paired = FALSE)

phospho.cells.Ins.ms <-
    medianScaling(phospho.cells.Ins.impute, scale = FALSE)

PYangLab/PhosR documentation built on Nov. 1, 2024, 1:47 p.m.