query.methodsNormalization: Run multiple normalization methods

In jhsiao999/ashbun: Methods and evaluation tools for single cell RNA-seq count data

Usage

query.methodsNormalization(counts, condition, methodsNormalize = c("LIB",
  "TMM", "RLE", "census", "SCnorm", "scran"))

Arguments

counts	Gene by sample expression count matrix (G by N). Use filtered count data.
condition	Binary vector of length N indicating sample biological condition.
methodsNormalize	Chararacter vector of evaluted methods. To run all methods, use c("normalize.cpm", "normalize.tmm", "normalize.rle", "normalize.census", "normalize.scnorm", "normalize.scran")

Value

libsize_factors List of multiple size factors.

Author(s)

Chiaowen Joyce Hsiao

Examples

ipsc_eset <- get(load(system.file("testdata", "HumanTungiPSC.rda", package = "ashbun")))
counts <- exprs(ipsc_eset)[sample(nrow(exprs(ipsc_eset)), 500), ]
condition <- pData(ipsc_eset)$replicate

----- Step 1: filtering
counts_filtered <- filter.excludeAllZeros(counts)
featuresToInclude <- filterFeatures.fractionExpressed(counts_filtered,
                                                     thresholdDetection = 1,
                                                     fractionExpressed = .01)$index_filter

samplesToInclude <-  filterSamples.fractionExpressed(counts_filtered,
                                                     thresholdDetection = 1,
                                                     fractionExpressed = .01)$index_filter

counts_filtered <- counts_filtered[featuresToInclude, samplesToInclude]
condition_filtered <- condition[samplesToInclude]

---- Step 2: compute library size factors
sizefactors <- query.methodsNormalization(counts_filtered,
                                          condition = condition_filtered,
                                          methodsNormalize = c("TMM", "RLE",
                                                                "census","scran"))

jhsiao999/ashbun documentation built on May 8, 2019, 11:17 p.m.