DEGordering: Variable ordering according to the p-values returned by...
In elqumsan/RNAseqMVA: Application of Multivariate Analysis methods to RNA-seq Data

Description Usage Arguments Details Value Author(s) Examples

Order the variables (columns) of a count table in preparation for supervised classification, by running differential expression analysis with either DESeq2 or edgeR.

DEGordering(
  countDataset,
  method = "DESeq2",
  permuteLabels = FALSE,
  alpha = 0.05
)

`countDataset`	an object of class DataTableWithClasses containing counts of reads per features. Importantly, edgeR and DESeq2 require non-normalized counts as input.
`permuteLabels=FALSE`	if TRUE, class labels are permuted before running the differential analysis, in order to run a negative control (empirical approximation of the null hypothesis)
`method="DESeq2"`	choice of method for differential expression analysis. Supported: "DESeq2" , "edgeR".
`alpha=0.05`	threshold on adjusted p-value (FDR) to call genes positive
`...`	all additional parmeters are passed to the differential expression method (DEseq2 or edgeR).

###################################################

a list with the following fields.

method: method specified in the function call
permuteLabels: Boolean variable indicating whether the test was led with permuted class labels (negative control)
classLabels: vector with the class labels used for the analysis (the ones from the original data, or the permuted ones if the option permuteLabels was TRUE)
geneOrder: a vector of gene names ordered by increasing p-value.
DEGtable a table with one row per gene, and one column per DEG statistics (mean, log-ratio, nominal and adjusted p-values, ...)
orderedDataTable ordered count table where rows (genes) have been ordered by increasing p-value.

Note that genes are ordered according to nominal p-value (pvalue) rather than adjusted p-value (padj) because DESeq2 produces NA values for the padj.

Mustafa AbuElQumsan and Jacques van Helden

###################################################
## loading required packages

recountID <- "SRP048759"
recountID <- "SRP042620"
studyCases <- loadCounts(recountID = recountID, parameters = project.parameters[[recountID]])
filteredCounts <- studyCases[[recountID]]$datasetsForTest$filtered
degOrderdPValues <- DEGordering(countDataset = filteredCounts, method = "edgeR")
## degPValues<- degPValues[order(degPValues$padj) ,]