calcComposition {CAGEfightR} | R Documentation |
For every feature, count in how many samples it is expressed above a certain fraction (e.g. 10 percent) within a grouping, usually genes. This count is refered to as the 'composition' value.
calcComposition(object, inputAssay = "counts", outputColumn = "composition", unexpressed = 0.1, genes = "geneID")
object |
RangedSummarizedExperiment: CAGE data quantified at CTSS, cluster or gene-level. |
inputAssay |
character: Name of assay holding input expression values. |
outputColumn |
character: Name of column in rowRanges to hold composition values. |
unexpressed |
numeric: Composition will be calculated based on features larger than this cutoff. |
genes |
character: Name of column in rowData holding genes (NAs are not currently allowed.) |
object with composition added as a column in rowData.
Other Calculation functions: calcBidirectionality
,
calcPooled
, calcShape
,
calcSupport
, calcTPM
,
calcTotalTags
,
subsetByBidirectionality
,
subsetByComposition
,
subsetBySupport
data(exampleUnidirectional) # Annotate clusters with geneIDs: library(TxDb.Mmusculus.UCSC.mm9.knownGene) txdb <- TxDb.Mmusculus.UCSC.mm9.knownGene exampleUnidirectional <- assignGeneID(exampleUnidirectional, geneModels=txdb, outputColumn='geneID', swap='thick') # Calculate composition values: exampleUnidirectional <- subset(exampleUnidirectional, !is.na(geneID)) calcComposition(exampleUnidirectional) # Use a lower threshold calcComposition(exampleUnidirectional, unexpressed=0.05, outputColumn='lenientComposition')