BootList {CNVPanelizer} | R Documentation |
Performs a hybrid bootstrapping subsampling procedure similar to random forest. It bootstraps the reference samples and subsamples the amplicons associated with each gene. Returns a distribution of sample/reference ratios for each gene and sample of interest combination.
BootList(geneNames, sampleMatrix, refmat, replicates)
geneNames |
A vector of gene names, with one entry for each sequenced amplicon. |
sampleMatrix |
A vector or matrix of the read counts from the sample of interest. In the case of a matrix columns represent samples and rows amplicons. |
refmat |
A matrix of the read counts obtianed from the reference samples. Columns represent reference samples and rows amplicons. |
replicates |
How many bootstrap replicates should be performed. |
Returns a list of numeric matrices: For each matrix a row represent a gene while each column represents a bootstrapping/subsampling iteration.
Thomas Wolf, Cristiano Oliveira
data(sampleReadCounts) data(referenceReadCounts) ## Gene names should be same size as row columns geneNames <- row.names(referenceReadCounts) ampliconNames <- NULL normalizedReadCounts <- CombinedNormalizedCounts(sampleReadCounts, referenceReadCounts, ampliconNames = ampliconNames) # After normalization data sets need to be splitted again to perform bootstrap samplesNormalizedReadCounts = normalizedReadCounts["samples"][[1]] referenceNormalizedReadCounts = normalizedReadCounts["reference"][[1]] # Should be used values above 10000 replicates <- 10 # Perform the bootstrap based analysis bootList <- BootList(geneNames, samplesNormalizedReadCounts, referenceNormalizedReadCounts, replicates = replicates)