CVEval: Method for the calculation of a correlation vector

In compmedlab/MCbiclust: Massive correlating biclusters for gene expression data and associated methods

Description

Upon identifying a bicluster seed with FindSeed, one of the next steps is to identify which genes not in your chosen gene set are also highly correlated to the bicluster found. This is done by CVEval, and the output is known as the correlation vector.

Usage

CVEval(gem.part, gem.all, seed, splits)

Arguments

gem.part	Part of gene expression matrix only containing gene set of interest with genes as rows and samples as columns
gem.all	All of gene expression matrix
seed	Seed of highly correlating samples
splits	Number of cuts from hierarchical clustering

Details

CVeval uses hierarchical clustering to select the genes most representative of the bicluster and then uses the average expression of these genes across the sample seed and calculates the correlation of every gene measured across the sample seed to this average expression value.

The correlation vector is the output of this calculation.

Value

Correlation vector

Examples

data(CCLE_small)
data(Mitochondrial_genes)

mito.loc <- (row.names(CCLE_small) %in% Mitochondrial_genes)
CCLE.mito <- CCLE_small[mito.loc,]

set.seed(102)
CCLE.seed <- FindSeed(gem = CCLE.mito,
                      seed.size = 10,
                      iterations = 100,
                      messages = 1000)

CCLE.sort <- SampleSort(gem = CCLE.mito,seed = CCLE.seed,sort.length = 11)

# Full ordering are in Vignette_sort in sysdata.rda
CCLE.samp.sort <- MCbiclust:::Vignette_sort[[1]]

CCLE.pc1 <- PC1VecFun(top.gem = CCLE.mito,
                      seed.sort = CCLE.samp.sort,
                      n = 10)

CCLE.cor.vec <-  CVEval(gem.part = CCLE.mito,
                            gem.all = CCLE_small,
                            seed = CCLE.seed,
                            splits = 10)

CCLE.bic <- ThresholdBic(cor.vec = CCLE.cor.vec,sort.order = CCLE.samp.sort,
                         pc1 = as.numeric(CCLE.pc1))

CCLE.pc1 <- PC1Align(gem = CCLE_small, pc1 = CCLE.pc1,
                     cor.vec = CCLE.cor.vec ,
                     sort.order = CCLE.samp.sort,
                     bic =CCLE.bic)

CCLE.fork <- ForkClassifier(CCLE.pc1, samp.num = length(CCLE.bic[[2]]))

compmedlab/MCbiclust documentation built on March 9, 2020, 12:14 a.m.