PCA graphs on codon coverage

Description

PCA graphs on codon coverage

Usage

1
codonPCA(data, typeData)

Arguments

data

a list of 2 data.frames: one with the number of times each codon type is found in each ORF and one with the number of reads for each codon in each ORF.

typeData

a character string. It is used as title for the PCA. Ex. typeData="codonCoverage"

Value

a list of length 2: PCA_scores - matrix of the scores on the first 4 principal components. PCA_plots - a list of 5 PCA scatterplots.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
#How to perform a PCA analysis based on codon coverage
#adapted from
#http://stackoverflow.com/questions/20260434/test-significance-of-clusters-on-a-pca-plot
#either get the codon frequency, coverage, and annotation using a function
#such as codonInfo in this package
#or create a list of matrices with the above information
data(codonDataCtrl)
codonData <- codonDataCtrl
codonUsage <- codonData[[1]]
codonCovMatrix <- codonData[[2]]

#keep only genes with a minimum number of reads
nbrReadsGene <- apply(codonCovMatrix, 1, sum)
ixExpGenes <- which(nbrReadsGene >= 50)
codonCovMatrix <- codonCovMatrix[ixExpGenes, ]

#get the PCA on the codon coverage
codonCovMatrixTransp <- t(codonCovMatrix)
rownames(codonCovMatrixTransp) <- colnames(codonCovMatrix)
colnames(codonCovMatrixTransp) <- rownames(codonCovMatrix)

listPCACodonCoverage <- codonPCA(codonCovMatrixTransp,"codonCoverage")
print(listPCACodonCoverage[[2]])
#See aditional examples in the pdf manual