Counts矩阵的标准化方法:TMM和VST、RLOG

  • TMM:The Trimmed Mean of M value by edgeR
  • VST:The variance stabilizing transformation by DESeq2
  • RLOG:The regularized-logarithm transformation by DESeq2

Counts矩阵来源于STAR匹配得到的结果df <- read.csv('GSE123379.csv', row.names = 1)

安装补充包

TMM方法

1
2
3
4
5
6
7
8
9
f_counts2TMM <- function(countsMatrix){
require(edgeR)
TMM <- DGEList(counts = countsMatrix)
TMM <- calcNormFactors(TMM, method = 'TMM')
cpm(TMM, normalized.lib.sizes = TRUE, log=F)
}
countsMatrix <- df[-(1:3)]
TMM <- f_counts2TMM(countsMatrix)
TMM

VST方法

1
2
3
4
5
6
7
8
9
10
11
12
13
f_counts2VST <- function(countsMatrix){
require(DESeq2)
conditions <- factor(rep("Control",ncol(countsMatrix)))
colData_b <- data.frame(row.names = colnames(countsMatrix), conditions)
dds <- DESeqDataSetFromMatrix(countData = countsMatrix,
colData = colData_b,
design = ~ 1)
vsd <- vst(object=dds, blind=T)
assay(vsd)
}
countsMatrix <- df[-(1:3)]
VST <- f_counts2VST(countsMatrix)
VST

RLOG方法

1
2
3
4
5
6
7
8
9
10
11
12
13
f_counts2RLOG <- function(countsMatrix){
require(DESeq2)
conditions <- factor(rep("Control",ncol(countsMatrix)))
colData_b <- data.frame(row.names = colnames(countsMatrix), conditions)
dds <- DESeqDataSetFromMatrix(countData = countsMatrix,
colData = colData_b,
design = ~ 1)
rld <- rlog(object=dds, blind=T)
assay(rld)
}
countsMatrix <- df[-(1:3)]
RLOG <- f_counts2RLOG(countsMatrix)
RLOG

Counts矩阵的标准化方法:TMM和VST、RLOG
https://b.limour.top/1934.html
Author
Limour
Posted on
July 27, 2022
Licensed under