tpm rpkm

    2. # countMatrix is a matrix of read counts from featureCounts or other similar tools
    3. # geneLength is a data frame with first column as gene id and second column as gene length
    5. # TPM ---------------------------------------------------------------------
    6. # 20190505 | update on 20200817
    7. TPM <- function(countMatrix, geneLength) {
    8.   library(tidyverse)
    9.   library(data.table)
    10.   geneLength <- geneLength %>% 
    11.     as.data.table() %>% 
    12.     set_names(c("geneID","gLength"))
    13.   # sort gene order 
    14.   geneLength <- geneLength[match(rownames(countMatrix), geneLength$geneID),] 
    15.   # calculate tpm
    16.   tmp <- countMatrix/(geneLength$gLength / 1000)
    17.   col_sum_perM <- colSums(tmp) / 1e6
    18.   tmp1 <- sweep(tmp, 2, col_sum_perM, `/`)
    19. }
    20. # RPKM --------------------------------------------------------------------
    21. # 20190604
    22. RPKM <- function(countMatrix, geneLength) {
    23.   library(tidyverse)
    24.   library(data.table)
    25.   geneLength <- geneLength %>% 
    26.     as.data.table() %>% 
    27.     set_names(c("geneID","gLength"))
    28.   # sort gene order 
    29.   geneLength <- geneLength[match(rownames(countMatrix), geneLength$geneID),] 
    30.   # calculate rpkm
    31.   col_sum_perM <- colSums(countMatrix) / 1e6
    32.   tmp <- sweep(countMatrix, 2, col_sum_perM, `/`)
    33.   tmp1 <- tmp/(geneLength$gLength / 1000)
    34. }