Raw counts to TPM in R

Question

[firestar]

Can someone verify if this R code for converting raw counts to TPM is correct?

#' @title Compute TPM for a read count matrix
#' @param dfr A numeric data.frame of read counts with samples (columns) and genes (rows).
#' @param len A vector of gene cds length equal to number of rows of dfr.
#' 
r_tpm <- function(dfr,len)
{
  dfr1 <- sweep(dfr,MARGIN=1,(len/10^4),`/`)
  scf <- colSums(dfr1)/(10^6)
  return(sweep(dfr1,2,scf,`/`))
}

[Ram]

Do you have a reason for suspecting it's not? Have you tested it? What do your tests reveal?

Answer 1

[ATpoint]

Use this code snippet from Michael Love (DESeq2 developer)

x <- counts.mat / gene.length
tpm.mat <- t( t(x) * 1e6 / colSums(x) )

[firestar]

Cool! I get the same result.

# michael's version
# https://support.bioconductor.org/p/91218/

tpm3 <- function(counts,len) {
  x <- counts/len
  return(t(t(x)*1e6/colSums(x)))
}

Michael's version is much faster despite all the transposes.

[Kevin Blighe]

If an answer was helpful you should upvote it, if the answer resolved your question you should mark it as accepted.