length scaled TPM counts in DEseq

Question

[mangchen922]

Hi，all

I was running a pipeline that contains Salmon and Tximport to generate length scaled TPM with the Tximport argument (countsFromAbundance="lengthScaledTPM").

Now I have a tsv for this file. How could I continue? Since I do not have the Txi file, I was not able to use the DESeqDataSetFromTximport function directly. (when I tried to use this file as txi argument, it shows the error in txi$counts : $ operator is invalid for atomic vector)

Appreciated any help!

Answer 1

[ATpoint]

You can use DESeqDataSetFromMatrix. The counts in it are adjusted for average length, see for details the manual: https://bioconductor.org/packages/release/bioc/vignettes/tximport/inst/doc/tximport.html

The difference from "classic" tximport pipeline is that in the classic mode the average length would be returned as a matrix which would be used by DESeq2 to produce offsets for the model rather than using the adjusted counts from lengthScaledTPM directly. Either approach is fine:

We could alternatively generate counts from abundances, using the argument countsFromAbundance, scaled to library size, "scaledTPM", or additionally scaled using the average transcript length, averaged over samples and to library size, "lengthScaledTPM". Using either of these approaches, the counts are not correlated with length, and so the length matrix should not be provided as an offset for downstream analysis packages. As of tximport version 1.10, we have added a new countsFromAbundance option "dtuScaledTPM". This scaling option is designed for use with txOut=TRUE for differential transcript usage analyses. See ?tximport for details on the various countsFromAbundance options.

[mangchen922]

Thank you for your help. My question is DESeqDataSetFromMatrix requires integer but the length scaled TPM has lots of genes with low abundance below 1. Then round(counts) will be 0 in a lot of genes. How could I get over this? Thanks.