Hello,

I have seen tximport will import TPM, estimated counts and effective lengths from transcript quantification tools into a list object.

As far as I know, DESeqDataSetFromTximport() can directly read this list ('txi' in the vignette) to create a DESeq object.

I understand TPM can be used by tximport() to recalculate the original estimated counts using 'lengthScaledTPM' or 'scaledTPM'.

Since DESeq will work on transcript counts, I wanted to know if calculated TPM are going to be somehow used later by DESeq or otherwise they will be completely ignored (I would say no, but wanted to be sure).

Many thanks in advance.

When DESeq2 is used with tximport via DESeqDataSetFromTximport, the TPM data (abundance matrix) is used to calculate a per-gene-and-sample offset. After that only the count and the offset are used. Or, if countsFromAbundance is used when calling tximport, then only the count is used.