Hi - apologies in advance if this is a really stupid question.

I'm trying to do some differential expression analysis between tumour/control lung RNA-seq data. Unfortunately I only have access to TPM values (not raw counts). I'm aware that this is extremely suboptimal.

My question is, can I get meaningful results using voom (followed by limma analyses) if I set library sizes to 1000000? Intuitively this seems to make sense, since the TPM values are per million, although of course it relies on the assumption that experimental conditions for the RNA-seq were similar.

If the above is a statistical crime, any advice on whether I can do anything meaningful with the TPM values would be appreciated.

Thank you!

See

• Differential expression analysis starting from TPM data

• What can I do if I only have TPM but not raw counts data?