Use TPM to compare between the samples

Question

[Kunal]

After going through multiple community posts and papers, I understand that TPM cannot be used for between sample analysis. But currently I am stuck with TPM values and don't have the raw sequence data. Then, I have to compare the transcripts between diseased and healthy samples. What could be the best possible way to move forward by limiting the amount of errors with this data?

Answer 1

[ATpoint]

If you have no alternative then use limma, see:

https://support.bioconductor.org/p/56275/#56299

TPM might or might not properly correct for changes in library composition, see:

TMM-Normalization

Be aware that for a publication you need the raw data, so better try to dig them out :)

[Kunal]

Thanks, I'll explore limma.