You dont want to be discarding too much too early, typically people will discard readlengths later in the analysis when they see they are not periodic (i.e not ribosome protected fragments). For point 2, change the

-k

and report back. I'd be interested to hear whats going on there.

Attempted to do so earlier but had issues with my HPC. Will report back as soon as they're resolved.

Hi @Jack Tierney, I was having the same problem with salmon not assigning any fragments to the transcripts. I have tried adjusting the kmer size to a lower value (23) for a new index, yet I still encounter the same problems. Do you know how I can solve this issue? Thanks!

Do they align when you use other aligners? Like a true rather than psuedo aligner?

Hmm I am checking the output from STAR and I see that the % of uniquely mapped reads was 0% and read length was around 75-79 (isn’t it a bit big for ribosome-protected reads?). I think these numbers look a bit abnormal? What do you think?

I think you haven't successfully trimmed the reads. The presence of adapters/umis etc would lead to this. Can you check those outputs?