Just for the record, the developer of

SortMeRNA

just confirmed that the tool factors in the total number of reads to determine hit significance, and as such the current situation is: “_The algorithm is just not designed to work for such marginal test cases (15M of the same very short reads)” ._ Rather unfortunately, however, Ribo-seq exactly is this marginal test case. They do have plans to improve the algorithm to meet newer test cases in the future, but for the time being we should indeed switch over to a

Bowtie2

and/or

RiboDetector

based filtering. Issue is here if anyone wants to learn more about the ‘feature’ of the algorithm: https://github.com/sortmerna/sortmerna/issues/453#issuecomment-3612239952

Not a dev here, but I just saw some discussion on bsky about this: https://github.com/bede/deacon It would obviously need testing to see if it's even helpful in this context but I wonder if a rRNA based index and depletion this way would be faster?

I can add some documentation to reflect the current state before the next sortmerna release