thanks Dr. Albert. I suspected as much. some strange Trinity normalization bug I think.
first I am going to answer my own question in so far as I have already run seqtk on interleaved fq file using default Mott algorithm. seemed to work fine!
then I ran into oddball occurences with trinity normalization after I split trimmed interleaved file into .1 and .2 files and fed to trinity for normalization. normalization program ran way too fast and practically obliterated my reads (from 33 GB for .1 fq file to 150 MB for .1 normalized file??)
so then I took interleaved file, split into .1 and .2 reads, then ran seqtk for trimming, and, bam, trinity normalization works fine.
so back to my original question, other than this experience, is there any reason why seqtk cannot be run on interleaved file? (it not only ran but seemed to trim appropriately, keep pairing, etc)
1 answer
Seqtk does not actually remove reads, it just trims them back as much as it is necessary or to a minimal lenght (whichever comes first).
So it should work just fine with both cases.
Log in to answer this question.