unikmer diff & uniqs: Ns

[sihellem]

Amazing tool! Exactly what I was looking for, and performing smoothly. (Is the tool published -I could not find much online?)

I am using unikmer as outlined in this workflow: https://www.biostars.org/p/475263/
I use it between intraspecific FASTQ libraries in order to extract reads (illumina PE150) bearing unique kmer to either libraries.

I have a few questions:

1. how does unikmer treat Ns? I compared one library to its exact reverse complement, and unikmer diff outputted a lot of uniques for both files. I was expecting to find 0 uniques. Could this be due to Ns? How does unikmer count treat Ns?
   EDIT: removing reads with Ns prior to unikmer count produces 0 uniques between a library and its reverse complement.

2. when extracting reads, does unikmer uniqs extract sequence bearing ONLY kmer unique to one library? Playing with the -m argument, I could see that when a reads holds two unique kmer for one library, the read is broken in two individual sequences. One could imagine that one single read could bear one kmer unique to indiv_1, another to indiv_2.
   Therefore, does unikmer uniqs would really output reads bearing ONLY kmer unique to indiv_1 (and none from indiv_2)?