Hi all!
I had a question about kneaddata filtering using the bowtie2 --un-conc option (which I noticed kneaddata uses when running on paired-end reads). From my understanding, this option writes out reads for which one or both of a pair do not align.
However, in many cases, we want to run more stringent filtering and only keep reads for which both mates failed to align (i.e. if even one pair aligns, it is considered a contaminant).
Does kneaddata have this option? If not, is there any reason why we would not want to do this?