recently I received quite big dataset of three pair-end libraries (350, 550, 700 is) and two mate-pairs libraries (Nextera, 3000 and 5000 is). I ve started with Trimmomatic creating really nicely trimmed pair-ended libs. This procedure failed badly for mate-pairs libraries, due to principle of mate-pair sequencing. So, I ve took quite recent and very nicely looking NxTrim, which does exactly what I need (cutting nextera adapters and sorting reads), but it outputs all mp and pe reads in one file. So, I assume, that mate pairs are coded in the headers of reads, but I am wondering how... And how to sort reads to R1 and R2 files respectively. Because, I still need to trim sequencing adapters from the sorted mate-pairs libraries and Trimmomatic expects paired reads in separated files on the input.
So, does anybody know, how exactly are paired reads recognised?