Filter Illumina reads & maintain PE ordering

[k-gun12]

Can anyone recommend a tool to filter my paired end Illumina reads (something like mean Q score), either shuffled together in the same file or in separate files, while maintaining the read order? i.e if 1 read of a pair fails, remove it's mate as well (and preferably stick it in a singlets file or something)? I'm aware of the Galaxy web tool but don't want to upload gigs and gigs of sequence data...

Thanks in advance!

[natstreet]

This thread might help. I've been using fastx to filter read1.fq and read2.fq and then the script provided in the thread to make a file of orphaned reads and a read1 and read2 correctly-paired set of files that can then optionally be interleaved. Hopefully tools like fastx can be updated to handle paired-end data out of the box (so to speak) but for now, this works well.

[k-gun12]

Thanks.. I didn't even realize galaxy used the fastx tools. I've only ever found the page for the web interface with no mention of the software.