Adapter trimming and trimming by quality question

[alisrpp]

Hi to all,

I've been experiencing some problems trimming the adapters and filtering by quality my paired-end Illumina reads.

As fas as i know if i want to use FASTX-TOOLKIT (included in Galaxy) i have to treat my data as single-end, trim the adapter of the forward reads first and then proceed with the reverse reads. And same thing for the filtering by quality. Right?

Does anyone know if i can combine forward/reverse reads in a unique file with the FASTQ joiner (also included in Galaxy) and then proceed with the trimming of the adapters and the filtering by quality? If i do that, am i going to be able to remove the adapters of both directions reads? Am i going to be able to keep the broken pairs?

I was reading that there is another toolkit, NGS QC TOOLKIT that allows you to work with paired-end reads and keep the broken pairs after the trimming and thinning. Did anyone try it? I think that has no galaxy option...

Thanks,

[westerman]

If you want non-Galaxy options then I recommend Trimmomatic.

As for Galaxy, I do not use it on a regular basis however I believe that there have been some comments recently on the Galaxy list as to how run fastx_quality_trimmer on paired-end sequences. There is also at least one published workflow -- look for 'Illumina Sequence Prep w/ Trim'. I think that may do what you want.

[alisrpp]

Thanks, i'm going to try Trimmomatic.
What do you recommend me for the adapter trimming: palindrome or simple.

[JackieBadger]

Trimmomatic is the best tool for this.

[prepagam]

You could also look at Scythe & Sickle. They can be run independently, or added to local source galaxy.
Scythe cuts out adaptors and contamination - but never removes an entire read so your pairs will remain in order.

https://github.com/vsbuffalo/scythe

Sickle does quality based trimming - and maintains the paired sequences order by outputting 3 files - read 1, read 2, and singletons. https://github.com/najoshi/sickle/blob/master/README.md

[ravipatel4]

TrimmingReads.pl tool of NGS QC Toolkit can trim reads from single-end and paired-end sequencing data. It provides two ways to trim the reads: 1) trim specified number of bases from either or both ends of read, 2) trim low quality bases from the 3' end based on give PHRED quality threshold value.

Usually, the base quality decreases as we go towards the 3' end of the reads. Such low quality bases affect the downstream analysis (eg. reference mapping, de novo and reference assembly, etc). Second option of TrimmingReads.pl takes care of this sequence artifact.