Trim FastQ

[nxtgenkid10]

Hi all, I would Like to know how could i trim a Fastq files and eliminate the reads lesser than QV30

[gringer]

Hi all, I would Like to know how could i trim a Fastq files and eliminate the reads lesser than QV30

FASTQ quality trimmer will do this:

http://hannonlab.cshl.edu/fastx_toolkit/

If you want a more hand-holding approach, I suppose you could use Galaxy instead.

[giampe]

Hi Gringer,
I'm writing to ask some question about FastQ quality trimmer.
I have loaded Fastx toolkit on my linux machine, I want to use that tool to trim bases from the 3' and 5' and with a low quality score <28.
I have a paired end data files of 101 bases, only in the R2 file I have a decrease of quality in the 3' and in the 5' end, I attached the relative quality box plot.
How should I set the parameter reported below of Fastq quality trimmer to obtain a good quality box plot as for R1 (attached)?

fastq_quality_trimmer -h
usage: fastq_quality_trimmer [-h] [-v] [-t N] [-l N] [-z] [-i INFILE] [-o OUTFILE]
Part of FASTX Toolkit 0.0.13 by A. Gordon ([email protected])

[-h] = This helpful help screen.
[-t N] = Quality threshold - nucleotides with lower
quality will be trimmed (from the end of the sequence).
[-l N] = Minimum length - sequences shorter than this (after trimming)
will be discarded. Default = 0 = no minimum length.
[-z] = Compress output with GZIP.
[-i INFILE] = FASTQ input file. default is STDIN.
[-o OUTFILE] = FASTQ output file. default is STDOUT.
[-v] = Verbose - report number of sequences.
If [-o] is specified, report will be printed to STDOUT.
If [-o] is not specified (and output goes to STDOUT),
report will be printed to STDERR.

I will appreciate your help.
Thanks

[giorgifm]

You can find a complete list of FASTQ trimmers, compared to each other, in this paper:

"An Extensive Evaluation of Read Trimming Effects on Illumina NGS Data Analysis"

[Brian Bushnell]

You can find a complete list of FASTQ trimmers, compared to each other, in this paper:

"An Extensive Evaluation of Read Trimming Effects on Illumina NGS Data Analysis"

That's not complete; BBDuk isn't in it

Granted, it was not publicly available when the paper was published, but it DID exist. Here's a comparison of mapping error rates after trimming with various trimmers:

http://seqanswers.com/forums/showpost.php?p=133436&postcount=16

[gringer]

FWIW, I'm now using Trimmomatic (which finally has a published paper out) because it has a pipeline-based interface that I find expressive enough for almost all of my potential uses. In particular, you can change the order of operations for different use cases.

http://bioinformatics.oxfordjournals.org/cgi/pmidlookup?view=long&pmid=24695404

It has fairly good statistics in the paper giorgifm linked, but you can probably find a better trimmer for each specific case. I notice that Brian Bushnell hasn't included Trimmomatic in his graph (the Trimmomatic paper suggests "Maximum Information" mode for the best statistics) -- any chance of adding that?

[Brian Bushnell]

The reason I did not include Trimmomatic is because its settings are more complex. Most of them, you can just give a quality cutoff; Trimmomatic requires multiple parameters so it's hard to grade objectively and plot on that graph where the only variable is quality cutoff.

That said, I would be happy to add it, if you can give me a typical command line.

[gringer]

Here's a command line I've used:

ILLUMINACLIP:TruSeq3-PE.fa:2:30:10:1:true LEADING:3 TRAILING:3 SLIDINGWINDOW:4:28

I haven't tried optimising it at all (i.e. by adjusting the sliding window quality cutoff), and have yet to experiment with MAXINFO mode. Replacing the sliding window with maximum information would be something like the following:

ILLUMINACLIP:TruSeq3-PE.fa:2:30:10:1:true LEADING:3 TRAILING:3 MAXINFO:50:0.5

[optimisation of this would involve making sure the minimum length of 50bp is sufficient for a good mapping, then adjusting the score strictness to see how it changes mapping]