trim data with trimmomatic

Question

[yaghoub.amraei]

hi all

i have 21 sample from RNA-seq single end. i want trim that with trimmomatic. whether i can use command for.

for f in *.fastq; do java -jar ~/sratoolkit.2.9.6-ubuntu64/apps/Trimmomatic-0.39/trimmomatic-0.39.jar  SE -baseout ${f/.fastq} HEADCROP:30 LEADING:30 TRAILING:30 MINLEN:36; done

Answer 1

[Kristoffer Vitting-Seerup]

You probably do not want to use HEADCROP:30 - that will remove the first 30 nucleotides (!) whereas LEADNING will only remove low quality bases in the 5' end.

In regards to using TRAILING the other option is to filter on the average quality in a window via the SLIDINGWINDOW command. According to this article is better than TRAILING (due to there sometimes being a single semi-low quality base followed by additional high quality bases - this can easily be mapped (as mappers do not require perfect matches) but will be thrown away if using TRAILING).

Other things to pay attention to are:

1. The order of arguments matters as that is the order in which the trimming is performed so you always want to end with MINLEN
2. You can also remove (any) adapter contamination via ILLUMINACLIP.
3. You can find more info on how to use Trimmomatic here.

[caggtaagtat]

Wanted to say the same, that you will lose a lot of good information if you trim undependent on base-calling quality!

Answer 2

[lakhujanivijay]

There are several options and recently I suggested using this shell script:

Shell script to automate trimmomatic for multiple samples