fastx_barcode_splitter

[biofreak]

I used fastx_barcode_splitter.pl to demultiplex the barcodes to generate separate fastq files. However the files still retain the barcodes (at the beginning of each sequence in my case) . I want to remove those before the alignment.
First, is there an option in fastx_barcode_splitter.pl to remove the barcodes while splitting the file? if not, I tried using:
fastx_trimmer -f 5 -l 50 -i input.fastq -o output.fastq
it gives the following error:
fastx_trimmer: Invalid quality score value (char '#' ord 35 quality value -29) on line 4

I used python script python trimmer.py -f input.fastq -s 5 -e 50 > try.fastq
but it trims every line (not just the sequence line) leading to wrong fastq format.

please help.
thanks

[ilenk]

Hi biofreak,
I have exactly the same problem. Did you find any solution to the problem?

[jme]

You need to offset the Q scores with the -Q33 option, so instead of

fastx_trimmer -f 5 -l 50 -i input.fastq -o output.fastq

try

fastx_trimmer -Q33 -f 5 -l 50 -i input.fastq -o output.fastq

I can't believe this isn't in their documentation, but it makes everything work!

[ilenk]

thanks jme,
but I should be more specific (and probably read the post carefully ;-) )
My problem is:
I want demultiplex illumina reads according to a set of 64 barcodes. I have stripped off the adapters from 3'. Subsequently I actually used sabre to demultiplex, and it worked smoothly. However, I'm unsure, whether the script retains the barcode or removes it. After looking into the fastx_barcode_splitter.pl code, I must conclude that barcode seqs are retained.
as to your answer: my problem was not the qual settings, thanks anyway.

[jme]

Sorry, my fault for skimming, just saw the error!