extract reads matching barcodes from fastq file?

[odoyle81]

Hi,
I'm trying to extract only certain barcoded reads from a multiplexed illumina run. I tried using the the fastx barcode splitter, but my barcodes are different lengths so it errors out:

Error: found barcodes in different lengths. this feature is not supported yet.

Are there other tools for this?
I started to write a python script, but this is more difficult than I thought.

Thanks!

[GenoMax]

Are these barcodes "inline" i.e. part of the actual sequence?

[odoyle81]

Yes, they are inline only - not in the header.

[odoyle81]

I started on a python script.. maybe others can comment on whether this will work?
I haven't thoroughly tested it yet..

import sys, os
from Bio.SeqIO.QualityIO import FastqGeneralIterator
def extract_reads(input_filename, barcodes):
        #create dictionary of barcodes
        barcode_dict = {}
        #ensure that file is closed after
        with open(barcodes, 'r') as barcodefile:
                # open barcode and create dictionary
                for line in barcodefile:
                        (barcode, sampleID) = line.split()
                        barcode_dict[barcode] = sampleID
        #search fastq file for matching barcodes, and dump into new files with sampleID names
        with open(input_filename, 'r') as fastqfile:
                for title, seq, qual in FastqGeneralIterator(fastqfile):
                        for barcode, sampleID in barcode_dict.iteritems():
			    if seq.startswith(barcode):
                                with open(sampleID + ".fq", "a") as outputfile:
					outputfile.write("@%s\n%s\n+\n%s\n" % (title,seq,qual))

[GenoMax]

I am not sure if this script will work but give it a try: http://creskolab.uoregon.edu/stacks/...ss_radtags.php

If you are in a position to ask the sequence provider to de-multiplex that data again (even inline barcodes can be done as described in this post: http://seqanswers.com/forums/showthread.php?t=18692) then that may be another possible option.

[odoyle81]

Thanks. Stacks is a great program but it is also limited by not being able to process barcodes of different lengths. You can of course create multiple barcode files and run sequentially, but it is not ideal.

[hkm128]

Hi odoyle81,

Did you find an answer to multiplexed illumina run with barcodes of different length?. If so please update?

Thanks,
hkm128

[odoyle81]

As I recall, I created separate keyfiles and processed all barcodes of the same length, and then the next set, etc, like that.. not ideal, but not too much effort..