unrecognized .txt format

[agc]

Hi,

I recently received data from illumina, and I am unable to determine the file format or how to convert it to a fastq for bwa. I found lots of documentation on a format called qseq, and several conversion scripts, but it doesn't seem to be the format I'm dealing with and the scripts don't work.

The lines in the files look like this:

NAME:3:1:1138:1083#0/1:ATATATGAGTGGGGTATAAACGGAGATTGATTTCGT:abababbababbbb`bbaaa]b_ab`aa]^_a_``^

[maubp]

It looks like SCARF using FASTQ style encoding for the quality scores.

Looks like you can just split the lines on the colon, something like this (in Python) should give you a FASTQ file:

#!/usr/bin/env python
"""
Simple script to generate FASTQ from colon separated SCARF input 
where the quality scores are ASCII encoded (FASTQ like).

Use at the command line, piping the input and output.
"""
import sys
for line in sys.stdin:
    name, seq, qual = line.rstrip("\n").rsplit(":",2)
    assert len(seq) == len(qual), line
    sys.stdout.write("@%s\n%s\n+\n%s\n" % (name, seq, qual))

The same idea would be equally trivial in Perl or your language of choice.

[agc]

Thanks a lot! It seems to be working. Any idea why these files are in this format? They're supposed to be illumina files.

[maubp]

SCARF is one of the many formats used by Solexa/Illumina, it gets confusing