Identifying and removing offending line in FASTQ

[JackieBadger]

Hi,

I am attempting to joint overlapping paired ends using FLASH.
The sequences have gone though some pre-processing and manipulation.

At around 4.5 milltion sequences the program exits saying that it failed to identify a '+' separator between the sequence and quality read.

Can anyone provide a clever way using Bash/perl to identify which line/sequence is causing the offence?

Hopefully, only one sequence is causing the problem which I can them remove.

[GenoMax]

http://genome.sph.umich.edu/wiki/FastQValidator

[maubp]

You could try parsing it with a relatively strict parser like the one in BioPerl, Biopython, BioRuby or BioJava and seeing what last good read identifier was?