QualityIO Sequence and quality captions differ

[jfallmann]

I am reporting a problem with Biopython 1.72 , Python 3.7.1 on Fedora 29
When reading in FASTQ files with SeqIO with differing lines 1 (header) and 3 (optional) we get the above error message. We use Biopython to post-process unmapped reads from metagenomics mapping and tend to use line three to add information about runs and metagenomes used, which seems to be a topic of controversial opinions, however, I fail to understand why

biopython/Bio/SeqIO/QualityIO.py

Lines 916 to 920 in e41e53e

	# The title here is optional, but if present must match!
	second_title = line[1:].rstrip()
	if second_title and second_title != title_line:
	raise ValueError("Sequence and quality captions differ.")
	break

is so strict about this. Is this really necessary? And if yes is there a way to circumvent this?
Thanks for any help, highly appreciated

[peterjc]

Because we wrote the "spec" http://dx.doi.org/10.1093/nar/gkp1137 ;)

Practically speaking, do you want access to the plus line information, or would downgrading to a warning be enough?

[jfallmann]

Well, good argument ;)

downgrading to warning would be ok, access would be great but being able to parse at all is most important.
How would I downgrade? And thanks for your help!

[peterjc]

If this is urgent, just comment out that if statement, or to reduce the exception to a warning, replace:

raise ValueError("Sequence and quality captions differ.")

with:

warnings.warn("Sequence and quality captions differ.", BiopythonParserWarning)

In terms of integrating this into Biopython, it might make sense as part of #1812 / #1864 where we're looking at a revised faster FASTQ parser which assumes four lines (it would not support line wrapping, which is almost never used).

[jfallmann]

@peterjc Thanks for forwarding this to the revised FASTQ parser issues, if somebody takes this up it would be great. In the meantime I'll just edit locally. Feel free to close this issue anytime