FASTA files with very large unwrapped records generate exceptions in agat_sp_extract_sequences.pl

[pmagwene]

I've encountered a bug in AGAT's agat_sp_extract_sequences.pl function (latest version from Conda as of 15 July 2021) where a Fasta input file that has a large entry with unwrapped lines generates the following error:

------------- EXCEPTION: Bio::Root::Exception -------------
MSG: Failed validation of sequence '[unidentified sequence]'. Invalid characters 

When I reformatted the Fasta file so that lines have a maximum of 70 chars the run completed without exception.

I suspect this is hitting some character limit in Perl string size or something similar?

I can provide sample files that raise this exception, but I haven't had a chance to trim them down to a MWE, so they're somewhat large.

[pmagwene]

I'm guessing this is the same issue discussed in #56

[Juke34]

Difficult to say, I don't get why the bioperl message is not MSG: Each line of the file must be less than 65,536 characters. Line 2 is 67824 chars. then.
The problem may be something related.
Someone had a similar issue, see #37.

Happy to hear you found a way to solve this.

[pmagwene]

Would be useful to add a note to the documentation regarding wrapping of FASTA files as #37 and #56 seem to suggest that this isn't a rare hiccup folks encounter.

[Juke34]

Good point, something to add in the Troubleshooting section.

Thank you for your feedback