[Bioperl-l] Re: SeqIO::swiss.pm + "gsequence"

[Kris Boulez]

Quoting Kris Boulez (krbou@pgsgent.be):
> Quoting hilmar.lapp@pharma.Novartis.com (hilmar.lapp@pharma.Novartis.com):
> > 
> > 
> > Hmm. What do others think / have encountered?
> > 
> > Could you volunteer to perform a couple of tests for various programs
> > reading FastA format files to see what they do upon the empty seq and
> > whether they choke? I'm not sure whether it's a good idea to produce files
> > that look sensible but make other programs choke because they violate
> > format specs.
> > 
> 
> I'll do.
> 

I did a quick check on some programs

I used the following input file

------- empty.fasta --------------
>1 d1
GCGGCGGCG
>2 d2

>3 d3
GCCGCCGCC
----------------------------------

Seqlab (GCG) : fromfasta inserts an empty sequence (no error)
Seqweb (GCG) : filters out the empty sequence
Blastn (v 2.0.9) : gives '0 letters', complains about a 'no valid query
			sequence' and continues
Clustalx (v. 1.64b) : reads in the empty sequence and uses it
FastA (web interface v. 3.3t06): complains about sequence with length
		<=0



Kris,