[Bioperl-l] GCG MSF format alignments

Ewan Birney birney@ebi.ac.uk
Mon, 12 Mar 2001 16:25:37 +0000 (GMT)


On Mon, 12 Mar 2001, Peter Schattner wrote:

> > > I actually like the exception being thrown there, even though I bumped
> > > into this several times, too. However, it always indicated a bug.
> > >
> > > I'd actually prefer this being solved by creating a specialized class
> > > that specifically allows for additional characters as they are used in
> > > e.g. alignments. This should go under Bio::Seq::<AlignedSeq>.pm
> > > (replace <...> with whatever you like better). Ideally Bio::PrimarySeq
> > > (or maybe better Bio::PrimarySeqI?) has a method boolean
> > > validate_seq(), and depending on its return value seq() throws an
> > > exception. So, in your deriving implementation you'd only have to
> > > override validate_seq(). (As a matter of fact, there is no
> > > validate_seq() right now :|
> > 
> > Yup. The refinement here is that Alignments use LocatableSeq so that's the
> > place to put it (perhaps LocatableSeq should migrate to AlignedSeq...)
> 
> Well, I'm clearly out-voted.  That's fine by me.  Do you guys envision
> implementing these enhancements (eg validate_seq() method) in 0.7.1 or
> only in 0.8?  In the latter case I'll submit a documentation-bug fix
> removing the statement that alignments can be read in with "~" or
> similar strange characters and fixed later.  

I think this is cool for 0.7.1 as well as on the main trunk -
Hilmar/Jason? Thoughts?

> 
> Cheers
> 
> Peter
> 

-----------------------------------------------------------------
Ewan Birney. Mobile: +44 (0)7970 151230, Work: +44 1223 494420
<birney@ebi.ac.uk>. 
-----------------------------------------------------------------