[Bioperl-l] BioPerl newbie question about exporting sub_SeqFe atures

Ewan Birney birney@ebi.ac.uk
Tue, 17 Apr 2001 22:51:55 +0100 (BST)


On Tue, 17 Apr 2001, Jason Stajich wrote:

> Problem is if you want to have a Feature A, and attach Feature B and C
> onto Feature A.  If you want to maintain a description for B, C, how would
> you do this?  add_sub_SeqFeature might have to get smarter and translate
> the contained Location object into a SplitLocation.

So this all comes down to "what do you mean" aka "how do you want this
represented in EMBL/GenBank" and/or "how do you want other programmars
access to your object". 

I think we've got to hear what Mark's real use case is, but it wouldn't
surprise me if the best solution is to inheriet off SeqFeature::gneeric
and overload to_FtHelper. Trying to make a one-size-fits-all SeqFeature
class without programming language extension is probably not going to
work.


(NB - we have the same problem with attributes getting muddled up with
eac_tag_value semantics. This is just the inherent polymorphism of feature
objects and us trying to tame them with interfaces and - even more
restrictive - data formats)

 > 
> BTW: I think the 'EXPAND' way of adding features looks messy to me,
> shouldn't all of this be captured in an object?
> 

EXPAND is messy. My fault. I *think* that we should be pointing people to
the location object as the first place to solve this - in fact, I would
not be against removing add_sub_SeqFeatures from SeqFeature::Generic and
making sub_SeqFeatures return () but this might be a little too dranconian
here.


This is all interesting grist to chew on as we encourage people to build
ontop of bioperl. I favour the sub-classing method, but it does require
programming expertise.  Hmmmm. Worth arguing about for sure.


PS - about to watch the european footie now. 




> -jason
> 
> On Tue, 17 Apr 2001, Ewan Birney wrote:
> 
> >
> >
> > Guys -
> >
> >
> > the "real" bug here in some sense is that Mark is not using
> > SplitLocations, and, in addition, that what "sub_SeqFeatures" mean
> > relative to a Split Location is really ill defined, and/or that the whole
> > mapping of "complex object" to EMBL/GenBank is a ... semantic mire.
> >
> >
> > Mark - what do you actually want to do? It may wellbe that you want to be
> > introduced to the wonders of the to_FTHelper function which will give you
> > full control of the way your features is dumped in EMBL and GenBank
> > format.
> >
> >
> >
> > e.
> >
> >
> >
> > -----------------------------------------------------------------
> > Ewan Birney. Mobile: +44 (0)7970 151230, Work: +44 1223 494420
> > <birney@ebi.ac.uk>.
> > -----------------------------------------------------------------
> >
> > _______________________________________________
> > Bioperl-l mailing list
> > Bioperl-l@bioperl.org
> > http://bioperl.org/mailman/listinfo/bioperl-l
> >
> 
> Jason Stajich
> jason@chg.mc.duke.edu
> Center for Human Genetics
> Duke University Medical Center
> http://www.chg.duke.edu/
> 
> 
> 
> 

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