[Bioperl-l] Alignment::Clustalw and Bio::RootI

Murad Nayal murad@godel.bioc.columbia.edu
Tue, 12 Dec 2000 15:18:29 +0100


> 
> > 2- how foolish is it to start mingling 0.62 and 0.7 objects. I really
> > need a clustalw wrapper and don't want to write one since it is already
> > available in 0.7. I suppose if I replace RootI 0.62 with RootI 0.7, then
> > 0.62 objects (which inherit from Object) will hook into Object methods
> > while 0.7 objects (really just Clustalw and co) will access the RootI
> > methods!!.
> 
> Going to be very difficult if not impossible to mix versions right now.
> Once we started down this path, we acknowledged there would be a lot of
> incompatibilities between versions because of Root::Object dependence. You
> will have an easier time checking out the main branch and learning with
> that then struggling with compatibility issues.  The Clustal module is
> just a wrapper around the clustalw program, are you trying to do something
> so tricky that you can't just run clustalw separately and then feed it
> into AlignIO (oh shoot, which is only on main-trunk too, well....).
> 
> Let me know if I can help steer you in the right direction.  Peter S (the
> author) is away on vacation for a few more weeks so I agreed to stand
> watch on the AlignIO and Bio::Tools::Alignment in the meantime.

Thanks Jason, I wasn't sure of the maturity of the 0.7 code so the
question was which option will give me the least amount of trouble. I
actually tried briefly to use RootI 0.7 in place of RootI 0.62 and it
seems to work for the examples I tried. the reason I think this is so is
that the 0.62 objects simply don't see the new RootI as Root::Object
stands in the way. in any event. I will retreat to 0.62 or some stable
version of the main truck the minute problems start arising.

needless to say, waiting anxiously for the release of 0.7 :-)

Murad