[Bioperl-l] Trace files and trace points

Aaron J Mackey Aaron J. Mackey" <amackey@virginia.edu
Wed, 10 Jul 2002 20:00:46 -0400 (EDT)


I have absolutely no use for trace data (peak height and peak indices and
all that jazz), but I'm sure others who do have a list of "usual
operations" they expect to be able to do (Chad's scf module already
provides a useful baseline to start out with).  Like the PrimaryQual
object, these would need to be tightly bound to their underlying sequence
(and altering the sequence itself should be a runtime error, IMHO).

I'm going to punt on this until someone begs for it at BOSC; it'll be
trivial to add to the read/write routines (since the io_lib is doing all
the grunt work) when we have a place to put the data.

-Aaron

P.S. Thanks to Robson de Souza, I now have some abi files, and SeqIO::abi
works (again, for input only - no output).  Still looking for these
mysterious 'ALF' format files ...

On Wed, 10 Jul 2002, Jason Stajich wrote:

> Way cool stuff Aaron.  I think this will be very very useful.  I like the
> idea of Seq::Trace object.  When you get to thinking about building one,
> can you send out a list of methods you think it would need - this is
> mostly for my edification about how one might want to design a trace
> object.
>
> -jason
>
> On Wed, 10 Jul 2002, Aaron J Mackey wrote:
>
> >
> > So the other large chunk of data in a trace file besides the quality
> > values are the trace point data.  Chad's scf.pm module is storing them in
> > the io object (not in the Seq object itself) (which of course makes
> > writing an SCF file with peak info a bit difficult - maybe I've
> > misunderstood something there).  Regardless, do we want to just trash this
> > info for now, put it in a data field of the seq object, or make a
> > Seq::Trace object to hold them (and potentially do interesting stuff with
> > it?) ... we don't have to answer this immediately, but until we do, 'ctf'
> > format output won't work (as it requires trace info).
> >
> > And, I've found that the io_lib routines don't write 'abi' or 'alf'
> > (multi-seq abi format) formats (and also no 'ctf' unless trace info is
> > available, as I said above).  Everything else is ready to go, just putting
> > finishing touches on make test (100 tests!) and whatnot.
> >
> > -Aaron
> >
> >
>
>

-- 
 Aaron J Mackey
 Pearson Laboratory
 University of Virginia
 (434) 924-2821
 amackey@virginia.edu