[Bioperl-l] Bio::Species, Bio::Taxonomy::Node overhaul
cjfields at uiuc.edu
Mon Aug 7 09:22:47 EDT 2006
That was about $seq->species->species() (i.e. Bio::Species). The old
method returns the species name (i.e. 'subtilis'). The genus()
method returns 'Bacillus', and binomial() returns 'Bacillus subtilis.'
Even if we ended up removing genus() and binomial(), the use of
species() here will be tainted by it's past use for returning the
species descriptor, so changing the way it behaves (returning the
binomial) is really an API change.
My thought was to tie genus(), species(), and binomial() to DB
lookups and have them return the previous information. But that's
also really an API change, and is it really necessary? I thought the
point of all this was to get rid of Bio::Species and have a much
cleaner container object take it's place (Bio::Taxon).
The lineage would be stored in Sendu's Bio::DB::Taxonomy::list
object, the organelle() moved to RichSeq (it's not involved in
taxonomic data), and the common names and scientific name added to
Bio::Taxon. So, Bio::Taxon could be used for writing output.
Bio::Species would be retained for the API and gradually phased out.
So, using your suggestion, a Bio::Taxon object would be returned here:
and a (soon to be deprecated) Bio::Species object would be returned
On Aug 7, 2006, at 7:54 AM, Hilmar Lapp wrote:
> On Aug 7, 2006, at 8:41 AM, Chris Fields wrote:
>> The only part I don't agree with is the use of species() to return a
>> binomial name, which I already responded to (and don't plan dragging
> Maybe I missed that? This is not about $seq->species() but a species
> () method on another object? If it's on a sequence object it's
> clearly a bad idea due to the API change, but I thought that's not
> the plan.
> : Hilmar Lapp -:- Durham, NC -:- hlapp at gmx dot net :
Lab of Dr. Robert Switzer
Dept of Biochemistry
University of Illinois Urbana-Champaign
More information about the Bioperl-l