Dear all,
I have written a small set of modules that have in common certain goals
with the bioperl modules Bio::DB::Taxonomy(::flatfile) and Bio::Taxon.
The main differences are:
+ No dependencies of non-standard Perl modules (VS tons of dependencies
of bioperl modules) -- This is important in certain (and convenient)
environments (like GRID or cloud systems) where you can't rely on
automatic installation of big bundles.
+ NCBI and RDP taxonomies support (VS only NCBI support of bioperl
modules) -- I plan to add support for additional taxonomies.
+ Very fast and low memory footprint (VS general poor performance of the
bioperl bundle). Orders of magnitude even for the simplest lookups.
+ Fast mapping of different identifiers (VS lack of this feature in bioperl)
Of course, these modules doesn't compete with Bio::DB::Taxonomy and
Bio::Taxon in completeness of methods or integration with other tools
(e.g. the rest of the bioperl bundle) but they are very handy (and
actually being used by several bioinformatics groups) for fast mapping
and large datasets analysis (frequent in bioinformatic analysis).
The current (local) name of these modules are "Taxonomy",
"Taxonomy::RDP", "Taxonomy::NCBI", "Taxonomy::NCBI::Gi2taxid"
I have been told to put them in CPAN. Their natural location would be
under the "Bio::" namespace. The bioperl developers don't have any
problem with this as long as i) the documentation clearly states that
these modules are not bioperl related and ii) the naming is sufficiently
distinguishable from existing (bioperl) Taxonomy modules. Bioperl
already have a Bio::Taxonomy::* (obsolete) and Bio::DB::Taxonomy::*
which makes difficult the naming decision.
I have in mind some alternatives like:
Bio::Taxonomy::Lite::*
Bio::DB::Taxonomy::Lite::*
Bio::TaxDB::*
Bio::TaxonomyDB::*
Bio::TaxLite::*
Bio::FastTaxonomy::*
I don't know what is your feel about this. Any suggestion would be welcome.
Best regards,
M;
