[Bioperl-l] Project: BioPerl 2.0 (and beyond)

[Chris Fields]

On Apr 8, 2010, at 11:24 AM, Robert Bradbury wrote:

> On 4/6/10, lin yu <joycelin12 at gmail.com> wrote:
>> Therefore, I would like to propose the use of BioPerl in access R
>> phlyogenetic tools or the combination of R and BioPerl in other similar
>> applications like microarray analysis etc. Would it be possible? I would
>> like to know the answer as soon as possible as I can edit my proposal in the
>> case.
> 
> Joyce, it isn't clear what you are really asking for here.  I did some
> digging and it appears as if "R" is a statistical language derived
> from S:
> http://www.r-project.org/
> It isn't clear whether there is currently a toolset of phylogenetics
> applications that use R (if so could you provide a pointer or some
> examples)?

Answering for her, but there are many.  Here's one just via a Google search using 'R phylogenetics':

http://www.r-phylo.org/wiki/Main_Page

> In theory one can interface perl programs to anything since it has
> complete system call capabilities.  Presumably the advantage from a
> BioPerl standpoint is to have BioPerl deal with the databases, genes,
> genomes, etc. (something R presumably does not do) and have R deal
> with the statistics side of any analysis.  But one needs a more
> complete examination of what statistical aspects of phylogenetics
> BioPerl does not currently handle (and could be better handled by
> taking advantage of features which R might have).
> 
> [In my opinion currently there is already way too much overlap in
> tools packages with BioPerl, BioPython and BioRuby -- and there is no
> matrix regarding the capabilities or lack thereof with each of these
> so one could easily tell which of the three could or should best be
> interfaced to R.]

Python has a very nice R interface I believe.

> It looks as if the phyloinformatics project is here:
> https://www.nescent.org/wg/phyloinformatics/index.php?title=Phyloinformatics_Summer_of_Code_2010
> but it isn't clear whether there is any coordination between the
> BioPerl side of things and the Phyloinformatics side of things.

There is overlap between the two organizations to a certain extent.  I could see having a direct set of bindings to R would be very nice from the BioPerl standpoint, but the project is really more a general Perl project in this case (R isn't just about microarrays or biostats).  The lone Perl-based interface, RSPerl, is no longer maintained that I know of (I never got it to work).

chris