[Bioperl-l] Questions about BioPerl

Nathan Torkington gnat@oreilly.com
Tue, 28 Aug 2001 11:00:19 -0600


Jason Stajich writes:
> This is right venue for querying implementors, providing suggestions and
> improvements and getting roped into helping with the projects... ;) 

Excellent!  I just had a long IRC conversation with Ewan, who was kind
enough to answer my questions about BioPerl.

I suggested some simple wrappers to hide the Scary Object Stuff from
the minds of the weak.  The easiest part would be exported functions
to wrap the object constructors:

  use Bio::Perl;
  $seq = newseq( -seq              => "gattaca",
                 -desc		   => "sample sequence",
		 -display_id	   => "something",
		 -accession_number => "12",
		 -moltype          => "dna",
		);

  $seq = genbank_seq( -id  => "MUSIGHBA1");
  $seq = genbank_seq( -acc => "AF303112");

  $in  = openseq(Fasta, "< inputfilename");
  $out = openseq(EMBL, "> outputfilename");
  $in  = openalign(fasta , "< inputfilename");
  $out = openalign(pfam, "> outputfilename");

I wouldn't be trying to make the whole of BioPerl available through
simple functions, but merely to provide some common basic
functionality.  This would be the sugar pill that gets new users
hooked.  Before you know it they'll be subclassing and multiply
inheriting with the best.

That's as far as I've thought through the idea.  Any obvious flaws?
Any other suggestions for easy ways to wrap the functionality?

If nobody points out that I've obviously been hitting the crack pipe
and has me committed to an asylum for the mentally cuckoo, I'll start
coding.

Nat