[BioPython] Me first!

Andrew Dalke dalke@bioreason.com
Thu, 02 Sep 1999 17:57:47 -0600


Bradley Marshall <bradbioperl@yahoo.com> said:
> I'm the first one on the list with an interesting post (although
> I suppose that depends on whether the archives are correct and
> what you consider interesting.)

Then does that make me the first one with an interesting response?

> I have the following suggestions for the up and coming sequence object:
> 1)  compatability with the LSR's BSA proposal to the OMG.

I agree.  In fact, I think that's even part of the FAQ.  Nope,
don't see it.

Anyway, I sat in on the OMG LSR DTF BSA WG RFP meetings last week and
think they are rather nicely designed, and a reasonable idea of
where to start.

OMG = Object Management Group
LSR = Life Sciences Research
DTF = Domain Task Force
BSA = Biomolecular Sequence Analysis
WG = Working Group
RFP = Request For Participation

For those interested, copies are at

though people are still going over the proposal with a fine-toothed
bug brush to flush out a few, final inconsistancies and needed functions.

> 2)  Pure python implementation (no c modules) so it retains the
> possibility of being compiled into java with jpython (So I can crash
> my browser more.)

Both Jeff and I agree that pure Python code, runnable by JPython and
compilable into .jar files for use by other Java code, would be the
best solution.  There are likely to be times where some algorithms,
like alignment code, may be implemented in C for performance.

I hope that in those cases either someone will also write the code for
Java or, better (IMHO) write a slower Python implementation with the
ability to use the C one, if available.  Eg, how pickle/cPickle works.

For my code, I expect to write everything in Python first.

> 3)  Annotated sequences.  Although this is in the BSA proposal.

Yep, though there is still discussion on how to add annotations and
how the iterator semantics are affected.  I believe the BSA will
make adding annotations be an optional compliance point.

> 4)  Seperate amino acid and nucleic acid classes.  This is also
> in the proposal.

Yes.  I do have a question though of, should DNA and RNA be in the
same class or different ones?  My feeling is that different is the
way to go, but I can't think of a good case scenario off-hand.

> 5)  Make the links on the web site work.  They all sound so
> promising!  -To that end I'm gonna go dig up some link to python
> software.

Great!  And do you want to work on content for the web site?  :)

I do plan to spend time this weekend helping finish up the pages.

						Andrew Dalke