[Biopython] SQL Alchemy based BioSQL

[Peter]

On Thu, Aug 20, 2009 at 10:59 PM, Kyle Ellrott<kellrott at gmail.com> wrote:
>
>> Ah. That ties in with the other changes on your github tree (to
>> work nicely with Jython) which had seemed unrelated to me.
>>
>> I guess MySQLdb etc uses C code which means it won't
>> work under Jython. I don't know enough about Jython to say
>> if there are any other alternatives to using SQLAlchemy.
>
> If you aren't using a layer of abstraction like SQLAlchemy, they you
> can use the standard Java SQL interfaces (JDBC).  But code written for
> that would only work within Jython and be useless for CPython.

Still, it might be worth while. Assuming it can be done as a
(Jython specific) modular backend to the existing BioSQL
framework, it should be a less invasive change.

>> This might be a silly question, but why not just use BioJava to
>> talk to BioSQL instead? Or use Biopython's pairwise alignment
>> code. Was the point just to demonstrate things working together?
>
> To prove that it could be done was part of the point.  But there is
> also a 'cross training' aditude about it.

Fair enough.

> BioPython seems more lightweight/easier to use, but has
> heavier requirements on installing external applications.

Biopython does often wrap external command line tools, yes.

> BioJava can be harder to use, but it has lots more embedded
> functionality ( built in dynamic programming and HMM code ).

Biopython does have its own HMM and pairwise alignment code
written in Python (and for Bio.pairwise2 we also have a faster C
code version, but you wouldn't get that under Jython). These
modules are not covered in the tutorial (if anyone wants to help).

> If I can get both working in the same environment, then I get
> the best of both worlds.

Absolutely.

Peter