[Biophp-dev] New question from the Newbie

[S Clark]

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello again!  Nice to see activity starting up again in here...
(Not that I can complain, having been pretty quiet myself)

On Friday 19 March 2004 06:25 pm, Nico Stuurman wrote:
[...]
> What is the local BLAST module?  I wrote an interface to primer3 which
> I now put into biophp/genephp/interfaces.  I guess that this is
> conceptually similar to a local BLAST interface (and for that matter,
> an interface to clustalw, and hmm, there must be more out there that we
> need to interface to).

fastDNAml...(I've really got to get my head wrapped around what the
algorithm should be for parsing and modelling newick trees...)

There are probably a variety of Java modules we could optionally link
directly to via the PHP-Java integration (in addition to just running
the 'whole programs' like normal).

The local BLAST module is a frontend for running a local copy of the BLAST
database search.  It works for me, and SHOULD work for Windows as well...
but the only person who's REPORTED using it on Windows has been having
odd problems getting it to work that I've not been able to track down.
(If anyone else here runs Windows and has an interest in playing with
the local BLAST frontend to help figure out what's going on, please let
me know...) (It LOOKS like the frontend is doing what it should be,=20
but the BLASTALL.EXE windows program isn't returning any data in the=20
output file...)

By default, both it and the NCBI BLAST module ask for XML output, so both
frontends can use the same BLAST results parser.
[...]
> The way it was last summer was that the bioinformatics cvs could not be
> accessed from Serge's windows box.  I will fire up the old beast in the
> basement this weekend and figure out if that is still the case.  If so,
> I propose to move the cvs repository to sourceforge so that everyone
> can contribute.  If not (that is bioinformatics cvs now works with
> windows), I invite everyone interested in biophp to contribute their
> code to that cvs.  In any case, Sean (and I, I guess) will need to do
> some re-organization of the code in cvs such that there is a logical,
> extendable structure in place.

I got the impression the problem there was more figuring out how to get
the client program configured (rather than a problem server-side on
either bioinformatics.org or sourceforge - though I know sourceforge
has had a lot of problems in the recent past).

Of course, it's all moot if we 'upgrade' to SubVersion instead :-)

> > Sometime later today, I'll try to get an email out to the list with an
> > overview of the importing modules (interfaces implemented by the
> > individual
> > 'file-type import' modules, and what needs to be added to the central
> > import class for e.g. auto-detection).

At least a rough overview as promised - if a day or two late - will follow
momentarily...I finally had a chance to pull up the files and look at
them again (I'd forgotten that I'd gotten a start on a seqIOExport module..=
=2E)


Sean
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.9.5 (GNU/Linux)

iD8DBQFAXTBBJ6yQLhNTzSkRAuXLAJ99EhYzbdLFjQXOsrq9B/Tg2KsxAACgrXmF
u8Bwrq3rjvZVzz7Sp90IlBk=3D
=3DiUEb
=2D----END PGP SIGNATURE-----