[Biophp-dev] Current Events - State of the Project, plans, etc.

S Clark biophp-dev@bioinformatics.org
Tue, 15 Jul 2003 16:46:06 -0600


CVS:

Sourceforge STILL appears to be mostly dead.  Only developers appear to be
able to get to it at all.  WebCVS updates are still dead, and
anonymous CVS access appears to be down completely.  This appears to be
sourceforge's whole system, not just GenePHP.

CVS at Bioinformatics.org should mirror what's in the sourceforge CVS, at 
least as of last time I updated, so current code can be viewed there on
webCVS or downloaded via anonymous CVS.

http://bioinformatics.org/docs/cvs/#anonymous

http://bioinformatics.org/cgi-bin/cvsweb.cgi/biophp/

Development:

Hello to Euan Adie, Gene Ioffe, and Dan Bolser who have recently
signed onto the list!

Nico (also of PHPLabware fame - http://phplabware.sourceforge.net ) has been
on vacation, but had been working on some BioSQL code.  I believe he gets back
this week, so we may hear more from him soon.

Serge is attending classes through October and said he won't be getting
much coding done meanwhile.  He mentioned that he did have
some updates to post, but only has access to his main computer
every two weeks or so.  Serge mentioned that he'd be focussing more
on "quick and dirty" (mainly standalone?) working scripts, to be
more 'formally' re-worked at a later date for incorporation as
BioPHP modules.
( http://bioinformatics.org/pipermail/biophp-dev/2003-June/000212.html )
He's gotten a lot of working scripts up on his GenePHP site that aren't
in CVS, so there are some good resources there:

http://genephp.sourceforge.net/demo_set.html
http://genephp.sourceforge.net/io_set.html


I (Sean) have checked in an NCBI BLAST interface, some minor
updates to the 'core' XML parser (and, incidentally, thanks
again to Greg Tyrelle for supplying a written parser for
me to look at earlier!) and ESearch and EUtils modules,
the first draft of the seqIOExport module, and the beginnings of
interface methods in seq.inc.php, and minor update to seq_factory
to make use of them.

My current plans (in an order roughly resembling the order I intend
to tackle them):

EFetch for sequences - necessary for some other things.  This won't be
"difficult" as such, but will be a bit tedious (the EFetch sequence
record is reasonably large).  As part of this, I'll probably add
one or two additional "parent tag searching" methods to the 'core' xml parser.

An "alternate" implementation of seq.inc.php - I want to experiment with
some different internal structures for data storage, handling of sequence
type, and so on, and I don't know yet how much of it will be feasible, so 
an "interface-method-compatible" object with the same class name but
different filename seems like the best way to avoid clobbering the existing
seq object and code that uses it with my potentially-non-functioning
incompatible experiments...

Import "wrappers" for EFetch, NCBI BLAST, and ESearch - provided there are
no objections to my "BioPHP URLs" idea, I can implement them as legal import
"targets", so one could, e.g., import the top 10 BLAST matches from NCBI
for a particular sequence as a set of seq objects for further analysis.

Export updates - my first draft of the seqIOexport module will probably need
a bit of tweaking, but after that, ability to accept seq_align and related
objects as input directly, and exports for clustal, phylip, and genbank seem
like important formats to add soon.

Frontends - for clustalw (Serge actually already has code for this I believe), 
fastDNAml, and a few other "compiled" analysis tools (particularly those
which wouldn't be feasible implemented in "native PHP").

Other Sites/Projects:

Greg Tyrelle (of nodalpoint.org fame) mentioned BioPHP at at 5-minute
talk at BOSC.  He mentioned that there was a fair amount of interest
among people there, particularly for building web-interfaces for
bioinformatics tasks (unsurprisingly).

I've been contacted by Hallgier Bergum, who has the biophp.org domain.  His
project has been a sequence annotation database system, but he mentions
that he's not had time to work on it, and suggested we pick up the domain
name from him "for a reasonable price" since he's not going to be able to
make use of it any time soon.  I sent back a reply expressing interest and
asking how much the transfer would cost - with my employer going out of
business from under me (and me stuck in southeastern Idaho trying, without a 
completed college degree, to find 'high-tech' work so I can afford to complete
my college degree...being a "pay as you go" older college student can
be a REAL pain sometimes...) I can't afford much.  No reply yet, but I'll keep
everyone posted.

As mentioned in my last post to the list, Zend Weekly News mentioned other
bioinformatics developers on PHP - I sent an email off asking Andrei Zmievski
if he could tell me who some of these developers are, so we might have more
information (and perhaps interested developers?) coming in at some point
about Bioinformatics on PHP...

As always, comments, suggestions, and corrections are welcome...

Sean