[Biojava-l] Biomolecular structure objects

[Michele Clamp]

On Fri, 7 Apr 2000, David Martin wrote:

> > 
> > Do you know of any Java parsers for the protein structure databases -
> > pdb, dssp and the like? I have just put together a rudimentary DSSP
> > parser, but it is very nasty. It would be nice to have a basic toolkit
> > of protein database-related stuff in BioJava but I don't fancy writing
> > it all from scratch.
> 
> I don't know if this would help but there is a set of C++ objects
> available from Birkbeck. These may be of some use.
> 
> I have tried to get hold of them before but the server is always down (or
> it seems like it). BCM library or some such.
> 
> There are some Java based PDB viewers. Any joy there?
> 
> ..d

I don't know whether you can get the code but the most polished java PDB
viewer I've seen is Dirk Walther's  - see somewhere around

http://www.cmpharm.ucsf.edu/~walther/

I don't know whether source is available.  There's always the rasmol code
of course :-)  Also the MSD group at the EBI having been doing heavy duty
PDB work and can probably do this in their sleep.  

I do know that PDB file parsing is an icky business :-)  I have a basic
one but that is based around my very first java program and also my first
c program that used linked lists.  Not to be recommended.

For what it's worth here's how it went.

All atoms represented as Atom objects.
All residues represented as Residue objects.

The structure was then defined by a set of Bond objects that
linked atoms.  These bond objects were the things drawn to the screen.

Different views (full atom, Ca only) were achieved using different sets of
bonds.

I've got off the point here.

M.

-- 
As the absent-minded zookeeper of time scrubs his loo with the 
startled bush-baby of hope, and the frisky king penguin of fate 
approaches the small nun of destiny, we come to the end of the mail. 

(Apologies to  Humphrey Lyttleton)