[Biojava-dev] Biojava Structure and its derivatives are PDB centered

[Andreas Prlic]

Seriously, it should not be that hard. Let's discuss your strategy first
before you start coding?

A


On Mon, Dec 9, 2013 at 8:45 PM, Amr AL-HOSSARY <amr_alhossary at hotmail.com>wrote:

> OK. I'll try to comply to it..
> However, I expect a (nice) amount of difficulties. :)
>
> Amr
>
> BTW
> Is anybody willing to aid?
>
> -----Original Message-----
> From: biojava-dev-bounces at lists.open-bio.org
> [mailto:biojava-dev-bounces at lists.open-bio.org] On Behalf Of Spencer
> Bliven
> Sent: Tuesday, December 10, 2013 7:05 AM
> To: Amr AL-HOSSARY
> Cc: biojava-dev
> Subject: Re: [Biojava-dev] Biojava Structure and its derivatives are PDB
> centered
>
> Amr-
>
> Yes, the Structure>Chain>Group>Atom data structure defined by BioJava
> contains concepts used in the PDB format, but that doesn't mean that a new
> implementation is needed to write a MOL2 parser. For instance, each
> MOLECULE
> could be used to create a model in the biojava Structure containing a
> single
> chain and group for the ligand. (I'm not sure whether there are conventions
> for protein-sized structures in MOL2, all the examples I've seen are for
> small ligands.) The intent of separating Atom from AtomImpl was primarily
> to
> provide a pure interface for the API, not necessarily to allow different
> concepts of how to organize 3D points into a meaningful system. Any new
> file
> parsers should use the existing data structures.
>
> -Spencer
>
>
> On Mon, Dec 9, 2013 at 2:33 AM, Amr AL-HOSSARY
> <amr_alhossary at hotmail.com>wrote:
>
> > Dear Biojava Development team,
> >
> >
> >
> > I noticed that we don't implement a reader/writer for MOL2 file format
> > and I thought of implementing one.
> >
> > But I noticed that the Structure interface (and almost all
> > structure-related classes are PDB centered).
> >
> >
> >
> > For example, AtomImpl is an implementation of the interface Atom..
> > sounds nice.
> >
> > AtomImpl is mentioned to be an implementation of an Atom of A PDB
> > file, and so it implements Atom, Serializable, and PdbRecord .. hum
> > that sounds logic, although it should have been called AtomPdBImpl,
> > but no problem.
> >
> > Now Atom itself extends Serializable and PDBRecord !!!!!!!
> >
> > In PDBRecord there are two overloaded forms of toPDB(). That is logic,
> > but what is not logic is letting Atom (as an abstract definition of
> > Atom) implements PDBRecord and passing toPDB() to all implementing
> classes.
> >
> >
> >
> > IMHO, the solution to this mess is to introduce an interface called
> > Parsable or Persistable, including methods read() and write(), and
> > implementing objects would implement it in different ways.
> >
> > Or simply to create a helper class that takes the Atom object and
> > write the appropriate format.
> >
> >
> >
> > A similar problem is present for Structure and StructureImpl in a less
> > severity form.
> >
> >
> >
> >
> >
> > Another solution is to rewrite the structure package in the
> > org.biojava3.structure with the new PDB-free mind orientation.
> >
> >
> >
> >
> >
> > Regards
> >
> >
> >
> > Amr
> >
> >
> >
> >
> >
> > Best Regards,
> > Amr AL-HOSSARY
> >
> >
> >
> >
> >
> >
> > Amr Ali AL-HOSSARY |   Ph.D. Student | International PhD program in
> > Computational Biology and Bioinformatics | School of Computer
> > Engineering | Nanyang Technological University | Singapore (GMT+8) |
> Email:
> > <mailto:aalhossary at pmail.ntu.edu.sg> aalhossary at pmail.ntu.edu.sg |
> >  Mobile:
> > +65-94572816
> >
> >
> >
> >
> >
> >
> >
> >
> > _______________________________________________
> > biojava-dev mailing list
> > biojava-dev at lists.open-bio.org
> > http://lists.open-bio.org/mailman/listinfo/biojava-dev
> >
> >
> _______________________________________________
> biojava-dev mailing list
> biojava-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biojava-dev
> _______________________________________________
> biojava-dev mailing list
> biojava-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biojava-dev
>