Re: [Cdk-devel] Problems with the Structure Diagram Generator

[Thomas K. <tho...@we...>]

Chris,

thanks for your answer. You are right with your assumption that =
bioclipse is
using a different cdk version because I just tested the whole thing =
again
with a new compiled JCP version from the current trunk and there I got =
the
same odd structures as I have got directly with the SDG.=20

So it seems that something broke the current SDG implementation.
But anyway I will add a test case therefore.

Cheers,
Thomas

-----Urspr=FCngliche Nachricht-----
Von: er...@do... [mailto:er...@do...]=20
Gesendet: Montag, 9. Juli 2007 12:26
An: Thomas Kuhn
Cc: 'cdkdevel'
Betreff: Re: [Cdk-devel] Problems with the Structure Diagram Generator

Thomas,

it could well be that Bioclipse has a slightly different version and
that something broke in between you version and the one that bioclipse =
uses.

Ideally, you would write a test case for this problem for the current
version of CDK (trunk). I would suggest copying one of the existing
tests which uses a SMILES as a starting point and putting in the SMILES
of your notorious problem.
People can then jump right in and test things.

Cheers,

Chris

Thomas Kuhn wrote:
> Hi everybody,
>=20
> =20
>=20
> I try to generate new 2D coordinates for a bunch of molecules but I am
> not really successful. The molecule still looks very ugly after using
> the SDG.
>=20
> Normally I would say that this is a problem of the algorithm within =
the
> cdk and think that this one is not capable for such molecules. But If =
I
> load this molecule within bioclipse and press the JCP button =
=93Relayout
> the structure=94 the structure really looks good! (see the pictures
attached)
>=20
> =20
>=20
> So where could be my problem? Here are some code snips which I tested:
>=20
> =20
>=20
> StructureDiagramGenerator str =3D *new* StructureDiagramGenerator();
>=20
>                   str.setTemplateHandler(*new*
> TemplateHandler(DefaultChemObjectBuilder./getInstance/()));
>=20
>                   str.setMolecule(mol, *true*);
>=20
> =20
>=20
> //                str.generateExperimentalCoordinates(new =
Vector2d(0,1));
>=20
> //                str.setBondLength(1.5);
>=20
> //                str.generateCoordinates();
>=20
>                   IMolecule newmol =3D str.getMolecule();
>=20
> =20
>=20
> I tried it with using the TemplateHandler and without, I tried
> generateExperimentalCoorindates and generateCoordinates, I tried to =
set
> the BondLength but nothing helped so far.
>=20
> Has anyone an idea?
>=20
> I attached also the example molecule as mol file!
>=20
> =20
>=20
> Cheers,
>=20
> Thomas
>=20
> =20
>=20
> =20
>=20
> =20
>=20
> Thomas Kuhn
>=20
> =20
>=20
> Institut f=FCr biologische und chemische Informatik / Institute for
> Bioinformatics and Chemoinformatics
>=20
> Fachhochschule Gelsenkirchen / University of Applied Sciences of
> Gelsenkirchen
> Fachbereich Angewandte Naturwissenschaften / Department of Applied
> Natural Sciences
> August-Schmidt-Ring 10
> D-45665 Recklinghausen
>=20
> =20
>=20
> Fon +49-(0)2361-4970053
>=20
> Mobile +49-(0)177-8085663
>=20
> =20
>=20
> E-Mail tho...@fh...
> <mailto:tho...@fh...>
>=20
> =20
>=20
> =20
>=20
>=20
> =
------------------------------------------------------------------------
>=20
>=20
> =
------------------------------------------------------------------------
>=20
>=20
> =
------------------------------------------------------------------------
>=20
>=20
> =
------------------------------------------------------------------------
>=20
>=20
> =
------------------------------------------------------------------------
>=20
> =
-------------------------------------------------------------------------=

> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
>=20
>=20
> =
------------------------------------------------------------------------
>=20
> _______________________________________________
> Cdk-devel mailing list
> Cdk...@li...
> https://lists.sourceforge.net/lists/listinfo/cdk-devel


--=20
PD Dr. Christoph Steinbeck
Lecturer in Chemoinformatics
Univ. Tuebingen, WSI-RA, Sand 1, D-72076 Tuebingen, Germany
Phone: (+49/0) 7071-29-78978   Fax: (+49/0) 7071-29-5091

What is man but that lofty spirit - that sense of enterprise.
... Kirk, "I, Mudd," stardate 4513.3..