[ghemical-devel] Exporting from GChemPaint to Ghemical

[Jean Bréfort]

--=-JNVAl1uupwYbotPGobUb
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Le mer 30/06/2004 =C3=A0 12:54, Tommi Hassinen a =C3=A9crit :
> On Tue, 29 Jun 2004, Jean Br=C3=A9fort wrote:
>=20
> > > But averaging the coordinates should do the centering trick.
> > >
> >
> > Thanks, it works. I must now try to transfer as far as possible the
> > stereochemistry (not so simple). At the moment, when I transfer a
> > molecule such as cholesterol with all z coordinates set at 0., I cannot
> > optimize the geometry (nan values everywhere). More simple molecules
> > work. Have a nice summer.
>=20
> Hummm... I guess the calculations fail because there is something bad in
> the initial geometry. It might be some of the following:
>=20
> 1) some atoms (most likely H and non-H atoms?) have identical coordinates
> (or otherwise overlap badly), creating bonds with zero length.
>=20
> 2) some bond angles have extreme values of 0.0 deg or 180.0 deg.
>=20
> The problems above cause computations to fail because there are some
> infinite or undefined intermediate values (that is, for example a divisio=
n
> by zero can happen in the calculations).
>=20
> How you could avoid this, is to modify the structure (even randomly!) a
> bit if the structure contains some errors mentioned above. Even a tiny
> modification is often enough to make the calculations work again.

There are some 180=C2=B0 angles (on quaternary carbons). This might disappe=
ar
when I'll change the Z coordinate.


--=-JNVAl1uupwYbotPGobUb
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: Ceci est une partie de message
	=?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQBA4vyiUkVLHK5zOkcRAnVzAJ9IU3xVtL1L629Hv4rRNsNWreMSigCgmZ9J
rquRGVQbyraJRx1f9Y2ISEI=
=w08p
-----END PGP SIGNATURE-----

--=-JNVAl1uupwYbotPGobUb--