symmetry-smart chem-draw/coordinate-transformation program
ASG
drawing/teaching/excercise program, i.e., a program that could
automatically identify symmetry elements of a molecule (after geometry
optimization or specification of equivalent atoms) and its symmetry
group? I find it a pain in the neck to have to transform Cartesian
coordinates manually from a MM program into the optimum coordinate frame
for QM (GAMESS) calculations, especially if the system is more
complex...
Thanks for any suggestions (Babel doesn't do what I want).
Tony
Michel PETITJEAN
Ref: <3820E8DB...@monmouth.com>
I have recently worked on the elaboration of such a programme. A way is
to compute a symmetry index, and deduce the symmetry element(s) from the
optimal superposition(s) of the set onto itself or onto its inverted image
(see my papers in: J.Math.Phys,1999,40[9],4587-4595, Compt.Rend.Acad.Sci.Paris,
serie IIc,1999,2,25-28, and J.Math.Chem.1997,22,185-201). At this time, I have
a programme computing the indices, but it is not yet a clean documented one;
the input molecule, i.e. structural formula plus cartesian coord., is entered
in a usual format, as hyperchem, sybyl mol2, or cambridge model; the output
is a list of pairwise atomic correspondences together with their associate
optimal affine transformation and the two indices (direct and indirect).
Until today, I thought that my programme has very few potential users.
Nevertheless, if you or any else has indeed some need of that, please
let me know, and I shall build a proper version to be exported. Also note
that I have worked on an heuristic algorithm working on cartesian coord only,
and able to recognize symmetry elements and pairwise correspondence without
knowing structural formula: for large molecules the heuristic version runs
faster than the latter programme, but it is less robust (the latter works even
when there is a large deviation from perfect symmetry).
Michel Petitjean, Email: peti...@itodys.jussieu.fr
ITODYS (CNRS, ESA 7086) ptit...@ccr.jussieu.fr
John Osborn
>
> I wonder if there's a symmetry-smart structure
> drawing/teaching/excercise program, i.e., a program that could
> automatically identify symmetry elements of a molecule (after geometry
> optimization or specification of equivalent atoms) and its symmetry
> group? I find it a pain in the neck to have to transform Cartesian
> coordinates manually from a MM program into the optimum coordinate frame
> for QM (GAMESS) calculations, especially if the system is more
> complex...
The Cerius2 molecular modelling package has a program which will
identify the point group of a molecule, and if you wish will transform
its coordinates into a standard reference frame (e.g. with the
principal axis, if there is one, aligned along z). The program is not
perfect and sometimes cannot identify the point group, but it usually
works. You can specify the tolerance you require to allow for slight
deviations from symmetry. This program comes with the quantum modules
available with Cerius2. Presumably other molecular modelling packages
have similar programs. Of course they are not free, and it would not
be worthwhile purchasing them simply to identify symmetry elements.
--
John Osborn
University of Bradford, UK.
To reply by email, replace "br" by "bradford" in my email address.
ASG
> The Cerius2 molecular modelling package has a program which will
> identify the point group of a molecule, and if you wish will transform
> its coordinates into a standard reference frame (e.g. with the
> principal axis, if there is one, aligned along z).
Thanks both to you and Michel for your suggestions. Since I'm toying around
with QM calculations in my "free" time, rather than do it for a living, I'd
rather get it for free, nice and easy. ;-) I'm actually looking for a
companion to Cotton's "Group Theory, Chp. 3", monograph, to exercise and check
on myself. "The Cerius2 component might be it, since the most onerous task
for me is to transform Cartesian coordinates obtained using C1 symmetry, i.e.,
no symmetry at all, into a standard reference frame, exactly as you
mentioned. Essentially, this is a simple geometrical translation/rotation
task which could easily be handled by a small QBasic utility program, but I'll
leave this for the winter evenings...
Regards, and thanks for your help.
Tony
John Osborn
>
> The Cerius2 molecular modelling package has a program which will
> identify the point group of a molecule, and if you wish will transform
> its coordinates into a standard reference frame (e.g. with the
> perfect and sometimes cannot identify the point group, but it usually
> works. You can specify the tolerance you require to allow for slight
> deviations from symmetry. This program comes with the quantum modules
> available with Cerius2. Presumably other molecular modelling packages
> have similar programs. Of course they are not free, and it would not
> be worthwhile purchasing them simply to identify symmetry elements.
I should have added that currently Cerius2 requires a unix workstation
(e.g. Silicon Graphics) to run. It does not run on a PC. At some time
in the future it will move to a client-server arrangement where the
client will be a PC. However, at the moment it is probably not a
viable option for someone who does not have access to a workstation.
Christopher B Specker
: I should have added that currently Cerius2 requires a unix workstation
: (e.g. Silicon Graphics) to run. It does not run on a PC. At some time
--
-----------------
Christopher Specker
John Osborn
>
> John Osborn (j.c.o...@br.ac.uk) wrote:
> : I should have added that currently Cerius2 requires a unix workstation
> : (e.g. Silicon Graphics) to run. It does not run on a PC. At some time
> So which category does a PC running Linux or *BSD fall into?
Cerius2 cannot run on such a system at least at the moment.
I don't know whether it will run on Linux or *BSD in the future when
it moves to a client-server arrangement.
ASG
> I should have added that currently Cerius2 requires a unix workstation
> (e.g. Silicon Graphics) to run. It does not run on a PC.
I use WinNT boxes both at work and at home and have acces to a Sun network,
but for this project being a self-improvement hobby ;-), my 32-bit versions
of Mopac 6.0, PC GAMESS-US and Molden are good enough. It's the symmetry
part I'd like to exercise... It galls me that Cotton's book has no answers
to its exercises (at least, I couldn't find them).
In place of such a program, where could I find a relatively large set of
molecules/complexes with assigned symmetry groups? Working with such
examples would be helpful, too--more so to a non-professional QM researcher
or crystallographer than relying on dry procedures only. One picture is
worth a thousand words, and Cotton's Chap. 3 doesn't cut it for me!
Tony
John Osborn
>
> In place of such a program, where could I find a relatively large set of
> molecules/complexes with assigned symmetry groups? Working with such
> examples would be helpful, too--more so to a non-professional QM researcher
> or crystallographer than relying on dry procedures only. One picture is
> worth a thousand words, and Cotton's Chap. 3 doesn't cut it for me!
A database of molecular structures determined by gas phase electron
diffraction and/or microwave spectroscopy would be likely to have
this information. I do not know of a recent compilation of such
structures (either published or online), but the following relatively
old compilation specifies the point groups for many of the entries:
Landolt-Bornstein Numerical Data and Functional Relationships in
Science and Technology, New Series (ed. A Encken): Group II, vol. 7 -
Structure Data of Free Polyatomic Molecules.
ASG
John Osborn <j.c.o...@br.ac.uk> wrote:
> the following relatively
> old compilation specifies the point groups for many of the entries:
> Landolt-Bornstein Numerical Data and Functional Relationships in
> Science and Technology, New Series (ed. A Encken): Group II, vol. 7 -
> Structure Data of Free Polyatomic Molecules.
Thanks; I'll look it up. BTW, are these standard QM or ctystallographic
point groups? I'm looking for the former.
Tony
Sent via Deja.com http://www.deja.com/
Before you buy.
John Osborn
>
> In article <3826F3...@br.ac.uk>,
> John Osborn <j.c.o...@br.ac.uk> wrote:
>
> > the following relatively
> > old compilation specifies the point groups for many of the entries:
> > Landolt-Bornstein Numerical Data and Functional Relationships in
> > Science and Technology, New Series (ed. A Encken): Group II, vol. 7 -
> > Structure Data of Free Polyatomic Molecules.
NB The editor given above is the editor of the entire series, not of
this particular volume: it may not be catalogued in a library under
the name of Encken.
> Thanks; I'll look it up. BTW, are these standard QM or ctystallographic
> point groups? I'm looking for the former.
I don't know what you mean by a standard QM point group. The point
groups in the above data set are simply the conventional point groups
of the molecules in question. They may or may not happen to be
crystallographic point groups, depending on whether or not
all symmetry elements in a given molecule are capable of existing in
a crystal (i.e. the point group will be a crystallographic point
group if the only proper or improper rotation axes in the molecule
are 1-, 2-, 3-, 4- or 6-fold axes; there may or may not be an
inversion centre and/or mirror plane(s) as well). If a particular
gas phase molecule does happen to have a crystallographic point
group (which many do), this does not mean that the compound will
crystallise with a crystal structure having that particular group:
some or all of the symmetry elements may be lost on crystallisation.
Luis Torres
dlzc
...
> How did you guys do since then?
See if this addresses your question:
https://math.nist.gov/mcsd/savg/tutorial/cqc.html
David A. Smith