Hi Michael,
I think we face confusion between different types of radii... However I'm
not a chemist.
In PyMOL, what the "sphere" representation shows is the van der Waals
radius, which by definition is determined from measurements of atomic
spacing between pairs of unbonded atoms in crystals (wiki). Hence in PyMOL
the van der Waals radius of Mg is 1.73 A (by definition not bound to any
element). However it seems that the charge is not taken into account by
PyMOL: I tried to include two positive charges to Mg and no change could be
seen (the H atom is still at the very border of the sphere of Mg).
962
molecule 2D
2 0 0 0 0 0 0 0 0 0999 V2000
0.0000 0.0000 0.0000 Mg 0 2 0 0 0 0 0 0 0 0 0 0
1.7300 0.0000 0.0000 H 0 0 0 0 0 0 0 0 0 0 0 0
M CHG 1 1 2
M END
What you want to depict with PyMOL is the ionic radius (71 pm?
http://www.webelements.com/webelements/elements/text/Mg/radii.html) which I
think is not implemented within PyMOL. However as proposed previously, with
the following command you should be able to obtain the representation you
want:
alter (elem Mg), vdw=0.71
rebuild
You might even add these "new" definitions in your pymolrc file.
Hope it helped!
Cheers,
Greg
-----Original Message-----
From: Michael Weber [mailto:web...@mailer.uni-marburg.de]
Sent: samedi, 1. juillet 2006 20:02
To: Grégori Gerebtzoff
Subject: Re: [PyMOL] spheres representation and atomic / ionic radii
Well, thank you for the analysis. I will check it out more in detail.
However, what do you mean by "bond length of 1.73A" with respect to
magnesium - bound to what? Anyway, I just visually inspected sizes of
spheres of Mg2+ which has a radius of around 65 pm compared to atoms of
phosphorus and oxygen and I had the impression that the size relations are
not correct.
All the best,
Michael.
- - - - - - - - - -
This e-mail was scanned for potential virus contamination.
----- Original Message -----
From: "Grégori Gerebtzoff" <gregori.gerebtz...@unibas.ch>
To: "'Michael Weber'" <web...@staff.uni-marburg.de>;
<pymol-users@lists.sourceforge.net>
Sent: Saturday, July 01, 2006 1:33 PM
Subject: RE: [PyMOL] spheres representation and atomic / ionic radii
Dear Michael,
I do not agree with you; I never observed any variation between sphere
representation and experimental data.
But to test the accuracy of the sphere size, I did the following sequence:
1) I created a molecule composed of one hydrogen and one oxygen, with a
"fake" distance of 1.52 A between the two atoms. Here is the MOL file
corresponding to this molecule (you might save it using any text editor, the
first line should be the '962', and save it with the .mol extension).
962
molecule 2D
2 1 0 0 0 0 0 0 0 0999 V2000
0.0000 0.0000 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0
1.5200 0.0000 0.0000 H 0 0 0 0 0 0 0 0 0 0 0 0
1 2 1 0 0 0 0
M END
2) I opened this molecule in PyMOL, and displayed it with line
representation.
3) I created a new object from the O only, displayed as sphere (with a
transparency of 0.5).
4) Ray-tracing... gives the image attached to this e-mail.
One can see that the O-H line, having a length of 1.52A, corresponds EXACTLY
to the sphere of the oxygen.
I checked using the same procedure the sphere size of Mg and here I also
obtained a perfect match between a bond length of 1.73A and the sphere. So I
don't understand how you came to the conclusion that there is a mismatch
between PyMOL values and experimental values.
Best regards,
Greg
-----Original Message-----
From: Michael Weber [mailto:web...@staff.uni-marburg.de]
Sent: vendredi, 30. juin 2006 08:23
To: pymol-users@lists.sourceforge.net
Subject: [PyMOL] spheres representation and atomic / ionic radii
Hi,
I have a question concerning the "spheres" representation. According to my
rough estimation the spheres drawn for individual atoms/ions don't seem not
match experimental atomic/ionic radii if I compare different elements. Is
that correct (I quickly checked for oxygen, phosphorus and magnesium2+)?
Furthermore, is it possible to manually assign radii to selected atoms?
Best regards,
Michael.
