[Molvis-list] (and JMol) polygons
Sinasi Ellialtioglu
sinasi at metu.edu.tr
Sun Feb 13 11:52:40 EST 2005
Dear Tamas,
More than ten years ago, when I discovered Rasmol in the internet, since
it is the default one, I have
been using the pdb format as I learned it from the tutorials at the
time. I know it has too many columns
for different properties to be used by Rasmol, and most of them I don't
even begin to understand, I still
did not switch to xyz format, for instance, although it would suffice
for my drawing purposes. As the
time went on I gained some confidence, I deviated from the standards of
the Rasmol world and started
to use its properties for my own wishes, like the 'temperature' column,
where I store the height of the
atom from a chosen plane, so that when I look at it from the
corresponding top view I can tell from the
red to blue spectrum of the atoms that which are closer to the surface
and which are deeper in the bulk
of the crystal or amorphous slab. I use the CPK column differently also,
since the colors of most atoms
I work with are either purple (undefined) or close to each other (like
in your version of perovskite all
three atoms are similar tones of red). So I noticed that different
letters in CPK column give different
atom colors, same goes for the CHAIN column, but this time with a
different set of color assignments.
If my aim was to construct a macromolecule with lots of atoms and use
the color codes arbitrarily
and expose it to the public in the internet, that would have been out of
standards and would have
been ugly to see a H-atom making more than one bonds, for instance.
Unfortunately, I could not
think of sending my example pdb file (I should have called it "toy.pdb"
instead of "perov.pdb" or
"abo3.pdb") to this list is also exposing it to public, so I hope I am
forgiven. It is such a toy file
that even the lattice parameter is far from the real values, so that
Rasmol does not bond the atoms
and I would be in control which one to bond to which, or which bonds to
show.
All these unscientific discussion took me away from the "polyhedron"
problem. Actually, for those
who might be wandering about the solution within the Rasmol domain, I
came up with a brute force
solution, it is not a smart way in practice, but just to say it exists:
Select all atoms on/within the polyhedron, spacefill off.
Wireframe 0.05 or so. So we just have the polyhedron picture.
Export as a gif.
Import gif into xfig.
Export as an eepic and edit it and compile it with latex.
Find out which lines make up which triangle, isolate them as closed
paths as "\path(1,2)(2,3)(3,1)".
Insert "\shade\color{cyan}" in front of "\path(1,2)(2,3)(3,1)".
Too much work.
And this reminds me of another wishful thinking (of mine!), and since I
don't know the
mechanisms and structure of any of these above software, I dare to ask:
Would it be possible technically Rasmol to export in "xfig" or "eepic"
formats?
Thanks for your time.
Sinasi
Tamas E. Gunda wrote:
>I don't know what software created this pdb file of perovskit, but it has
>many mistakes and different other programs may behave oddly. HETATM should
>be instead of ATOM (small bug), the CONECT records are incomplete (all bonds
>should be there twice, A-->B and B-->A), elemental symbols are in the column
>of residue names and vice versa, the residue names are in the column of
>atom names (big bugs). According to the recent PDB spefications in columns
>77 and 78 the elemental symbol should be present in order to avoid
>confusions when unusual atom names are tried to translate to elemental
>symbols.
>
>It should like for example:
>
>HEADER perovskite
>COMPND perovskite
>SOURCE e-mail
>SOURCE Created by Mol2Mol 5.4
>SOURCE 10:22:01, Sunday, Feb 13, 2005
>HETATM 1 Ti o 1 8.000 8.000 8.000 1.00 0.00
>Ti
>HETATM 2 O2 s 2 8.000 8.000 4.000 1.00 0.00
>O
>HETATM 3 O3 s 2 8.000 4.000 8.000 1.00 0.00
>O
>HETATM 4 O4 s 2 4.000 8.000 8.000 1.00 0.00
>O
>HETATM 5 O5 s 2 12.000 8.000 8.000 1.00 0.00
>O
>HETATM 6 O6 s 2 8.000 12.000 8.000 1.00 0.00
>O
>HETATM 7 O7 s 2 8.000 8.000 12.000 1.00 0.00
>O
>HETATM 8 Sr8 d 3 4.000 4.000 4.000 1.00 0.00
>Sr
>HETATM 9 Sr9 d 3 12.000 4.000 4.000 1.00 0.00
>Sr
>HETATM 10 Sr10 d 3 4.000 12.000 4.000 1.00 0.00
>Sr
>HETATM 11 Sr11 d 3 12.000 12.000 4.000 1.00 0.00
>Sr
>HETATM 12 Sr12 d 3 4.000 4.000 12.000 1.00 0.00
>Sr
>HETATM 13 Sr13 d 3 12.000 4.000 12.000 1.00 0.00
>Sr
>HETATM 14 Sr14 d 3 4.000 12.000 12.000 1.00 0.00
>Sr
>HETATM 15 Sr15 d 3 12.000 12.000 12.000 1.00 0.00
>Sr
>CONECT 2 3 4 5 6
>CONECT 3 2 4 5 7
>CONECT 4 2 3 6 7
>CONECT 5 2 3 6 7
>CONECT 6 2 4 5 7
>CONECT 7 3 4 5 6
>END
>
>Dr. Tamas E. Gunda
>Dept. of Pharmaceutical Chemistry
>University of Debrecen
>POBox 36
>H-4010 Debrecen, Hungary
>
>tgunda (.AT.) puma.unideb.hu
>
>----- Original Message -----
>From: "Sinasi Ellialtioglu" <sinasi at metu.edu.tr>
>To: "Molecular Visualization, especially in education with freeware"
><molvis-list at bioinformatics.org>
>Sent: Saturday, February 12, 2005 12:39 PM
>Subject: Re: [Molvis-list] polygons
>
>
>
>
>>Dear Miguel, Abhijit, and Inbal,
>>
>>I thank you all for your comments and guidiance.
>>With all I know about Rasmol I can obtain octahedra but I don't know a
>>way to make its
>>surfaces as solid or semi-transparent. I enclose a simple case of a
>>perovskite pdb file, and
>>a script file to be implemented afterwards:
>>
>>abo3.pdb
>>
>>ATOM 1 o Ti d 2 8.000 8.000 8.000 1 24.00
>>ATOM 2 s O b 3 8.000 8.000 4.000 1 60.00
>>ATOM 3 s O b 3 8.000 4.000 8.000 1 48.00
>>ATOM 4 s O b 3 4.000 8.000 8.000 1 36.00
>>ATOM 5 s O c 3 12.000 8.000 8.000 1 36.00
>>ATOM 6 s O c 3 8.000 12.000 8.000 1 48.00
>>ATOM 7 s O c 3 8.000 8.000 12.000 1 60.00
>>ATOM 8 d Sr a 1 4.000 4.000 4.000 1 12.00
>>ATOM 9 d Sr a 1 12.000 4.000 4.000 1 12.00
>>ATOM 10 d Sr a 1 4.000 12.000 4.000 1 12.00
>>ATOM 11 d Sr a 1 12.000 12.000 4.000 1 12.00
>>ATOM 12 d Sr a 1 4.000 4.000 12.000 1 12.00
>>ATOM 13 d Sr a 1 12.000 4.000 12.000 1 12.00
>>ATOM 14 d Sr a 1 4.000 12.000 12.000 1 12.00
>>ATOM 15 d Sr g 1 12.000 12.000 12.000 1 12.00
>>TER
>>#CONECT 1 2 3 4 5 0 0
>>#CONECT 6 1 0 0 0 0 0
>>#CONECT 7 1 0 0 0 0 0
>>#
>>CONECT 2 3 4 5 6 0 0
>>CONECT 7 3 4 5 6 0 0
>>CONECT 6 4 5 0 0 0 0
>>CONECT 3 4 5 0 0 0 0
>>
>>and abo3.spt
>>
>>set ambient 60
>>background white
>>set boundbox on
>>color boundbox black
>>spacefill 0.45
>>wireframe 0.15
>>color atoms cpk
>>rotate y 15
>>rotate x -15
>>select 3
>>spacefill 0.10
>>wireframe 0.01
>>
>>If I had only a way to shade a closed surface bounded by CONECT'ed
>>
>>
>lines....
>
>
>>Anyway, I will visit JMOL solution since it is already there.
>>
>>Thank you all again.
>>Sinasi
>>
>>Miguel wrote:
>>
>>
>>
>>>>Hi everybody,
>>>>
>>>>I am a Rasmol user, and since this list is now bringing together a
>>>>multi-software society
>>>>I would like to ask if drawing polygons around oxygen octahedra or
>>>>tetrahedra in various
>>>>crystal structures (like in YBCO) is possible in any of Rasmol, Jmol,
>>>>Chime, Protein Explorer, or else.
>>>>
>>>>
>>>>
>>>>
>>>I have recently added rendering of polyhedrons (tetrahedrons and
>>>octahedrons) to Jmol. They can be rendered as solid or translucent.
>>>
>>>The code has not yet been officially released, but is currently in
>>>prerelease test phase.
>>>
>>>There is not much documentation either, but there are only a few commands
>>>and it is pretty straightforward.
>>>
>>>Check out www.jmol.org
>>>
>>>Download the prerelease version from sourceforge at:
>>>
>>>http://sourceforge.net/project/showfiles.php?group_id=23629
>>>
>>>Then you can post specific questions on the jmol-users list. Sign up at
>>>www.jmol.org/lists
>>>
>>>
>>>Miguel
>>>
>>>_______________________________________________
>>>Molvis-list mailing list
>>>Molvis-list at bioinformatics.org
>>>https://bioinformatics.org/mailman/listinfo/molvis-list
>>>
>>>
>>>
>>>
>>>
>>_______________________________________________
>>Molvis-list mailing list
>>Molvis-list at bioinformatics.org
>>https://bioinformatics.org/mailman/listinfo/molvis-list
>>
>>
>>
>
>_______________________________________________
>Molvis-list mailing list
>Molvis-list at bioinformatics.org
>https://bioinformatics.org/mailman/listinfo/molvis-list
>
>
>
More information about the Molvis-list
mailing list