Menu
▾
▴
#263 minimum bounding box
Hello, Jmol developers and users
I'm interested in building a minimum bounding box of the vdw isosurface of molecule.
The default setting for bounding box of the vdw isosurface is defined by the minimum and maximum coordinate values of the vdw isosurface.
This bounding box is not a minimum bounding box.
Could you please teach me to build a minimum bounding box of the vdw isosurface with a sample Jmol script ?
Thanks for advice,
Sawada
Hello, Toshihiko
Could you please define more precisely what is yor idea of "minimum" bounding box?
Based on atom coordinates rather than vdW sphere outer coordinates, maybe?
[feature-requests:#263]https://sourceforge.net/p/jmol/feature-requests/263/ minimum bounding box
Status: open
Group: v14
Created: Fri May 19, 2023 05:54 AM UTC by Toshihiko Sawada
Last Updated: Fri May 19, 2023 05:54 AM UTC
Owner: nobody
Hello, Jmol developers and users
I'm interested in building a minimum bounding box of the vdw isosurface of molecule.
The default setting for bounding box of the vdw isosurface is defined by the minimum and maximum coordinate values of the vdw isosurface.
This bounding box is not a minimum bounding box.
Could you please teach me to build a minimum bounding box of the vdw isosurface with a sample Jmol script ?
Thanks for advice,
Sawada
Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/jmol/feature-requests/263/
To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/
Related
Feature Requests: #263
You mean "best" in any orientation, not just xyz axes?
On Fri, May 19, 2023, 12:54 AM Toshihiko Sawada gif220@users.sourceforge.net wrote:
Related
Feature Requests: #263
Hi, Angel and Bob
Thank you so much for response and discussion.
To share the idea, i post samples of Jmol command (CASE 1 and 2), input file (sample_S.pdb), and BoundingBox.jpg (CASE 1-3).
-- i am able to build a bounding box of molecular vdW sphere (see CASE 2), BUT this bounding box is not the "minimum" bounding box.
-- Minimum bounding box is defined as the minimal volume oriented bounding 3D box.
-- CASE3.jpg illustrates my freehand-made minimum bounding box. My goal is an automatic building of this minimum bounding box by using Jmol and Jmol script.
-- Algorithms to find the minimum bounding 3D box are reported. but it is too difficult for me to implement the algorithm in Jmol.
CASE 1. default setting of bounding box of molecule
$ load sample_S.pdb
$ boundbox on
$ show boundbox
boundbox {10.139999 -11.542 -0.0024999976} {18.886 16.107 0.99050003}
boundbox corners {-8.746 -27.649 -0.99300003} {29.026 4.5650005 0.98800004} # volume = 2410.4556
-> CASE1_BoundingBox.jpg
CASE 2. default setting of bounding box of molecular vdW sphere
$ load sample_S.pdb
$ isosurface ISO resolution 6 color white vdw sasurface 0
$ boundbox $ISO
$ boundbox on
$ show boundbox
boundbox {10.140104 -11.5915365 0.017665863} {19.9828 17.255083 2.1100953}
boundbox corners {-9.842695 -28.846619 -2.0924294} {30.122904 5.6635475 2.1277611} # volume = 5820.569
-> CASE2_BoundingBox_vdW_sphere.jpg
CASE 3. the minimum bounding box of molecular vdW sphere
i illustrate freehand-made minimum bounding box of molecular vdW sphere by using image editor.
My goal is an automatic building of this minimum bounding box by using Jmol and Jmol script.
-> CASE3_BoundingBox_vdW_sphere_freehand_minimum.jpg
Thanks for advice,
Sawada
Hi everyone
Finding "minimum perimeter bounding box" is one of the famous optimization
problems. So, as far as I know, there is no way you can do that in Jmole as
Jmol has not been designed to go that deep in optimization methods.
In a previous study (https://pubs.acs.org/doi/abs/10.1021/acsami.2c12485),
I applied Jmol to read the structure file and to export the xyz coordinates
to a text file. I then applied the MATLAB Global Optimization Toolbox to
read the coordinates and to calculate the minimum perimeter bounding
vertices/edges. You find an algorith and write your own code to do that
(There are many algorithms for the problem. You should look for the one
that works more efficiently. ). I managed to do this in a batch mode and
in a parallel mode for over 20,000 periodic Metal-Organic gigantic
structures in a couple of days.
I hope this helps and this finds you well.
Best regards
Farhad
On Sun, May 21, 2023 at 9:43 AM Toshihiko Sawada gif220@users.sourceforge.net wrote:
Related
Feature Requests: #263
Hello, Sawada
Not perfect, but it approximates to what you are looking for (at least wih your example molecule):
rotate selected best;
boundbox {all};
The 1st command changes the coordinates of all atoms.
The 2nd forces a recalculation of the boundbox to the new coordinates.
Not sure how it will work with other molecules.
Jmol can do this. But it is specifically designed for orienting molecules in the "best" orientation. It does so by testing a set of 1495 quaternion-based orientations. So it is not a formal mathematical "best" box. But it is pretty close.
As Angel points out, that means that you can rotate the molecule to the "best" orientation using the ROTATE SELECTED option, which does change the coordinates, by the way.
rotate SELECTED BEST
and if you then create a temporary vdw surface, you can create the bounding box for that using
isosurface ID s VDW
boundbox $s
isosurface ID s delete
I'm curious what your use case is here. What is your plan for this bounding box? Do you just need its corner coordinates? Its volume? Just making a nice-looking box? Or are you doing this for some other purpose?
Hi, Bob, Angel, Farhad and all
thank you so much for your advice and opinions.
i have confirmed the workings of Jmol commands "rotate selected best", and accepted this approach as the second best of.
Thank you again for your time.
Sawada
Sawada,
How will you use bounding box?
Miguel
On Tue, May 23, 2023 at 02:57 Toshihiko Sawada gif220@users.sourceforge.net
wrote:
Related
Feature Requests: #263
Hi, Bob, Michael, and all
The minimum bounding box (MBB) of the molecule descripts its molecular dimensions.
In many cases, molecular dimensions are related to molecular properties.
Calculation of MBB with molecular design will motivate chemists to control molecular properties.
Sawada
Ah. Ok. So it is the calculation not the visual. Then we need the vdw
isosurface and just the calculation of the box. And not do the quaternion
rotation itself at all.
And we would need to figure out how good the quaternion analysis is,
perhaps allowing for a high precision option.
That is doable. How exacting does it need to be?
Do you already know of a quaternion solution?
The reason to go with quaternions is that they provide an unbiased sampling
of orientation.
Not that I have any time at all for this right now....
On Tue, May 23, 2023, 7:00 AM Toshihiko Sawada gif220@users.sourceforge.net
wrote:
Related
Feature Requests: #263
Toshihiko
I am not still convinced with the solution Bob offered because it is not
clear how Jmol does calculate the box and whether there is a mathematical
proof that it ends with the minimum bounding box.
As I mentioned in my earlier email, finding the minimum perimeter bounding
box is a well-discussed optimization problem. In order for it to be used as
a molecular descriptor/property, the method must give the real minimum
bounding box not just an approximated vale of that.
Best regards
Farhad
On Tue, May 23, 2023 at 8:00 AM Toshihiko Sawada gif220@users.sourceforge.net wrote:
Related
Feature Requests: #263
Hi, Bob, Farhad, and to whom are interested in MBB of molecule
Thank you for your interest in the Minimum-perimeter Bounding Box (MBB) of molecule.
I would like to cooperate in implementation of MBB calculation function to Jmol with keeping the original Jmol license (the GNU Lesser General Public License).
I agree with Farhad's suggestions. We need to search the reported mathematical algorithms with their approximations in advance of the implementation. There is offen a trade-off between computational complexity and accuracy.
To Bob and Jmol developers:
Would it be possible to set a formal discussion room more closed than this sourceforge room ?
this sourceforge room is too open...
Best regards,
Toshihiko Sawada
Sure. Switch to private email for a start.
Ps my naive intuition is that this problem as an adequate approximation is
just the minimization of the desired metric on a quaternion S2 sphere,
which suggests that an incrementally increasing geodesic search should pin
it down sufficiently rapidly.
I'm surprised I'm not seeing a quaternion based algorithm out there.
That would be relatively easy to implement in Jmol for testing and
visualization of the solution space. And for validation.
I have developed a simplex method using quaternions that is proving
interesting. It allows visualization of the "minimum bounding box
landscape" For example, with $isopranol, I can see this minimum box
[image: image.png]
and a quaternion landscape that looks like this:
[image: image.png]
This indicates multiple very nearly minimal bounding boxes. It's part of a
much larger set, where the colors here indicate volume, with red being
smallest:
[image: image.png]
(this is 20000 samples)
On Thu, May 25, 2023 at 6:55 PM Bob Hanson hansonr@users.sourceforge.net
wrote:
--
Robert M. Hanson
Professor of Chemistry
St. Olaf College
Northfield, MN
http://www.stolaf.edu/people/hansonr
If nature does not answer first what we want,
it is better to take what answer we get.
-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
We stand on the homelands of the Wahpekute Band of the Dakota Nation. We
honor with gratitude the people who have stewarded the land throughout the
generations and their ongoing contributions to this region. We acknowledge
the ongoing injustices that we have committed against the Dakota Nation,
and we wish to interrupt this legacy, beginning with acts of healing and
honest storytelling about this place.
Related
Feature Requests: #263