On Sat, 2004-12-18 at 02:29 -0500, Miguel wrote:
> Lets look at a few cases:
>
> Case 1
> ------
> Both molecules have atoms clustered around (0,0,0)
> The average of all the points is somewhere near (0,0,0)
> We center at the average of all the points.
>
>
> The result is that the two molecules are rendered with one directly on top
> of the other.
>
> Q: Is this a desirable thing? (I honestly don't know)
This might be acceptable behavior - assuming that there was a way to
select a specific molecule (see below) and then move it around to get a
better view.
I can also see one situation where this behavior would be expected - say
I do an alignment of two structures and save these structures to two
files. The alignment procedure will generate coordinates such that the
two molecules lie on top of each other (their center of gravity will lie
at the same point) - when these two are loaded into Jmol the user should
see that the molecules lying on top of each other and hence view the
alignment
>
> Case 2
> ------
> Molecule A has atoms clustered around (1000,1000,1000)
> Molecule B has atoms clustered around (-1000,-1000,-1000)
> The average of all the points is somewhere near 0,0,0
> We center at the average of all the points
>
> The result is that the molecules on-screen are very small ... two small
> clusters spread very far apart.
>
Once again if this situation did occur then the user could move the
molecules closer to each other by hand after loading.
> Q: Do atom coordinates every get calculated so that they are located far
> away from (0,0,0)? I don't know.
I have'nt seen HIN files with very large coordinates
One solution to the above problems could be to translate coordinates of
any loaded molecule to the origin. But this might then create problems
with cases where say a file contains 2 molecules which are placed at
specific positions. If Jmol considers all the molecules in a given file
as 1 system and translates the *whole* system to the origin then it
would work. If Jmol loads a file with multiple molecules as distinct
molecules then this idea would not be good.
I tend to the option of letting the user load multiple molecules and
Jmol placing then wherever the coordinates indicate and let the use do
the maniuplation.
To make things easier for the user there could be menu options that
allow a molecule to be centered, or some aspect of the molecule set as
the origin etc. So an example workflow could be:
1) load a molecule. This will set the origin at some point
2) load another molecule. Say the centroid of this molecule lies at
10,10,10.
3) The user can now
3a) drag the second molecule closer to the first one
3b) choose a menu option 'Center' that translates the molecule to
the origin of the first molecule
> Q: Do people ever cut out pieces of a .pdb file just to look at section?
> Probably.
I don't know
> > b) the center of the coordinate system that constitutes the current
> > viewport
Well I made this comment based on some OpenGL code I had written a very
long time ago - so my terminology and/or understanding might be way off!
My thinking was this: a XYZ file specifies some set of coordinates. To
dsiplay these coordinates, they need to be transformed to some world
coordinate system. Say the transformation is given by the unit matrix
(so the file coordinates are the world coordinates). The next step is
then to translate the world coordinates to screen coordinates - this
last set of screen coordinates would define the viewport.
> Not sure what you mean by this.
>
> Q: Are you saying that selecting a molecule means selecting a viewport?
No. My thinking is that that the viewport coordinate system (essentially
the screen coordinate system) is the same for all molecules that have
been loaded. So selection is not done by choosing viewports - rather
selections are done on the basis of atom/molecules and the center of
rotation would be defined as the center of the viewport (or as whatever
origin the used defined by using the 'Define Center' menu option)
> Q: If so, does this mean that the display 'shifts' when the user switches
> focus from one molecule to the next?
See above.
However, as I mentioned, I have little experience with 3D graphics and
my understanding of the viewport may be totally wrong. So feel free to
ignore the above
>
> Separately, and more imporantly ...
>
> Q: Is there any other molecular vision package that does this?
PyMOL
MOE
(though they're not just visualization packages)
I read somewhere that the SWISS PDB viewer also does something like this
but as I have'nt used it I can't say much more
> Q: Is there any other molecular vision packate that does it *well* ?
I think the mechanisms by which the above packages work are quite nice.
In the case of PyMOL, switch to mouse editing mode and then shift+right
click+drag rotates the molecule containing the atom that the mouse
pointer is nearest too
In the case of MOE, select the molecule (by selecting an atom and then
extending the selection to the entire molecule) and then Alt+Shift+drag
middle button translates the selected molecule, Alt+drag middle button
rotates the selected molecule
(Personally, I like the idea of extending selections - makes working
with large proteins very handy - does Jmol have this capability?)
-------------------------------------------------------------------
Rajarshi Guha <[EMAIL PROTECTED]> <http://jijo.cjb.net>
GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04 06F7 1BB9 E634 9B87 56EE
-------------------------------------------------------------------
He who is in love with himself has at least this advantage -- he won't
encounter many rivals.
-- Georg Lichtenberg, "Aphorisms"
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://productguide.itmanagersjournal.com/
_______________________________________________
Jmol-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jmol-users
