[Molvis-list] Docking in Rasmol?

Michael Sternberg sternberg at anl.gov
Mon Apr 24 10:12:22 EDT 2006


Herbert J. Bernstein wrote:
> Adding a command to undo the initial centering at any time is feasible.
> That is just what is done to write out the original coordinates when
> writing a PDB file from RasMol.  Naturally, to preserve the alignment
> among molecules that have been docked the coordinates of all but
> one molecule would have to be modified.   Here are the commands
> I would propose to add:
> 
> 1.  recenter off -- would turn off recentering for all current 
>		and subsequent loads of molecules
> 
> 2.  recenter on --  would restore the recentering of all current and
>		subsequent loads of molecules
> 
> 3.  write modified <type> <filename> -- would write out the current
>		molecule with the coordinates modified to incorporate
>		the changes made to create the display
 >
 > You would have to put in a recenter off command.  I
 > don't want to disable existing scripts by making recent off the
 > default, but it is easy to set up an initialization script for
 > RasMol, to effectively make this the default behavior.  Making RasMol
 > do the merge of multiple datasets internally on to one will take
 > a little longer, and will probably need to be to WPDB or mmCIF format

Great, that would do it!

I wonder, however, if the "write modified" command would actually be 
needed.  Maybe I'm overlooking something, but as I see it, the behavior 
of "write" could be governed wholly by the "recenter" flag (this also 
keeps parsing simple):

(a) If off, nothing changes.   This is the default.
(b) If on, the "write" command outputs the moved structure.  Upon load 
in another instance, the behavior transparently depends on the 
"recenter" flag at the time of loading, as expected.

A question is what to do when the user resets the flag from "on" to 
"off" -- should the display respond immediately?  I would suggest to 
defer this to the next "center" or "reset" command.


Regards, Michael


More information about the Molvis-list mailing list