[Chimera-users] Centro-symmetric space groups

[Thomas Goddard]

Hi Mike,

   If there is a bug I want to fix it.  Give me the example PDB and the 
correct answer, say from some other software -- best if it were another 
PDB file with the other asymmetric units in the unit cell.  If that is 
difficult and image might suffice.

   I'm not clear on what you are saying is wrong.  Do you mean the 
placement of the asymmetric units produced by the Unit Cell dialog for 
space group "P -1" is wrong?  That would be pretty odd -- I'd expect 
many space groups to give wrong results if there was a bug.  If that is 
the issue, you may be getting tripped up by the fact that the asymmetric 
units are placed according to the symmetries but then moved so that the 
center of each asymmetric unit being placed lies within the unit cell 
box having origin at the location given in the dialog.  The purpose of 
this is to get a packed unit cell.  Probably should have an option to 
disable that behavior so you just get exactly the standard symmetry 
placement.

	Tom


-------- Original Message --------
Subject: Re: [Chimera-users] Centro-symmetric space groups
From: Michael Day
To: Tom Goddard <goddard at cgl.ucsf.edu>
Date: 8/31/09 4:45 PM

 > Tom,
 >
 > Thanks for the info! Especially the speedy response! The PDB definitely
 > allows centrosymmetric groups even though they never appear in proteins.
 >
 > Centrosymmetric space groups still get mangled. Your file appears to
 > have the right symmetry operators. I ran into this on a P-1 structure
 > that is a stack of planar molecules but when I expand the unit cell I
 > get some thing completely different.
 >
 > I've tried the PDB file with other viewers and I can get the correct
 > representation but I much prefer Chimera to everything else I've tried
 > (there's really no comparison).
 >
 > It seems one of the the most straightforward options would be to start
 > with CIF files because they have the symmetry operators in them and the
 > coordinates are the actual crystallographic coordinates and therefore
 > wouldn't need to be transformed/deorthogonalized.
 >
 > It may be a bug with your deorthogonalization routine.
 >
 > Cheers,
 > Mike
 >