Bonds between symmetry partners

[tic20@…]

The "sym" command should establish bonds between symmetry partners that are indicated in the original data. This is probably not easy. The info is in the SSBOND and LINK records for PDB files and who-knows-what for mmCIF files.

[tic20@…]

This is going to be tricky in a few different ways:
- when adding hydrogens. If you only have the asymmetric unit loaded but there is a bond to a symmetry equivalent molecule in the crystal, do you add hydrogens as if the bond is there, or not?
- Does ChimeraX actually allow bonds between atoms in two different AtomicStructure molecules, or would they need to be merged into a single, larger molecule?

 
 
Tristan Croll
Research Fellow
Cambridge Institute for Medical Research
University of Cambridge CB2 0XY
 

 

[Eric Pettersen]

Yes, tricky.  Certainly for any bonds they have to be combined.  Chimera(X) does not allow bonds between AtomicStructures.

—Eric

[Tom Goddard]

Not easy. Bonds cannot be made between different models. So the symmetry copies would have to combined in a single model. The sym command does not currently support building a single structure that includes all the copies. It would have to change the chain identifiers for all the copies. For visually depicting such a bond a pseudobond can be used since that can join different models. But then none of the tools that use chemistry will work as expected on it (e.g. add hydrogens) since the pseudobond is not a real bond.

A second aspect of this is reading the SSBOND and LINK data from PDB files and corresponding information from mmCIF files.

A related problem is to have just one copy of atoms that are on a symmetry axis. Tristan says he has seen such cases.

The best hope for handling this stuff is that the sym command would have an option to combine all copies into a single model and add bonds crossing asym units using PDB / mmCIF file data.