automatic force field restraint setup
Hi all,
I was wondering if there is a quick way to set up force field restraints on a hierarchy of atoms generated by read_pdb()? Right now I have to explicitly set up three or four types of restraints myself (following the protocol in the charm_forcefield.py example in the atom module). This is likely to be re-implemented often and it would be nice if IMP provided an easy way to do this.
Maybe less useful to others, but potentially to me, would be if you could specify a region of the protein to which to add the forcefield restraints (instead of the whole PDB you just read in).
thanks, db
Yeah, it seems like one should be able to just add one restraint, not the current convoluted steps. Something like the ExcludedVolumeRestraint (which wraps a ClosePairContainer and a PairsRestraint). Easy enough\ for a simple case like the example (but I can't say I've looked at the code beyond the example).
For the second question, it is not obvious to me how to create a force field for something that is not a whole subtree in a hierarchy (eg, parts of several different atom.Hierarchies, or a few parts of the same hierarchy). Ben?
On Jul 12, 2010, at 3:31 PM, Dave Barkan wrote:
> Hi all, > > I was wondering if there is a quick way to set up force field restraints > on a hierarchy of atoms generated by read_pdb()? Right now I have to > explicitly set up three or four types of restraints myself (following > the protocol in the charm_forcefield.py example in the atom module). > This is likely to be re-implemented often and it would be nice if IMP > provided an easy way to do this. > > Maybe less useful to others, but potentially to me, would be if you > could specify a region of the protein to which to add the forcefield > restraints (instead of the whole PDB you just read in). > > thanks, > db > _______________________________________________ > IMP-dev mailing list > IMP-dev@salilab.org > https://salilab.org/mailman/listinfo/imp-dev
On 7/12/10 3:31 PM, Dave Barkan wrote: > I was wondering if there is a quick way to set up force field restraints > on a hierarchy of atoms generated by read_pdb()? Right now I have to > explicitly set up three or four types of restraints myself (following > the protocol in the charm_forcefield.py example in the atom module). > This is likely to be re-implemented often and it would be nice if IMP > provided an easy way to do this.
Indeed - this is a common task. I have such a simplified wrapper available but haven't added it yet. Note that there are several steps because "adding a forcefield" isn't just a case of adding a restraint, but also requires topology, etc. to be set up properly.
> Maybe less useful to others, but potentially to me, would be if you > could specify a region of the protein to which to add the forcefield > restraints (instead of the whole PDB you just read in).
You can create topology for any Hierarchy - doesn't have to be the entire PDB, but could be a single chain, for example.
Ben
> Indeed - this is a common task. I have such a simplified wrapper > available but haven't added it yet. Note that there are several steps > because "adding a forcefield" isn't just a case of adding a restraint, > but also requires topology, etc. to be set up properly. One should be able to just cache the topology in the particles. The rigid body related scores do similar things (a hierarchy is built if it is not found and cached in the particle).
> >> Maybe less useful to others, but potentially to me, would be if you >> could specify a region of the protein to which to add the forcefield >> restraints (instead of the whole PDB you just read in). > > You can create topology for any Hierarchy - doesn't have to be the > entire PDB, but could be a single chain, for example. Part of a chain?
participants (3)
-
Ben Webb -
Daniel Russel -
Dave Barkan