What type of bond determines a molecule in a barostat?

[lewis martin]

Hi all,
I'd like to restrain a ligand to a bound pose in a (also restrained) protein pocket, in a system along with a barostat. I could use CustomExternalForce, but that is locked to an absolute position in space - and a barostat might keep moving the COMs of the ligand and protein away from that absolute position. With periodic distance in the energy esxpression, they should consistently spring back into place, but it feels like this is suboptimal - perhaps it affects the barostat efficiency, for instance?

Apparently one way around this is to add 'dummy' particles, with zero mass (and zero epsilon, q in the NonbondedForce), and then restrain the ligand atoms to the dummy particles. If the dummy particles are bonded to the real atoms, they will be considered a part of the molecule, and be moved by the barostat - thus moving the restraint relative to the rest of the system - great!

What determeines a "bond" here? A bond in the topology when creating the system? or a term in the harmonic bond force? Do CustomBondForce bonds also link atoms into a molecule? Does the MonteCarloBarostat require updating, after it's created, to perceive new bonds added later?

Thank you!
Lewis

[Peter Eastman]

The technical answer is that every Force class (or still more technically, the ForceImpl it creates) defines a getBondedParticles() method. It reports which particles are bonded to each other by that Force. In practice, the Forces that define bonds are HarmonicBondForce, CustomBondForce, CustomCentroidBondForce, and DrudeForce.

[lewis martin]

Thank you! Resolved