ASE simulations that don't use OpenMM don't have a simulation class suitable for the OmniRunner

[hjstroud]

Currently we can run simulations using OpenMM with or without ASE (OpenMMSimulation and ASEOpenMMSimulation, respectively), but we don't have a corresponding ASESimulation class for running simulations using ASE via the OmniRunner. This is relevant for the basic_example.ipynb and openmm_neuraminidase.ipynb notebooks in the ase examples folder.

[Ragzouken]

Looking over it, these are examples that deliberately avoid using the runners, instead using NanoverASEDynamics. I think it can be made to work with the Omni runners easily, but I think need to avoid reconstructing the dynamics when resetting.

[Ragzouken]

Wait for #255