Consider using DCD or other formats for trajectories

[benmwebb]

RMF trajectories are slow to read in Chimera, and RMF support is unlikely to be added to ChimeraX unless we really really push for it. We can't reasonably use mmCIF files to look at trajectories (since it's a text file, they'd be enormous and probably even slower to use). There are however a number of established formats for MD trajectories, which Chimera has good support for. Since many of these (like DCD) are basically just binary dumps of the coordinates, they are compact and very fast. They are, however, designed for atoms having 3D (xyz) coordinates.

Investigate whether we can adapt a format like DCD to store coordinates of our coarse-grained models (potentially including non-Cartesian data like Bayesian nuisances). This could be used in concert with an mmCIF file (or modified PSF file) to efficiently display trajectories in ChimeraX.