Implement more general angle and torsion types for heteropolymers

[cwalker7]

When we look at more complex heteropolymers (say, more than 3 residue types), setting angle and torsion parameters gets tedious, as one needs to figure out all the possible sequences and assign the appropriate parameters. Setting default parameters and then modifying the rest is less effective for such systems.

Just as all-atom force fields often have more general bonded types that encompass multiple atom types, each coarse-grained particle type could belong to broader classes for determining bonded interactions.

For example, if we have a sequence of particles bb1-bb2-bb3 along the backbone, where each particle has slightly different LJ parameters, and want bb1-bb2-bb3 to have the same angle parameters as bb2-bb1-bb3, bb1-bb3-bb2, etc..., we would assign an angle type of 'B' to bb1, bb2, and bb3. Then we can just set the angle parameters for B-B-B.