Ring Conformers: fixing major degrees-to-radians bug in my previous push
118 degrees != 118 radians
Carbohydrate ring conformers now look good, even in strained cases.
I also edited cyclohexylglycine params so that AtomTree does not flip out on
ring conformation changes.
(I was too quick to celebrate on my previous push; it is more complicated to
make a ring system work with the RingConformationMover. One needs to ensure
the AtomTree behaves properly by defining a CUT_BOND and at least one virtual
atom. I'll document this all later once I refine things more.)
Only the rings integration test changes, as expected.