Handle Models about pdbtbx

[douweschulte]

As of now Chimera and PyMol can both open the ensemble file, so I guess it was linked to #38. But the question remains how strict to be on validating of models, so do HETATMs also need to correspond to each other

[douweschulte]

Based on the file format definition (v3.30) also the HETATMs should correspond, but I have seen examples where this is not the case. So I would propose to issue a LooseWarning when finding HETATMs that do not correspond.

This record is used only when more than one model appears in an entry. Generally, it is
employed mainly for NMR structures. The chemical connectivity should be the same for each
model. ATOM, HETATM, ANISOU, and TER records for each model structure and are
interspersed as needed between MODEL and ENDMDL records.