The DOPELoopModel class is derived from LoopModel, and is very similar in operation, except that a newer DOPE-based loop modeling protocol is used (see Selection.assess_dope()). The main differences are:
To use, simply use 'DOPELoopModel' rather than 'LoopModel' in your Python scripts (see Section 2.3 for examples). Note that it will be significantly slower than the regular loop modeling procedure, primarily due to the GB/SA interaction.