Specifying component B to A in ligand mapping

[ijpulidos]

Currently the LigandAtomMapping class has a componentA_to_componentB parameter that we can use to define the mapping. Some packages or tools might expect a different "directionality" in the mapping, for example in perses is inverted compared to the convention chosen by gufe (hopefully for convenience, but maybe it was just arbitrary on the perses side).

Do we want to have a componentB_to_componentA option for the constructor of a LigandAtomMapping object? That said, I'd think that we want at least to have it unambiguously defined in the docstring, even for the current parameter.

[richardjgowers]

@ijpulidos you mean this? https://github.com/OpenFreeEnergy/gufe/blob/main/gufe/mapping/ligandatommapping.py#L59

[ijpulidos]

Oh, that helps, but I meant this https://github.com/OpenFreeEnergy/gufe/blob/main/gufe/mapping/ligandatommapping.py#L25 , that is, when creating the object.

[dotsdl]

Conclusion: a clarifying docstring specifying what the keys, values are in the componentA_to_componentB and componentB_to_componentA property dicts.