Consider allowing mixed resolution graphs for protein-ligand graphs #546
Description
It would be nice to allow for residue-chains to be at residue level while non-residue chains are atomic level. Unclear how this can be implemented and whether it is worth the effort.
Depending also on how we will solve this issue, for the resolutions it may make sense to have finer (atomic) and coarser (residue/bases/mixed) resolutions instead of atomic/residue ones.
Blocked by #544