Hi (especially Greg),
I’m using the RDKit R-Group Decomposition node pretty extensively, albeit in a non-standard way, and I’m wishing there was a way to have it exhaustively match the provided cores, i.e. to give all possible matches for a given core, rather than just the first one it finds.
What I’m doing is using it to chop up long molecules (degraders), into predefined sections (ligand 1, linker section A, linker section B, [linker section …], ligand 2) with the incoming and outgoing vectors labelled [R1, R2], [R2, R3], […]. My intention is then to use molzip to stick all the combinations together, while preserving the section order (hence why I’m not just splitting on bond types). Anyway, the issue I have is that some of the definitions of “linker section B” are the same as “linker section A”, so the scaffold might match to the wrong part of the molecule. If the node gave me all matches it would be easy to filter out the wrong ones, but that doesn’t seem to be an option. If I’m right, then please could it be added at some point?
My current workaround is loop nodes, which can get pretty slow for large sets of molecules.
Cheers,
Richard