Implement Surge structural isomer-finding algorithm

[bertiewooster]

I recently learned about Surge, an open-source chemical graph generator aka structural isomer finder, and wondered if it's been considered for incorporation into the RDKit. Its performance seems impressive: according to Surge's GitHub page, it can "generate the 452458 isomers of C10H16O in 0.1s on some vanilla flavor year-2021 PC" and seems ~50x faster than MOLGEN. It has the Apache License Version 2.0 and is implemented in C. Perhaps this could be a GSOC project.

I couldn't find any structural isomer-generation algorithm in the RDKit; is there one?

BTW I learned about Surge due to its incorporation into WolframChemistry.

[greglandrum]

I don't think something like this is a great fit for the core RDKit, but it would be cool to have some examples of using one of these libraries together with the RDKit to do something interesting.

BTW: here's another one, from our group: https://jcheminf.biomedcentral.com/articles/10.1186/s13321-022-00677-6

[bertiewooster]

Thanks. Good to know about your group's enu, which I understand enumerates stereoisomers as well as constitutional isomers, and is part of CombiFF on GitHub.

Has your group considered putting CombiFF on PyPI to simplify adoption, e.g. including pre-built wheels for chips including Apple Silicon (ARM chips M1, M2, and M3)?