François Bérenger
François Bérenger
I think there is a google summer of code related to this project. Probably involved people could help you if they see the code.
I agree the multiple consecutive parenthesis is the only weird thing in the proposed syntax. If ML can "understand" ring size, I guess that it could also understand "branch length".
Some methods, but not all, will certainly generate DeepSMILES which cannot be decoded to SMILES. It would be nice if the implementation is robust to those. Crashing early and loud...
there is a PR for this one!
I think this could be merged.
a -o option to say where the molecules passing std should be written to would be nice
-o FILENAME
mol_std should be printed out (in SMILES), rather than the SMILES line from the input file which passed standardization. I guess, people are interested in molecules _after_ standardization, rather than...
I vote for both.
No special use case. I just like that it's an open, high bit-rate, block cipher.