search

rxn4chemistry / rxnmapper

RXNMapper: Unsupervised attention-guided atom-mapping. Code complementing our Science Advances publication on "Extraction of organic chemistry grammar from unsupervised learning of chemical reactions" (https://advances.sciencemag.org/content/7/15/eabe4166).
http://rxnmapper.ai
MIT License
286 stars 68 forks source link

Can you share the code to reproduce the figure 4(A)? #35

Closed jaewanlee93 closed 1 year ago

jaewanlee93 commented 1 year ago

Related the previous issue #34 Can you share the code to reproduce the figure 4(a) of Schwaller et al., Sci. Adv. 2021 ? As you said, I tried to reproduce the result with the same rdkit version('2020.03.3'), but I couldn't reproduce it. Or can you tell me what functions in the code you used?

pschwllr commented 1 year ago

If it matched the original mapping or one of the alternative correct_maps after canonicalization, it was counted as correct. That's how we got this figure.

It was a loop with a combination of https://github.com/rxn4chemistry/rxnmapper/blob/448118388a5f33c525731b9d40df45e9a1a3760d/rxnmapper/smiles_utils.py#L340 and https://github.com/rxn4chemistry/rxnmapper/blob/448118388a5f33c525731b9d40df45e9a1a3760d/rxnmapper/smiles_utils.py#L301 to check all possibilities.

I would share the exact code. However, when I left IBM Research, I had to leave all my code behind. So, I do not have access to this specific piece of code anymore. Sorry for that - I will make it better in future publications.

jaewanlee93 commented 1 year ago

Thanks for replying. I'll try it in various ways. I'll close this issue now.