patronov

Currently, there is no information about the batch/step number logged out. However, we are preparing to release a new version where this is available. It should be ready within the...

Hi, thank you for using Reinvent! Yes this is an issue as we dont currently provide a standardizer that would help you to sanitize the dataset you want to conduct...

We have released version 3.0 which should be (hopefully) compatible with newer versions of pytorch - specifically with 1.7. Note that there are some other third party dependencies so best...

I suspect you have installed version of reinvent-scoring that is different from `reinvent-scoring==0.0.41`. We have fixed the typo in the later library releases but we havent updated the public git...

Thanks for bringing this up, it is indeed a good point. I always thought It is a bit tricky to penalize a specific tautomer generation since the molecules are still...

Looking forward to hear about the outcome 👍

Hi, thanks for reporting this. The file has been moved to: https://github.com/MolecularAI/ReinventCommunity/tree/master/notebooks/models @GuoJeff could you please update this example: https://github.com/MolecularAI/ReinventCommunity/blob/master/notebooks/Complete_Use-Case-DRD2_Demo.ipynb it points to `/data/drd2.pkl` instead of pointing to `/models/drd2.pkl`

Which version of scikit are you using? Is it scikit-learn=0.21.3 ?

Current models are compatible with up to 0.21.3 They wont work with newer versions. In the future we will aim to become compatible with the recently released v 1.0 of...

Thanks for the positive feedback! Reinvent doesnt implement reaction rules in its standard modes. The SMILES are simply evaluated as valid if they can be correctly transformed into valid molecules...