Pretrained Models For Inference?

[arazthexd]

Hi! I was looking for my options to have a rough estimation of distributions of different protomers of some molecules and I found your great paper and repo. I wanted to try to test your model that was reported in the paper on my molecules, but I don't seem to find it.

Is there a chance to have access to the trained models for inference?

Thanks,
Araz

[Benzoin96485]

We're sorry for the late reply due to the article submission and revision process. We hope this demo could help you. Another module for direct micro-pKa and macro-pKa prediction is also under active development.

Our team is also discussing about publishing model weights. I will sync you with the latest update.

[arazthexd]

Thank you for your response. I will definitely check that out.

I look forward to seeing more of your work in the future and hope to hear from you soon.

[mchakra8]

I have also noted with great interest your recent publication regarding Uni-pKa on July 17th in JACS Au. As arazthexd above has mentioned, I too am interested in using your pre-trained model to perform inference calculations. Now that the paper has been available for a couple of months, is there any plan to release a model checkpoint or some other file so that we could use the model that you describe in your publication?

[ZhouGengmo]

I have also noted with great interest your recent publication regarding Uni-pKa on July 17th in JACS Au. As arazthexd above has mentioned, I too am interested in using your pre-trained model to perform inference calculations. Now that the paper has been available for a couple of months, is there any plan to release a model checkpoint or some other file so that we could use the model that you describe in your publication?

Hi,
The files related to data preprocessing and the details of the reproduction workflow are ready to be added to the repository.
You can refer to this pr #6.

And we will also release the model ckps ASAP.