zaixizhang / FAIR

NeurIPS 2023 Spotlight paper: Full atom protein pocket design via iterative refinement
https://arxiv.org/abs/2310.02553
42 stars 5 forks source link

How do I run the model? #1

Closed bio-rat closed 8 months ago

bio-rat commented 9 months ago

Hi first sorry for asking a seemingly dumb question. I am new to the deep learning space from a biochemist background. The provided code only have "train.py" and "test.py" but no easy "inference.py" that specify which type of inputs to feed into the model.

Thank you!

GanQiao1990 commented 8 months ago

i also seek the information

bio-rat commented 8 months ago

If the author doesn't respond soon I will write my own inference.py in a few days!

zaixizhang commented 8 months ago

Thanks for your question! In the test.py, the input is the test split of the processed protein and ligand molecule data. We are developing code that can design protein pocket for new protein-ligand data not in the processed dataset.

bio-rat commented 8 months ago

Oh thank you for the response. When will the code for that be available? I really appreciate your work on this and it may apply to my current problem so I wish you the best!

zaixizhang commented 8 months ago

It may take several weeks. I am working on other projects and thanks for your patience.