Could you please give us more details about how to use your code base?

[jdp22]

For example,you can give us an example about where to put the checkpoint and data dir.

[jdp22]

Besides, how to transfer the .pt files into the pdb files so that we can visualize the protein.

[Ced3-han]

You can use models_con/sample.py to sample, and models_con/inference.py to reconstruct PDB files. We promise to add more demonstrations and cleaned eval scripts in the near future.

[amelie-iska]

Any updates on this? 🙏

[amelie-iska]

Just FYI for anyone who needs help, I was able to set the model up, organize the repo with the correct data, and run the model, but my run script for running on new inputs isn't producing the results I expected. I'm getting peptides with all of the residues overlapping, and the peptide is not placed near the target or near the origin. I thought maybe the peptide needed to be reconstructed using the rotation and translation matrices that the model returns, but I was not really able to figure that out. If anyone wants to discuss, feel free to reply to this.