Hannes Stärk

@amfaber Can you look into these issues that seem to appear with your multiligand inference file?

Hi! The OT matrix was used in computing the returned ot_dist. The model then receives the ot_dist as additional loss and is "encouraged" to decrease it by producing keypoints that...

Hey! 1. I am not sure how a one-to-tone correspondence between Keypoints and pocketpoints could be possible considering that the sets have different cardinality. This is the reason why the...

You could investigate the results in PyMol or a similar tool. EquiBind itself does not generate any predictions of which interactions should be happening or something similar.

Hi! 1. unfortunately, I missed to note down the PDB ids of these examples and do not remember them. 2. We do not use any code to isolate ligands and...

1.1 To clarify why there is no mismatch, here are the lines of code through which rdkit coords are used instead of the PDBBind structure: Here during preprocessing: https://github.com/HannesStark/EquiBind/blob/bdc9c4c32d49681c670abf63bc11a9d4d2a8e090/commons/process_mols.py#L882-L883 https://github.com/HannesStark/EquiBind/blob/bdc9c4c32d49681c670abf63bc11a9d4d2a8e090/commons/process_mols.py#L820-L826...

1. EquiBind U indeed refers to running EquiBind without the point cloud fitting. You are correct that we do not apply any distance geometry regularization in EquiBind-R since it only...

Did you observe the generated sequences often being quite "unrealistic" with e.g. the same residue types repeated over and over?

Are you sure that you are using the same .fasta file for both steps? @JuLieAlgebra Would you be able to describe the specific setting and procedure you run for one...

This sounds to me like there were issues when running inference. Then the results were not placed in “--out_dir results/user_predictions_small” Then they are not in the list when listdir lists...