Interpreting DiffDock-PP output as a docked protein complex

[cshih14]

This issue is regarding DiffDock-PP: Rigid Protein-Protein Docking.

I was able to fix the output PDB files according to ketatam/DiffDock-PP#10 to get visuals that look more like proteins. However, when combining the ligand and receptor PDBs, the ligand and receptor appear to have overlapping coordinates as if they are just smashed together rather than docked as a complex. Here is a screenshot of the combined PDB visualized:

In the visualization/ output directory I get ligand PDB files with numbers appended 0 through 40, plus a ‘-gt' appended ligand PDB and a receptor PDB. Still trying to figure out the meaning of these outputs.

Is combining the ligand PDB appended with ‘-40’ (final timestep?) with the receptor output PDB into one single "complex" PDB the correct way to get the output complex?