docking results from AutoDock Vina are not aligned with the experimental structure

[SoodabehGhaffari]

Hello, I have used AutoDock-vina (vina scoring function) for docking a ligand inside a target which I have the experimental result. The docking result is not aligned with the experimental result even though I used exhaustiveness=64 and I used the center of the grid box as the box from the experimental structure (the ligand is placed in a different angle compared to the experimental structure). I was wondering if there is anything that I can modify to improve the result.

I have a couple of questions:

1. The ligand is initially placed in the binding pocket but not exactly in the same way as it is in the experimental structure. The reason is that I am going to run docking for other molecules for which I do not have experimental results. I believe the docking should do this job.
2. The output of docking shows 20 modes, but the pdbqt generated in the output only shows one structure. Could you please explain it?
3. How can I take flexible residues and energy range cut off into account in the Python script?

Any suggestion is greatly appreciated. Best

[diogomart]

1. i agree
2. if the energies are positive it is possible that the box is too small
3. see the python docs in the vina readthedocs page

[SoodabehGhaffari]

Hello, Thanks for the response. I believe I did not clearly explain my first question. My ligand is in the binding pocket, but the resultant structure from docking is not aligned with the experimental structure. The angle between them is significantly large. The binding energy is negative so the box size is appropriate. I was wondering if you have any suggestions to improve the docking result. The ligands have 3 rings and are prepared with default flags (without rigid macromolecule).

[diogomart]

See "Why do I not get the correct bound conformation?" in https://autodock-vina.readthedocs.io/en/latest/faq.html for some pointers. To that I'll add keeping/removing water molecules, and note that meeko+vina can sample macrocycles but can't sample small rings (e.g. chair conformers in 6-member rings).

[SoodabehGhaffari]

Thank you for the help. I have two more questions:

1. One of the reasons mentioned in that link is the protonation state of ligands and receptors. I prepared the receptor by adding the flag (-A "hydrogens") and I added hydrogen to the ligand using RDkit library before preparation. Do you have any suggestions about evaluating the protonation state?
2. To dock small rings, do you suggest using AutoDock 4 or other tools?

[diogomart]

1. There's a lot of literature on this topic, for example: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4437766/ https://jcheminf.biomedcentral.com/articles/10.1186/1758-2946-6-12

2. No, autodock4 can't sample small rings either.