Randomize ligand conformation - results include starting pose

[and-tos]

I want to randomize the starting pose of my ligand to remove bias from the initial pose, but it seems that the original conformation is still included when I run the following lines. Is there an error in my code? Or am I misinterpreting the meaning of the randomize method?

n_poses = 20
v = Vina()
v.set_receptor(str(rec_pdbqt))
v.set_ligand_from_file(str(lig_pdbqt))
v.compute_vina_maps([0.0, 0, 0], [30, 30, 30])
v.randomize()
v.dock(exhaustiveness=16)

[diogomart]

The randomization affects only rotatable bonds and the overall orientation of the ligand. Bond lengths and angles are not changed. Does this explain what you observe?

Our usual protocol for removing bias from the x-ray positions of the ligand is to create 3D coordinates from smiles with https://github.com/forlilab/scrubber

[and-tos]

Thanks for clarifying, is there a way to export the randomized structure?

[diogomart]

Unfortunately, we overlooked that and didn't expose it to Python. Only from command line. Sorry about that..