ADL: Binding energy in AutoDock

Amarendra Nath Maity anmaity at
Wed Nov 25 06:46:55 PST 2020

Dear All,
Recently, I have performed docking of the same ligand several times using
identical parameters using AutoDock 4. However, the docking results produce
different binding energy (For example, in my case: -8.88; -9.27; -9.41,
-8.98) each time for the lowest energy conformation of ligand. The ligand
that I am using at present has 12 torsions. Later I would like to
use ligands having 14 torsions. I am using ga_run 100; ga_pop_size 150;
ga_num_evals 25000000; ga_num_generations 27000. It takes approximately 15
hours. Stefano Forli informed in a previous communication that my problem
arises because the search is not sufficient, and I have to repeat the
calculation more times and/or increase the search parameters (more ga_runs,
more ga_evals, etc.) Now my questions are as follows: Which parameters
would you suggest to change and to what extent, keeping the computation
time in mind, to get a consistent binding energy? Are there any maximum
limits for those above mentioned parameters? Should I change other
Thank you in advance.
With Regards,
Amarendra Nath Maity

More information about the autodock mailing list