Please read :

http://www.yasara.org/docking.htm

http://www.yasara.org/dock_runensemble.mcr

To find ligands that might bind to your protein, you either identify potential ligands according to the biological role of your protein, e.g different metabolites in the metabolic pathway you are studying, or you do a virtual screening of a compound library e.g. the ZINC library or a subset thereof:

Article ZINC − A Free Database of Commercially Available Compounds f...

https://zinc.docking.org

As you have done homology modeling based on a template protein, you can check all the available template structures if any of those has a bound ligand in them. If there is any, you can use that ligand as a positive control for docking. You can also search in the literature if there are any known drugs/inhibitors, which target any of the template proteins. Those molecules may have a good chance of binding your homology model. Those can also be used as positive controls in docking and doing control experiments is a good practice. Finally, if you want to find new inhibitors for your hypothetical protein, you can go for virtual screening (many many dockings, usually automated) against a large and unbiased dataset of ligands such as ZINC as suggested by Annemarie Honegger.

check literature for ligands or inhibitors for that enzyme, hopefully with experimentally measured affinity (IC50, Kd) so you can do validation for docking technique

https://www.bindingdb.org/bind/index.jsp