You should prepare .gro file containing all three ligands and the target. I assume that you have already the topology for one ligand (the .itp file). Then it is more than simple - just edit the [ molecules ] directive in your .top file and put there the right number of ligands, i.e. 3.

Guys I don't know about GROMACS, but I have done similar stuff using NAMD using CGENFF forcefield. What I have done is that, after obtaining a docking poses I first generated their topology and parameters. There is online server called PARAMCHEM (https://www.paramchem.org/index.php) that generates topology and parameters for ligand using analogy of previously determined parameters from existing ligand. I would suggest to better have a look at parameter file for reliability score for each parameter before you proceed for simulations. Once you have proper topology and parameters for ligand, you just need to create a single PDB file containing your protein as well as all the three of your ligand. And finally you are ready to roll... Moreover, so far as i know topology and parameters file generated by ParamChem is compatible to gromacs as well. But you better check it out. I am providing a pdf tutorial file for your quick start. Hope this helps

Best luck for your MD simulation

Please provide detail methodology and research article on multi ligand simulation