Hello everybody,

I would like to perform MD simulations of protein-ligand-complexes using AMBER22 and I need some advise on how to calculate RESP charges for the ligands using Gaussian16.

I know how to generate the input script for gaussian and so on. That is not the problem.

What I just can't figure out is, wether I should perform the optimization step of the ligand on the molecule alone or with the binding pocket of the protein around it (frozen for G16 calculation).

Everything that I can find about that topic (AMBER tutorials and manual, mailing list etc) says to optimize the ligand and shows examples on isolated molecules. But the ligand won't fit in the binding pocket any longer, if I optimize it on it's own.

What am I missing here?

I appreciate every help that I can get.