This does not seem to be a question concerning any of the examples in this forum.
The most reliable methods to compute free energies require a special simulation ("free-energy perturbation","thermodynamic integration","potential of mean force") and are not used with a standard trajectory.
Your question should have been asked BEFORE you started your simulations.
MM-PB/SA (or variants) can be used, but may be of lesser quality. The LIE method of Johan Åqvist is another possibility, but you still have to have a trajectory of the free ligand, and you should check for convergence of your results in all cases.
