You have two main commands:

g_lie -> computes a free energy estimate;

g_bar -> calculates the free-energy estimates

But you'll must have a trajectory without the ligand because the ligand free energy binding is calculated subtracting the calculated free-energies of the protein only MD from the Protein+Ligand MD.

Another way, a lot longer and more expensive in time and computational resorts, is for you to do a Umbrella Sampling.

PS: you have PLUMED, an add-on for GROMACS, exclusively for free-energy calculations.

You have two main commands:

g_lie -> computes a free energy estimate;

g_bar -> calculates the free-energy estimates

But you'll must have a trajectory without the ligand because the ligand free energy binding is calculated subtracting the calculated free-energies of the protein only MD from the Protein+Ligand MD.

Another way, a lot longer and more expensive in time and computational resorts, is for you to do a Umbrella Sampling.

PS: you have PLUMED, an add-on for GROMACS, exclusively for free-energy calculations.

New update:

you now have g_mmpbsa, an add-on to GROMACS, that uses APBS to calculate free energies. Look here: http://rashmikumari.github.io/g_mmpbsa

Ricardo