Its possible to calculate the changes in the translational, rotational, vibrational, and total entropy of ligand while binding to the receptor using Schrodinger suite.

I think the primary purpose of molecular docking is to obtain 'reasonable' protein-ligand complexes with low computational cost. In order to obtain 'reliable' prediction of the energetics of protein-ligand association one needs to perform molecular dynamics simulations of the 'ligand' in water/solution and of the 'protein-ligand' complex in water/solution and calculate the binding free energy from the differences in non-bonded interaction energies in the two environments, e.g. using Linear Interaction Energy or Linear Response Approximation methods.

As mentioned by Martin, in order to compute the thermodynamics with great accuracy, you have to go for Molecular Dynamics of the protein-ligand complex. However, there are other approaches too, in build in the suites like GBSA and free energy perturbations that also works on the same principles