The potential of mean force (PMF) between two nanocrystals (NCs) represents an effective interaction potential that is essential when explaining the assembly of NCs to superstructures. For a given temperature, the PMF is obtained best from molecular dynamics simulations. Based on a density functional approach, this study proposes three methods of predicting the PMF for any given temperature based on a single molecular dynamics simulation for one temperature. The three methods construct the PMF by considering the ligands as an ideal gas, as hard-sphere chains, or as Lennard-Jones interaction sites. To apply this methodology, the density of the interaction centers must be extracted from the simulation data. For the ideal gas model, a straightforward sampling procedure with a fixed lattice in space leads to free energies that are too large in order to consistently explain the simulation data for different temperatures. Naive sampling does not account for the small momenta added to the NCs when coupled to a thermostat. A method is proposed that corrects for the unphysical steps during the simulation. The ideal gas contribution computed for the corrected density is significantly smaller than the one obtained from naive sampling and can thus explain the temperature dependence of the PMF correctly. For the hard-sphere chain model, where a weighted density is used, the correction of the particle density is not essential. However, the PMF calculated based on the corrected density confirms our approach. All three models predict PMF curves in very good agreement with simulation results, but they differ in the number of input parameters and the computational effort. Based on the modeling results, we predict the existence of an additional attractive force at small distances of the NCs - a depletion force.