Azurin protein potentially plays an important role as an anti-cancer therapeutic agent, particularly in treating breast cancer in experiments and showing without having a negative effect on normal cells. Although the interaction mechanism between protein and lipid membrane is complicated, it can be modeled as protein-lipid interaction. Since the all-atom (AA) model simulation is cost computing, we apply a coarse-grained (CG-MARTINI) model to calculate the protein-lipid interaction. We investigate the binding free energy value dependency by varying the windows separation and electrostatic scale parameters. After scaling the electrostatic interactions by a factor of 0.04, the best result in terms of free energy is -140.831 kcal/mol, while after window-separation optimization, it reaches -71.859 kcal/mol. This scaling was necessary because the structures from the CG MARTINI model have a higher density than the corresponding all-atom structures. We thus postulate that electrostatic interactions should be scaled down in this case of CG-MARTINI simulations.