Protein-protein interactions play a very important part of various biological process,in the study of protein-protein interactions,it is very important to calculate the binding free energy.We can judge whether the bound structure is stable by combining binding free energy,but it is still a huge challenge to theoretically calculate the free energy of protein-protein binding.Traditional computational methods that calculating binding free energy,such as Free Energy Perturbation,Thermodynamic Integration,and Umbrella Sampling,often involve computing a series of physically non-existent intermediate states,which are computationally expensive and computationally effective and may not be ideal.In this paper,we used an interaction entropy method,which recently was used to protein-protein and protein-ligand binding free energy computation successfully,to research the influence of protein internal dielectric constant for computational free energy.Interaction entropy through calculating the fluctuation of interaction energies of a series of protein-protein binding structure extracting from molecular dynamics(MD)simulation to get the loss of entropy in protein-protein binding process,others can be evaluated by standard molecular mechanics/generalized Born surface area(MMPB/GBSA)approach.At the same time,many studies have found that different dielectric constants should be used inside different residues within proteins.Through analysis the computational result of twenty realistic protein-protein interaction systems,we find the protein internal dielectric constant act a significant role in binding free energy computation,and the primary reason is charged residue(side chain)need a bigger dielectric constant. |