In the process of obtaining the nonbond potential of the Coarse-grained(CG) model, I need to simulate the CG model mapped from the conformation of all atom(AA). AA simulation process is equilibrated in a constant pressure− temperature (NPT) at 500K and atmospheric pressure(1atm), then quenched to 300K, and finally NPT at 300K and 1atm. It was originally intended that CG process should be simulated according to this process. However, according to the simulated iterative Boltzmann inversion for coarse-grained of polyurea published by Agrawal et al., I found that the author mentioned that the density and temperature should be kept constant during CG simulation. However, when I tried to apply this method to a small system polyethylene at that time, the bond length, bond angle, dihedral angle and other information was obtained smoothly, but when I used the same method to fit the nonbond potential energy, I found that the density obtained from CG file mapped directly by AA model was far less than that obtained from all atom model. According to what I know, the density of the system is only fixed by the pressure-control command. How can I keep the system density unchanged in CG simulation?