[gmx-users] Significant smaller area per lipid in bilayer simulations using CHARMM force field

[凌未风]

Hello:


I am simulating a stearol sphingomyelin bilayer with 81 lipids in each leaflet using the CHARMM force field. I used the CHARMM-GUI to generate the initial conformation, as well as the parameters for mdp file. I conducted the simulation by gmx5.0.4. 


I found a significant smaller area per lipid (APL) in my simulations. The APL kept on decreasing, and started to equilibriated after 400 ns. The expected value is ~55.4 A^2, but in my simulation, the value was ~50.8 A^2 after equilibrated. For me this difference is significant. I looked up in the mailing list, and found someone said that it is because of the ions using or it is because of the motion of the two leaflets. So I tried simulations without ions, and with NaCl/KCl. I also tried to remove the centers of mass of two leaflets separately in my simulations. But all of the simulations generated similar APL. 


I am really appreciate for any possible suggestions. This looks really strange for me, as I am using the "standard parameters" (i.e., the parameters generated by CHARMM-GUI) for CHARMM force filed in gromacs, and the difference should not be so larger, if there are any differences....


Thanks again for any help you may provide. 


Cheers, 
RXG
University of Calgary