From: 翁林岽 (wenglindong_at_yahoo.com.cn)
Date: Fri Jun 07 2013 - 21:30:12 CDT
Thanks for your reply. I use langevin dynamics to control the temperature and I run NPT ensemble. Actually, I did restrain the lipid for several ns and release it. But the shift always happens. I may just leave it as it is and try recentering the trajectory for visualization as you said.
发件人： Aron Broom <broomsday_at_gmail.com>
收件人： 翁林岽 <wenglindong_at_yahoo.com.cn>
抄送： "namd-l_at_ks.uiuc.edu" <namd-l_at_ks.uiuc.edu>
发送日期： 2013年6月7日, 星期五, 9:56 下午
主题: Re: namd-l: Shift (or offset) of the lipid bilayer along Z axis during equilibration
it will have zero impact on your simulation, it's just a visualization issue. There are a number of emails in the thread about re-centering a trajectory for visualization.
As to why it happens, certainly part of it is just diffusion, but why does it continually drift in the same direction? Maybe it isn't fully equilibrated/minimized and some clash gives it some initial momentum. What thermostat are you using? Also, you can initially restrain the lipids and equilibrate the water around it then release that constraint. This may help if you are very concerned with the drift, but the position of the periodic boundary is arbitrary.
On Fri, Jun 7, 2013 at 9:36 PM, 翁林岽 <wenglindong_at_yahoo.com.cn> wrote:
>When I equilibrate a DPPC/water system, the center of the lipid bilayer always shifts by several Ås along the Z axis. Since the center of the simulation box is kept at (0, 0, 0), the consequence is that 1) the water layer on one side is squeezed into the neighbor periodical box and gradually shrinks and 2) the water layer on the other side correspondingly becomes thicker due to the periodical conditions I think. As the equilibration goes on, the shrinking water layer becomes a very thin layer. Does anyone observe the same problem in your simulation?
>I am afraid that the lipid bilayer may eventually be seperated by the periodical boundary so that the original water<--->lipid bilayer<--->water complex will become a lipid monolayer<--->water<--->lipid monolayer one in the simulation box. Probably, with periodical conditions, such changes will not negatively affect the simulation accuracy. But can anyone tell me why there is such shift? Are these changes normal or acceptable? Are there any ways we can avoid such shifts and make the simulation box look normal?
>Any suggestions would be really appreciated.
Aron Broom M.Sc
Department of Chemistry
University of Waterloo
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:17 CST