[gmx-users] pressure based replica exchange

[Michael Shirts]

Pressure replica exchange is not in currently. It will be added
eventually (it's not that hard, just takes time, and hasn't been a
priority.).

Note that pressure temperature control with berendsen pressure control
will be 100% wrong.  For pressure replica exchange to work, the
barostat has to give the correct volume fluctuations, and berendsen
does not give correct volume fluctuations.

Of course, that's irrelevant since pressure replica exchange isn't in yet.

The exchange criteria is always of the form min(1,exp(-X), where X =
\beta(Delta (beta U) + Delta (beta P V)).  Where P, beta, and U can
vary with state.

What is in is temperature replica exchange with NPT, where beta
changes, but P stays the same.

On Thu, Dec 12, 2013 at 1:40 AM, Jianguo Li <ljggmx at yahoo.com.sg> wrote:
> Thanks very much. As you suggested, making T with small increments such as 300.0, 300.0001, 300.0002 ... leads to complaints :-)
>
> I am trying to do replica exchange at different surface tensions for my membrane system.
> To achieve different surface tensions for different replica, I only changed Pxy, but keep Pz the same for all replica.
> pcoupl          = berendsen
> pcoupltype      = semi-isotropic
> tau_p           = 0.5 0.5
> ref_p           = 1.0 1.0  ; Pxy and Pz
> compressibility = 4.5e-5  4.5e-5
>
> I noticed that in the manual, the eq (3.141) shows the exchange probability. Can you explain what does the P1 and P2 represent? Pxy or Pz?
>
> Also, the manual mentioned about using Hamiltonian and temperature replica exchange simultaneously (eq 3.143). But there is no pressure term in the acceptance criteria (eq 3.143). I am wondering if pressure+temperature+Hamiltonian replica exchange has been implemented in Gromacs. If yes, what is the general equation of acceptance criteria?
>
> Cheers,
> Jianguo
>
>
>
>
>
>
>
> On Tuesday, 10 December 2013, 17:49, Mark Abraham <mark.j.abraham at gmail.com> wrote:
>
>
>
>
>
>
> On Tue, Dec 10, 2013 at 1:47 PM, Jianguo Li <ljggmx at yahoo.com.sg> wrote:
>
> Dear All,
>>
>>Is it possible to do pressure based replica exchange simulations in gromacs? Basically I want to do replica exchange simulations for my membrane system at different surface tensions. If I just set different pressures in the mpd file, then the mdrun will complain "systems are all the same, there is nothing to exchange". Is there any other parameters I need to set?
>>
>
> Yes, the implementation is a bit broken, and requires that one of ref-t or lambda varies before it even considers the pressure. If you want T constant and P as the replica control variable, you can can only do that by fooling the implementation by having T vary in tiny increments.
>
> Mark
>
>
>>Thanks a lot!
>>
>>Jianguo
>>--
>>Gromacs Users mailing list
>>
>>* Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>>
>>* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>>* For (un)subscribe requests visit
>>https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.
>>
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.