[gmx-users] Disabling the pair-list update in REMD

[Mark Abraham]

On Thu, Apr 16, 2015 at 5:18 PM, Mateusz Marianski <
marianski at fhi-berlin.mpg.de> wrote:

> Hi all,
>
> I am trying to circumvent  the update of the list of neighbor atoms for
> which the long-range interactions in Cutoff scheme are evaluated. In
> essence, I want to maintain the initial neighbor-list for long-range
> interaction for whole simulations. In a simple MD run setting the nstlist
> parameter to 0 will prohibit the update of the list.
>
> Nevertheless, when I try remd, it does not work anymore, since - as it
> states in the manual:
> "The velocities are scaled and neighbor searching is performed after every
> exchange."
>
> Is there any way to disable the neighbor-search performed during the
> exchange of replicas?
>

No. GROMACS exchanges the coordinates (unlike various other MD packages, I
understand), and does not exchange the neighbour-search data structures,
which thus requires the latter to be rebuilt. One could turn that off, thus
using the neighbour-search structure originally built for whichever replica
was in the destination ensemble, but that sounds pretty crazy (e.g. solvent
diffusing differently in every replica). But maybe you have that sorted,
somehow? It would not be all that hard to rework to swap the ensemble data
instead of the coordinates, but getting that right for all possible
algorithm and integrator combinations doesn't sound like fun...

Mark


> PS. Yes, I am aware of all potential flaws/problems with not updating the
> neighbor-list in simulations. It's (hopefully) under control :).
>
> Best,
> Mateusz
> --
> 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.
>