[gmx-users] Warnings related to Berendsen pressure coupling and use the refcoord_scaling option

[Mark Abraham]

On 7/06/2012 8:33 PM, PAVAN PAYGHAN wrote:
>
>
>  Dear Gromacs Users,
>
>  I am using gromacs version 4.5.5.and running my jobs on single 
> node with 8 cores. My system contains about 425000 atoms (protein + 
> Lipid +SOL). I have successfully reached up to Energy minimization 
> step.As per the suggestion by Dear Mark, I am starting my NPT 
> equilibration with Berendsen and further with Parinello-Rahman to get 
> the right ensemble.But when I am trying for NPT equilibration with 
> Berendsen, getting following Warnings.
>
> WARNING 1 [file npt.mdp]: Using Berendsen pressure coupling 
> invalidates the true ensemble for the thermostat
>
> WARNING 2 [file npt.mdp]: You are using pressure coupling with 
> absolute position restraints, this will give artifacts. Use the 
> refcoord_scaling option.
>
> And about 10 such LINCS warnings while running the job as shown below:
>
> Step 39, time 0.078 (ps)LINCS WARNING
>
> relative constraint deviation after LINCS:
>
> rms 0.000029, max 0.000473 (between atoms 890 and 891)
> bonds that rotated more than 30 degrees:
>  atom 1 atom 2  angle  previous, current, constraint length
>   12050  12049   31.2    0.1000   0.1000      0.1000
>   11721  11720   46.0    0.1000   0.1000      0.1000
>    5496   5495   36.8    0.1000   0.1000      0.1000
>
> With --maxwarn option I am able to run it, and output density and 
> pressure seems perfectly alright.
>
> Based on the same I want to know:-
>
>  1. How does pressure coupling with Berendsen invalidates the true
>     ensemble?
>

See manual sections for T and P coupling for introductory discussion.

> At least for initial fixing of density and pressure?
>
> Whether to bother for above mentioned warning or ignore it?
>

Since you're going to do more equilibration after this, you be the judge.

>  2. Pressure coupling with absolute position restraints warning how it
>     will lead to artifacts?
>
> How one can use refcoord_scaling option in this situation?
>

See manual.

> Whether to bother for above mentioned warning or ignore it?
>

Position restraints and NPT is messy. Choose your poison.

>  3. Using Berendsen with semiisotropic couple type is wrong method or
>     no such problem?
>
> Please suggest me whether to move ahead with ignorance to the warnings 
> or to change some parameters in mdp file?
>

That's your judgement to make. How does your preparation protocol 
compare to the ones you have read about in the recent literature?

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20120607/bb5024ac/attachment.html>