[gmx-users] pressure and potential energy

[Justin Lemkul]

On 3/24/16 10:00 AM, abhishek khetan wrote:
> No I was never warned about anything like that.
> The output I get after grompp for the NVT continuation run is:
> -------------------------------------------------------------------------------------------------
> Command line:
>    gmx_mpi grompp -f nvt.mdp -c dme_nvt.gro -t dme_nvt.cpt -o nvt.tpr
>
> Back Off! I just backed up mdout.mdp to ./#mdout.mdp.1#
> Setting the LD random seed to 451659651
> Generated 330891 of the 330891 non-bonded parameter combinations
> Generating 1-4 interactions: fudge = 0.5
> Generated 330891 of the 330891 1-4 parameter combinations
> Excluding 3 bonded neighbours molecule type 'DME'
> Removing all charge groups because cutoff-scheme=Verlet
> Analysing residue names:
> There are:    64      Other residues
> Analysing residues not classified as Protein/DNA/RNA/Water and splitting
> into groups...
> Number of degrees of freedom in T-Coupling group System is 3069.00
> Determining Verlet buffer for a tolerance of 0.005 kJ/mol/ps at 300 K
> Calculated rlist for 1x1 atom pair-list as 1.100 nm, buffer size 0.000 nm
> Set rlist, assuming 4x4 atom pair-list, to 1.100 nm, buffer size 0.000 nm
> Note that mdrun will redetermine rlist based on the actual pair-list setup
> Reading Coordinates, Velocities and Box size from old trajectory
> Will read whole trajectory
> Last frame         -1 time 1000.000
> Using frame at t = 1000 ps
> Starting time for run is 0 ps
> Calculating fourier grid dimensions for X Y Z
> Using a fourier grid of 20x20x20, spacing 0.111 0.111 0.111
> Estimate for the relative computational load of the PME mesh part: 0.22
> This run will generate roughly 6 Mb of data
> -------------------------------------------------------------------------------------------------
>
> And the output when using the NVT data in the NPT grompp is:
>
> -------------------------------------------------------------------------------------------------
> Command line:
>    gmx_mpi grompp -f npt.mdp -c dme_nvt.gro -t dme_nvt.cpt -o npt.tpr
>
> Back Off! I just backed up mdout.mdp to ./#mdout.mdp.1#
> Setting the LD random seed to 846700063
> Generated 330891 of the 330891 non-bonded parameter combinations
> Generating 1-4 interactions: fudge = 0.5
> Generated 330891 of the 330891 1-4 parameter combinations
> Excluding 3 bonded neighbours molecule type 'DME'
> Removing all charge groups because cutoff-scheme=Verlet
> Analysing residue names:
> There are:    64      Other residues
> Analysing residues not classified as Protein/DNA/RNA/Water and splitting
> into groups...
> Number of degrees of freedom in T-Coupling group System is 3069.00
> Determining Verlet buffer for a tolerance of 0.005 kJ/mol/ps at 300 K
> Calculated rlist for 1x1 atom pair-list as 1.100 nm, buffer size 0.000 nm
> Set rlist, assuming 4x4 atom pair-list, to 1.100 nm, buffer size 0.000 nm
> Note that mdrun will redetermine rlist based on the actual pair-list setup
> Reading Coordinates, Velocities and Box size from old trajectory
> Will read whole trajectory
> Last frame         -1 time 1000.000
> Using frame at t = 1000 ps
> Starting time for run is 0 ps
> Calculating fourier grid dimensions for X Y Z
> Using a fourier grid of 20x20x20, spacing 0.111 0.111 0.111
> Estimate for the relative computational load of the PME mesh part: 0.22
> This run will generate roughly 6 Mb of data
> -------------------------------------------------------------------------------------------------
>
> Both times when I use mdrun, there are no special lines saying that the
> time step is large or whatever. yeah, you're spot on, I'm not using any
> constraints and/or restraints, nor any refcoord_scaling. Do you think that
> is where the problem originates? I discussed this at length on the forum (
> http://permalink.gmane.org/gmane.science.biology.gromacs.user/83291) and
> from what i learned, i really had no reason to use constraints/restraints
> for my nonaqueous solvent system.
>

Let me clarify my remark regarding constraints and restraints.  Your previous 
post mentioned a concern that your molecule would "disintegrate" somehow.  I 
stated that since you have bonds defined, that can't happen.  You don't need to 
use constraints or restraints to prevent "disintegration" of your molecule. 
That does *not* mean that constraints aren't appropriate.  For any time step > 
0.5 fs or so, you should probably be at least constraining bonds to hydrogens. 
Without constraints, you can get instability, which usually manifests itself in 
a crash.

> Im trying to increase the fourier grid density and then rerun the
> simulations. It would be very kind if you can educate me a bit here on
> getting to procedure to equilibration and production MD runs. Thanks very
> much in advance.
>

Your system is extremely small, and you're using coupling algorithms that allow 
for very large fluctuations.  I don't think this is a wise combination.  Build a 
larger box and you will get more sensible properties, especially those related 
to the pressure.  Your simulation cell is simply too tiny to reliably be using 
the algorithms you're applying.

-Justin

> Hi,
>>
>> Did grompp warn you that your time step was too large for a system without
>> constraints? What did you do about that?
>>
>> Mark
>>
>> On Thu, 24 Mar 2016 00:28 abhishek khetan <askhetan at gmail.com> wrote:
>>
>
>
>

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================