[gmx-users] TFE query

[Mark Abraham]

Dilraj Lama wrote:
> Hello gmx users,
>                 I have a query and would be greatfull if I could get some
> suggestions on the subject.
>                            I constructed a "tfe.itp" file with the
> parameters mentioned by Fioroni et. al., J.
> Phys. Chem. B 2000, 104, 12347-12354. I then
> generated a "tfe.gro" file using InisghtII
> molecular modeling package followed  by
> genconf.
> 
> I ran a "energy minimization" fand then a 2ns NPT ensemble Molecular
> Dynamics simulation on the system.

Why are you not equilibrating before running a production simulation? 
See http://wiki.gromacs.org/index.php/Steps_to_Perform_a_Simulation

> I did not specify "Lincs algorithm" in the parameter file to constrian the
> bonds, rather I used [constraints] for bond distance in the "tfe.itp" file
> itself.
> 
> I tried using tau_p as 1ps to 3 ps, but in doing so, my simulation crashed
> after a certain period with the error message as "high pressure scaling,
> and the failure of the constraint algorithm (either lincs or shake)".

See above.

> So I used tau_p as 4 ps which is the value used for simulation in the
> paper mentioned above.
> 
> The isothermal compressibility for liquid TFE at 298K is 1.22e-4
> The density of TFE at 298K is 1383 Kg/m3.
> 
> These values are mentioned in the paper.
> 
> I then calculated the Properties "Temperature, pressure and density" for
> the simulation using "g_energy".The average vaules for the properties are
> as:
> 
> Temperature: 300.67
> Density: 1402.95
> Pressure: -2.59
> 
> Is the density value reasonable???

No, because of the above.

> Why is the pressure value so different?????

For a 2ns simulation starting from generated velocities this is 
surprisingly close to ref_p. See the above, and also 
http://wiki.gromacs.org/index.php/Pressure_Coupling for a partial 
discussion, and the GROMACS manual for a fuller one.

> I would be very greatfull If I could get some suggestions on the
> parameters and on the values of the properties mentioned.

It sounds like some doing some more background reading and doing some 
tutorial material would be helpful to you.

Mark