On Thu, Feb 25, 2021 at 11:17 PM Pragati Sharma <
praga...@gmail.com> wrote:
>
> Thanks Christoph, I got a python script here in group chat and after modifying it, I can easily make mapp.xml for 500 polymer.
>
> I am running NPT simulation of the polymer with the obtained tabulated potentials. However I have a few queries:
>
> 1. For obtaining potentials from IBI, NVT simulations were performed and It reached good convergence (<0.003) but my pressure values were negative. Later I added pressure correction to the settings file. With NVT, values did not change much and although pressure becomes positive, It didn't reach one (1 bar is required pressure). But with NPT and pressure correction, pressure values reached to 1, however the convergence decreased. Should one perform NPT or NVT for potentials with pressure correction.
I am not 100% sure IBI in NPT is really well researched. Pressure
correction is bit magically especially when you have multiple bead
types, in principle one would need to match the partial pressure, but
as no MD code has partial pressures implemented, VOTCA cannot do it
either. The only advice I can give for the pressure correction is to
make sure your trajectories are long enough, the error bars/
fluctuations on the pressure are usually huge.
>
> 2. After obtaining potentials, when I run CG NPT simulation. The command is gmx mdrun -s topol.tpr -tableb table_a0.xvg table_a1.xvg table_a2.xvg table_b0.xvg table_b1.xvg table_b2.xvg -c confout.gro
> While we provide bonded tables using -tableb, Is there an option to give non bonded potentials in the command itself. Does this command line will take all non-bonded potentials by itself since we have put energygroup in mdp file.
Yes, I believe you don't need an exxtra option for the non-bonded
tables (at least in gromacs-2019).
>
> 3. If I want to use GPU for NPT simulations, It gives error that you can't use GPU with user defined tables. You have to give the PME or cut-off. Since we have to use tabulated potentials, does that mean we can't use GPU for such simulations?
That is unfortunately right, if you have any question on this issue,
ask on the gromacs user list (
gmx-...@gromacs.org) Alternatively you
could use LAMMPS or hoomd-blue for your md simulation.
Christoph
> To view this discussion on the web visit
https://groups.google.com/d/msgid/votca/CAMpF9Q16J4EBwuN-WTTqNRC%2B_y%2BCTK-cwiu-nuHrjVM%3DW%3DXMKA%40mail.gmail.com.