How to determine the number of water molecules in a box? gmx insert-molecules or gmx solvate?

User discussions
1

GROMACS version: 2025.3
GROMACS modification: Yes/No
Here post your question

Hello,

I am planning to simulate the self-assembly of surfactants in water. I created a 40 x 40 x 40 nm box and added some surfactants into it. However, I don’t know how to fill the remaining room with water. Using the gmx solvate -cp no_solvant.gro -cs standard_water.gro -o initial_system.gro command, only about 60,000 water beads were added, and a Density = 314.045 (g/l) was obtained.

When calculating how many water beads can be added by considering 64,000 nm³/0.12 nm³ (of course some room has been occupied by surfactants), about 500,000 beads should be added. gmx insert-molecules -f no_solvent.gro -ci standard_water.gro -nmol 473000 -box 40 40 40 -o with_water.gro -try 500 command can indeed add this amount of water, but this many lead to a very high density and pressure.

So how many water should be added?

Thanks in advance.

2

How much water you needs depends on the concentration of surfactants you are after. You don’t write what you want.

When you have randomly placed surfactacts there is very little space to add water molecules without overlap. You will have to start with a larger box that what you finally want to be able to insert sufficient water and then scale down the box slowly in the NPT ensemble.

3

Thanks for the help. I inserted 6000 surfactants (each consisting of 9 beads) and 3000 counterions (each consisting of 2 beads) into the 40 x 40 x 40 nm box, which should be a relatively high concentration.

Will the box size change upon NPT optimization?

Meanwhile, I created the standard_water.gro file myself (as below), should I download it from some official website?

Standard water
1
1W W 1 0.000 0.000 0.000
1.000 1.000 1.000

4

The box size will shrink during NPT.

Using a single water molecule for solvation is the worst choice. Just use the standard spc216.gro.

5

Thanks. Sorry for the insufficient explanation. I want to run coarse-grained simulation with the MARTINI force field, and all-atom model may no be suitable.

6

The same things apply there. Solvate using a box of equilibrated solvent.

7

I think first filling with solvent and then trying gmx insert-molecules with the -replace option to stick in your surfactants might work here.

8

Appreciate it. I created a box of water, made it energy-minimized and NPT-optimized, and then used it for gmx-solvate. Water was indeed added in a quite big amount.

9

Thanks for the help. I used a energy-minimized and NPT-optimized water box to solvate the system. I will try the -replace option if this solvated system is hard to converge.