Measuring the RMSD of the system is too crude a metric to be useful. The RMSD of the ligand and/or binding site residues is much more sensitive and informative.

I assume you mean hydrogen bonds, not hydrogen atoms, so your results indicate any of the following are possible:

1. The initial docked pose is wrong (note that the starting number of hydrogen bonds is already quite a bit lower than the experimental reference value).

2. The reference structure from experiments is incorrect (hydrogen bonds in crystal structures can be spurious because the H atoms can't usually be resolved).

3. The force field parameters for the ligand and/or protein are not suitable. More likely the ligand is to blame than anything else.

4. The length of the simulation is inadequate to refine the structure.

you can follow this link

http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/

good luck

I edited the question .

The description is still lacking in any useful information. What are you hoping to observe in the simulation? What defines "bad" - protein unstable, some value from the literature not being reproduced, etc?

I try to validate docking results . and I started by that one that has experimental data .

I docked it and try to perform MD simulation . but number of hydrogen atoms decreased from 10 to 5 . (in experimental it was about 15) but the pose didn't change and RMSD of all system stabilize after about 0.5 ns

Measuring the RMSD of the system is too crude a metric to be useful. The RMSD of the ligand and/or binding site residues is much more sensitive and informative.

I assume you mean hydrogen bonds, not hydrogen atoms, so your results indicate any of the following are possible:

1. The initial docked pose is wrong (note that the starting number of hydrogen bonds is already quite a bit lower than the experimental reference value).

2. The reference structure from experiments is incorrect (hydrogen bonds in crystal structures can be spurious because the H atoms can't usually be resolved).

3. The force field parameters for the ligand and/or protein are not suitable. More likely the ligand is to blame than anything else.

4. The length of the simulation is inadequate to refine the structure.

Gromacs is no doubt a unique software to remove so called "back ground noises" in bioinformatics processing of data. But it has its own restrictions too. It does not have universal application. You seem to have applied censorship to edit your question. In that case you must do intensive search on the net.

Compare simulation results with differents parameters and you will see how sensitive they are.

You can use travis, if you export the trajectory to an xyz coordinate file.

The use of this program is self-explanatory, and it is very powerful.

Http://www.travis-analyzer.de/

You can type command followed by -h to get instruction, for example, "g_energy -h"

As repeatedly stated, you need to provide much more information. What we're given to go on is virtually nothing. The one thing we know about your system is that the "[...]number of hydrogen atoms decreased from 10 to 5[…]", which is basically nonsense. Please provide more detail and context, or no one will be able to help you with this.