I created new project, and prepared for simulation by “Gromacs FULL MD setup”.
Using this prepared structure, I ran “Simple Box Solvent Molecular Dynamics (NPT)” 2 times.
After simulation, I noticed that these 2 simulation seeds(written in log file as ld_seed) are same.
But other results (gro, trr, xtc, pdb) were different.
And using analysis tool, I checked RMSd and Bfactor. These were also different.
Then using analysis tool “Get a trajectory snapshot”, I checked structures first step and last step.
These structures were same between first and second simulation.
If seed is same, should simulation results be same?
Can I change seed for simulation?
Trajectories are different, why are snapshots same?
If seed is same, should simulation results be same?
In theory yes, but MD has a stochastic behaviour that makes trajectories diverge quickly even using the same seed. See this thread from CHARMM forum as an example:
Can I change seed for simulation?
Unfortunately this is not possible with MDWeb. I would recommend you to take a look at the new work done by BioExcel on MD usability here: http://mmb.irbbarcelona.org/biobb/
This building blocks library is much more flexible than MDWeb in terms of GROMACS parameters, for example it will allow you to change the seed.
Trajectories are different, why are snapshots same?
Actually the final snapshots of the two trajectories are different. Try to download both trajectories to your own computer and visualize them using a software like VMD. There must have been some issues with the “Get a trajectory snapshot” tool.
Thank you very much for your interest in our tool!