GROMACS version:
GROMACS modification: Yes/No
I am running some replica simulations of a protein in complex with DNA . In the previous simulation the energy minimisation was done for 500 steps , is there any explanation as to why this specific number of steps? I think it has to do with the emtol or maximum force applied to the system. kindly correct if I am wrong .
Also, is the equilibration step necessary in all the simulations? can it be skipped?
Regarding the minimization steps, it does have to do with emtol. The energy minimization process involves making sure that no force on any atom is higher than the maximum force set by emtol and does this by perturbing each atom a tiny bit during each energy step. The minimization stops once no force exceeds emtol, so if your minimization was done in 500 steps, it’s either because all forces are successfully below emtol or the max number of steps to perform the minimization was set to 500.
You definitely shouldn’t skip the equilibration step, your results depend on the protein complex being equilibrated with the rest of the system. Without equilibration, you’ve essentially teleported your protein from what may effectively be a completely different system into the one you have now. Values such as the kinetic energy of the atoms on the protein complex, the pressure exerted by the surrounding solvent, and the placement of solvent molecules around the protein complex will be set up in ways you would never find in a real system.
what if I don’t carry out a separate equilibration step? I start my simulation and let the system equilibrate for a few nanoseconds and then start analyzing the data.
You’re still starting from an entirely arbitrary state with unphysical water molecules and random velocities. If you don’t properly equilibrate, the entire trajectory could be poisoned by bad forces in the first few frames. Even a stable protein could badly distort if you don’t equilibrate sensibly. Don’t skip it. Equilibration to achieve an appropriate statistical mechanical ensemble and physical properties is different from discarding the first few nanoseconds of a run during the unrestrained simulation.