GROMACS version: 2021.4
GROMACS modification: Yes/No
Here post your question : I have run my simulation for 50 ns using Gromacs 2021.4. For extending the run, i used another system with Gromacs version 2022.2. Will this affect my results somehow?
With run-of-the-mill equilibrium simulations, same .mdp settings, and two versions half a year apart - for all intents and purposes, no.
@milosz.wieczor As I have longer simulations to run, i am extending the simulation on hpc, is the noPBC.xtc file needed to be uploaded for the extending the simulation? I uploaded only the previous run.tpr and .cpt file.
In general, I recommend using the -noappend flag with gmx mdrun, which writes each single iteration to a separate set of files (numbered with .part000x.); this is safer as .xtcs can get corrupted or overwritten. If you do so, you don’t need to have the previous .xtc locally. If you skip this flag, though, I’m inclined to guess Gromacs will complain about the lack of a file to append to while restarting from a .cpt.
@milosz.wieczor I want to simulate my protein-water and protein-solvent system at different temperatures. For that do I need to change the temperature at equilibration steps or at the production run step only? In the md.mdp, I need to change the ref_temp to the desired temp. that I need to study? If it’s needed to change temperature at equilibration steps, then in the nvt.mdp do I need to change both reference (ref_temp) and Maxwell distribution (gen_temp) ? And do I need to change the ref_temp in npt.mdp too ?
You need to set the desired temperature correctly at all stages, from velocity generation onward. Equilibrating at one temperature and then attempting to simulate (production) at another negates the value of equilibration and introduces instability.
Okay, I understood now. I did not change the ref_temp in the nvt.mdp parameter file, i had just changed the gen_temp only.
Thank you @jalemkul for the timely help.