Usually md simulations start with 3d structure either from crystallography or nmr experiment. Then this structure is solvated, typically water + salt (NaCl), to mimic biological environments. Since the density of the solvent is not initially the same as the density of the standard conditions (temperature around 289 K and pressure 1 atm) and also the volume of the biomolecule, we first equilibrate the system by first restraining the heavy atoms of the protein positions close to the experimental structure, and letting the solvent and/or hydrogens to relax to the standard conditions. The equilibration protocol may include minimization, heating/cooling, restraint/constraint simulations for 500-1000 ps. Then, when the density of the solvent is close the standard conditions, the restraints or constraints are removed and the all system is relaxed in either nvt or npt ensemble for 1-2 ns depending on the size of system. Then, after that starts the production run, which is collection of data for analyzing.
How to set up the system for each section of protocol you should refer to the user manual of the corresponding md engine.
The numerical integrator time step for Namd using single step numerical integrators AND SHAKE for bonds involving hydrogens is 1-2 fs, during the production run 2fs should be stable depending on the ensemble and thermostat/barostat method you have selected.
The time step between snapshots for analysing properties is recomended around 1 ps (500 steps), this depends also on the time correlation length of the property of interest.
Thanks a lot for the answers!!
Just one more thing. What is the diference beteween equilibration and production run? In my case I am studying the impact of mutations on the structure of my protein. How is correct to use this terms?
Regards
Usually md simulations start with 3d structure either from crystallography or nmr experiment. Then this structure is solvated, typically water + salt (NaCl), to mimic biological environments. Since the density of the solvent is not initially the same as the density of the standard conditions (temperature around 289 K and pressure 1 atm) and also the volume of the biomolecule, we first equilibrate the system by first restraining the heavy atoms of the protein positions close to the experimental structure, and letting the solvent and/or hydrogens to relax to the standard conditions. The equilibration protocol may include minimization, heating/cooling, restraint/constraint simulations for 500-1000 ps. Then, when the density of the solvent is close the standard conditions, the restraints or constraints are removed and the all system is relaxed in either nvt or npt ensemble for 1-2 ns depending on the size of system. Then, after that starts the production run, which is collection of data for analyzing.
How to set up the system for each section of protocol you should refer to the user manual of the corresponding md engine.
If your simulations are not so time expensive, you can also evaluate the best time step (starting from the typical values just suggested you in previous answers) by performing several runs with different time steps. The best time step is the greater one which does not show any significant changes in statistics.
- Badges
- Science topic