[gmx-users] Need help on NVT equilibration for MD simulation of Protein-RNA complex

Justin Lemkul jalemkul at vt.edu
Fri Sep 8 13:45:52 CEST 2017

On 9/8/17 7:10 AM, Saravanan Parameswaran wrote:
> ​​
> ​Dear Simulators,,
> I am new to MD simulations. I am still learning Gromacs from tutorials and
> documentation.
> I want to run MD simulation of protein (1700 amino acids) with ssRNA
> (around 80 nucleotides) and another dsRNA/dsDNA (around 20-40 nucleotides).
> I am trying to understand how protein of interest interacts with
> dsDNA/dsRNA/ssDNA/ssRNA. what will be the difference in interactions when
> it interacts with different dsDNA/dsRNA/ssDNA/ssRNA and mutational studies
> on both protein and dsDNA/dsRNA/ssDNA/ssRNA.
> And I am stuck with parameters in ‘initial phase’ NVT equilibration using
> Langevin bath which I have found from literature.
> I am struggling to find the exact parameters for the protein-DNA/RNA system
> to run the three steps of the initial phase: (a) 10ps with increasing
> temperature from 0 to 100 K keeping protein & RNA fixed, (b) 25ps with
> temperature was further increased upto 200 K keeping backbone atoms of
> protein & RNA fixed and (c) 25ps to reach 298 K. Then NPT equilibration for
> 100 ps at 298 K
> After this initial phase, NPT equilibration for 40-50ns followed by
> production run in microseconds.
> I am not sure how to set parameters for MDP files for the three steps of
> the initial phase and NPT equilibration . I have googled it for few days,
> but not of much help. I end up with mdp files for free energy analysis only.
> Why the NVT equilbration or initial phase is done with three steps in
> Langevin bath?

There are many ways to prepare simulation systems.  Some people believe 
in slow warming of a system, others don't.

> Why long runs for NPT equilibration?

You have to obtain a stable thermodynamic ensemble; I don't know why 
those authors chose such a time frame (tip: ask them, that's what 
corresponding author contact information is for) but there had to be a 

> Is it really necessary to do production run in microseconds?

Depends on the questions being asked in the simulation.  You have to 
simulate on a time scale sufficient to answer whatever biological 
questions are of interest.

> I request you all to provide me the ‘mdp’ files for NVT and NPT
> equilibration as well as MD production to run the molecular
> dynamics simulations of protein/RNA complex.

If you want to replicate exactly what the authors did, you need to apply 
what is called "simulated annealing."  It is described in the manual 
along with an example.  Custom position restraint files are generated 
with gmx genrestr.  You will need different treatment of restraints in 
each phase described above.



Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129


More information about the gromacs.org_gmx-users mailing list