GROMACS version: 2019
How to validate the MD simulation results obtained by gromacs? My simulated systems are 5 protein-ligand complex.
Unfortunately, I can’t access to experimental data for comparison and validation. I’m beginner in gromacs. Is doing duplicate or triplicate MD simulation good answer or are there some special analysis tools for this issue? Please guide me about that.
Merry,
Best
How long is a piece of string?
The answer depends on what exactly you’re wanting to validate. Is the force field you are using suitable for the nature of the system you are studying? Do the physical/physiological process you are studying occur on the timescales you are able to simulate with your available resources?
You mention protein-ligand complexes. Is the environment in your simulation box (e.g. solvent, lipids, ions) close to the environment you would find that protein in naturally? Is your cell size large enough so that the protein is not forming unphysical self-interactions with its periodic images? Does the RMSD of the protein backbone reach equilibrium during the simulation? The latter two you can check with the bundled gmx analysis tools.
Are the force fields you’ve chosen for the protein compatible with the ligand force field? Does your process involve water? What water model are you using? Is this model sufficient to model that process?
Did the proteins come cocrystallised with the ligand? If not, how did they end up in/on the protein? If you docked the ligand, do you have experimental evidence suggesting this ligand belongs in/on this protein and in/on the place you have chosen to dock it?
Where are your ligand force field parameters from? Does your GROMACS-minimised ligand accurately reproduce any available crystal structures of the ligand? How about compared to a reasonably accurate (e.g. M06-2X/6-311G(d)) QM-optimised structure of the ligand? Does gmx nmeig reproduce the normal modes accurately compared to an IR spectrum of the ligand? Do the partial charges on the atoms make sense to your chemical intuition? If it’s unclear you can reference with MK charges. Can you reproduce in GROMACS the relative energies of the ligand conformers compared to QM?
Running replicates is always a good idea, but if your system has other issues this will not fix them. Garbage in = garbage out.