10 Common Molecular Dynamics Simulation Mistakes (And How to Avoid Every One)

10 Common Molecular Dynamics Simulation Mistakes (And How to Avoid Every One)

Most failed MD simulations don’t fail with an obvious error message. They run to completion, produce a plausible-looking trajectory, and return results you might trust — until someone more experienced looks at your methods. These are the ten mistakes that appear most often in troubleshooting requests, reviewer comments, and retracted computational papers.

Severity: Critical — invalidates results High — significantly degrades accuracy Medium — produces artifacts or misleading output
01
Critical
Simulation box too small — protein interacts with its own periodic image

Periodic boundary conditions require the simulation box to be large enough that the protein never comes within interaction distance of its own periodic image. When the minimum distance between the protein and the box edge is smaller than the non-bonded cutoff (typically 1.0–1.2 nm), the protein effectively interacts with itself — an unphysical artifact that corrupts electrostatics, introduces false forces, and produces meaningless dynamics.

This is especially problematic for elongated proteins, flexible proteins that sample extended conformations during the simulation, and any system that undergoes significant conformational change. A box that was large enough at the start may become too small after 100 ns if the protein extends.

The fix
Set the minimum distance from protein to box edge to at least 1.0 nm using gmx editconf -d 1.0. For flexible proteins or long simulations, use 1.2 nm. Monitor the minimum image distance throughout the simulation with gmx mindist -pi — if it drops below your non-bonded cutoff at any point, the simulation is compromised.
02
Critical
Not equilibrating long enough before production

The most common source of non-reproducible MD results. A system that hasn’t equilibrated is not representative of physiological conditions — temperature, pressure, and density are still drifting, and the protein may be in a metastable state far from its true equilibrium structure. Running production MD from a poorly equilibrated system means collecting data from an artifact rather than a real simulation.

The telltale sign: RMSD that rises steadily throughout what you’re calling the “production” phase. If RMSD hasn’t plateaued, the system hasn’t equilibrated — regardless of how long the equilibration phases nominally ran.

The fix
Run NVT and NPT equilibration long enough for temperature, pressure, and density to plateau — typically 100–500 ps for each phase. Then check RMSD at the start of production and discard any portion that shows systematic drift. For large or flexible proteins, equilibration can require several nanoseconds. Always plot temperature, pressure, density, and potential energy vs time before calling a system equilibrated.
03
Critical
Non-zero system charge left uncorrected

Particle Mesh Ewald (PME) — the standard long-range electrostatics method — requires the simulation box to be charge-neutral. A system with a non-zero net charge generates a divergent electrostatic energy under PBC that is handled by adding a uniform background charge, introducing a systematic error in the electrostatic calculation. The simulation runs and produces no error message, but all electrostatic energies are shifted by an unquantified constant.

This affects binding free energies, protein-ligand electrostatic interactions, and ion binding. It is especially problematic when the charged species is the ligand — a charged ligand in an uncorrected system will have spurious electrostatic interactions with the background charge.

The fix
Always run gmx genion -neutral to add counterions and neutralize the system before simulation. Verify with grep "qtot" topol.top | tail -1 — the total charge must be zero. For ligands with a net charge, confirm the formal charge is correctly assigned in the ligand parameterization step and that the corresponding counterion is added.
04
Critical
Mismatched force field and water model

Every biomolecular force field is parameterized against a specific water model. The protein parameters and water parameters are interdependent — they were fitted together to reproduce experimental observables including protein-water interactions, hydration free energies, and structural properties. Using a mismatched combination (CHARMM36 protein with TIP4P water, for example) introduces systematic errors in protein-water interactions that propagate into every property you measure.

The fix
Use the recommended water model for your force field: AMBER force fields → TIP3P; CHARMM36/36m → CHARMM-modified TIP3P (tip3p in GROMACS); GROMOS → SPC or SPC/E; OPLS-AA → TIP4P. Never mix these. The correct water model for each force field is specified in the force field documentation and in the pdb2gmx prompt when you choose your force field.
05
Critical
Wrong or incompatible ligand force field parameters

Protein force fields cover the 20 amino acids. They say nothing about arbitrary drug molecules. Using protein-derived atom types for a ligand, mixing GAFF2 ligand parameters with CHARMM36 protein parameters, or using poorly validated ACPYPE output without checking the atom type assignments all produce physically wrong interactions between the ligand and protein. The simulation runs, but the binding affinity, ligand dynamics, and protein-ligand contacts are all artifacts.

The fix
Use a force field pair designed to work together: AMBER protein force fields + GAFF2 ligand parameters, or CHARMM36 protein + CGenFF ligand parameters. After parameterization, inspect the ACPYPE or CGenFF output: check atom type assignments for all unusual functional groups, verify the net charge matches your expected formal charge, and inspect partial charges on key atoms. For CGenFF, check penalty scores — values above 50 warrant manual review.
06
High
Timestep too large — simulation instability

The standard 2 femtosecond timestep is safe when X-H bonds are constrained with LINCS or SHAKE. Without these constraints, the fastest atomic motions (X-H bond vibrations, ~10 fs period) require a timestep of 1 fs or smaller. Using 2 fs without hydrogen constraints can cause numerical instability that either crashes the simulation with a Stepsize too small error or, more insidiously, produces a trajectory that looks normal but has accumulated integration errors.

The opposite problem also exists: some researchers use very long timesteps (4–5 fs with virtual sites) without properly validating that structural properties are preserved. A timestep that’s too long for the system’s fastest degrees of freedom produces systematic errors that don’t announce themselves.

The fix
Always use constraints = h-bonds in your MDP file when running at 2 fs. If you want to use a 4 fs timestep for performance, also constrain heavy-atom bonds (constraints = all-bonds) and consider using virtual sites for aromatic rings. After any change to timestep, run a short NVT simulation and check that total energy is conserved and temperature is stable.
07
High
Using Berendsen thermostat or barostat for production MD

The Berendsen thermostat and barostat equilibrate temperature and pressure quickly and stably — which is why they’re appropriate for equilibration. But they do not produce a correct thermodynamic ensemble: they suppress the natural fluctuations that define the canonical (NVT) or isothermal-isobaric (NPT) ensemble. Using Berendsen for production MD gives the wrong heat capacity, wrong compressibility, wrong free energies, and potentially wrong conformational populations. The simulation looks fine; the thermodynamics are wrong.

The fix
Use Berendsen only for equilibration. Switch to Nosé-Hoover thermostat and Parrinello-Rahman barostat for all production runs. In your production MDP: tcoupl = Nose-Hoover and pcoupl = Parrinello-Rahman. This is one of the most common errors in published MD papers and increasingly flagged by reviewers.
08
High
Analyzing or visualizing the raw uncentered trajectory

Periodic boundary conditions cause atoms to appear to teleport across the box boundary in the raw trajectory. This is not a physical artifact — it’s a coordinate wrapping convention. But it causes real problems when you run analysis on the raw trajectory: RMSD calculations give wrong values when atoms jump across boundaries, hydrogen bond distances appear impossibly large, and visualization shows a fragmented protein. The simulation is fine; the analysis is wrong.

The fix
Always process with gmx trjconv -pbc whole followed by gmx trjconv -center -pbc mol -ur compact before any analysis or visualization. Make this the first step after every simulation — before RMSD, before RMSF, before hydrogen bond analysis, before loading into VMD or PyMOL.
09
Medium
Not checking energy conservation during NVT runs

In NVT simulations with no thermostat coupling (or very weak coupling), total energy should be approximately conserved over time. Systematic energy drift — the total energy rising or falling steadily throughout the simulation — indicates a problem with the integrator, timestep, or force field parameters. It’s a diagnostic signal that most researchers never check, leaving subtle simulation problems undetected.

Energy drift is measured in kJ/mol/atom/ns. A drift rate below 0.01 kJ/mol/atom/ns is excellent. Above 0.1 suggests a problem worth investigating.

The fix
Run a short NVT simulation with tcoupl = no (no thermostat) and check total energy conservation with echo "Total-Energy" | gmx energy -f nve.edr -o nve_energy.xvg. If energy is not conserved, reduce the timestep, increase the LINCS order, or check for poorly parameterized atoms in your system.
10
Medium
Interpreting a short simulation as representative — convergence not checked

A 100 ns simulation that samples a local energy minimum looks identical to a 100 ns simulation that has genuinely explored the relevant conformational space — unless you check convergence. Conclusions drawn from unconverged simulations are simply observations of where the system happened to be, not statements about equilibrium properties. This is the most common way MD studies overstate their confidence in results.

Common convergence checks include: comparing RMSF profiles from the first and second halves of the trajectory (they should be similar), checking whether RMSD has genuinely plateaued or is still slowly rising, and running multiple independent replicas with different initial velocities to confirm key observations are reproducible.

The fix
Before reporting any result from an MD trajectory, split it in half and compare RMSF, key distances, and secondary structure populations between the two halves. Run at least two independent replicas. If the two halves or two replicas disagree on key observables, run longer or report the disagreement explicitly. State the convergence check in your methods section.

Quick reference: all 10 mistakes at a glance

MistakeSeverityCore fix
Box too smallCritical≥ 1.0 nm edge distance; monitor with gmx mindist -pi
Insufficient equilibrationCriticalCheck RMSD plateau; discard non-equilibrated frames
Non-zero system chargeCriticalgmx genion -neutral; verify qtot = 0 in topology
Mismatched FF + water modelCriticalUse recommended pairs: AMBER+TIP3P, CHARMM36+TIP3P
Wrong ligand parametersCriticalMatch GAFF2 to AMBER or CGenFF to CHARMM36
Timestep too largeHighUse constraints = h-bonds with 2 fs timestep
Berendsen in productionHighSwitch to Nosé-Hoover + Parrinello-Rahman for production
Raw trajectory analysisHighAlways run gmx trjconv -pbc whole then -center first
No energy conservation checkMediumRun short NVE test; drift < 0.01 kJ/mol/atom/ns
No convergence checkMediumCompare first and second half RMSF; run 2+ replicas
The pattern behind most of these mistakes
Seven of the ten mistakes above produce plausible-looking output with no error message. GROMACS will run to completion and return a trajectory regardless of whether the box is too small, the system is charged, or you’re using the wrong thermostat. The only defense is a systematic pre-simulation checklist and post-simulation convergence check — not trusting that a finished run equals a correct one.

The one-paragraph version

MD simulation mistakes are dangerous precisely because most of them are silent — the simulation runs, produces numbers, and returns results that look reasonable. The five critical mistakes (box size, equilibration, system charge, force field mismatch, ligand parameters) can each independently invalidate an entire study. Build a preparation checklist, run convergence checks on every trajectory, and verify the basics before trusting any result. The researchers who catch these mistakes in their own work before publication are the ones whose reviewers don’t catch them afterwards.

Last updated on

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *