Guide · Simulation
Molecular dynamics
Solvate a protein-ligand complex, run energy minimization and equilibration, then produce an MD trajectory with live progress tracking.
Prerequisites
A docked complex or a cleaned protein in your project At least 16 GB RAM (32 GB recommended); GPU optional but strongly recommended for long runs
Step 1: Select the system to simulate
From your project, open the docked complex you want to simulate (or just the cleaned receptor for apo-protein MD). Click Start MD to open the simulation form.
Step 2: Choose a force field and water model
Select a force field for the protein (AMBER ff14SB is the default). For the ligand, GAFF2 parameters are generated automatically via OpenFF. Choose a water model (TIP3P default) and set the solvent box padding (1.2 nm recommended).
Step 3: Set simulation parameters
Configure: total simulation time (ns), timestep (2 fs default), temperature (310 K default), pressure coupling (NPT for production). For a first run, 10–50 ns is a reasonable target; you can extend or restart from a checkpoint later.
Step 4: Submit and monitor
Click Run MD. The system is built (solvation, ionisation, force field assignment), then minimised, equilibrated in NVT then NPT, and finally run in production. A live chart shows energy, RMSD, and temperature as they update via WebSocket. Each phase checkpoint is saved so you can resume if the job is interrupted.
Step 5: Review the trajectory
When production is complete, click View trajectory. The trajectory player streams frames into the Mol* viewer. Use the timeline scrubber to jump to any frame. The analytics panel shows RMSD, RMSF, radius of gyration, and ligand contact plots.
Step 6: Download outputs
Download the trajectory (.dcd), final frame (.pdb), and energy CSV. These can be used directly with MDAnalysis, VMD, or as input for ABFE/RBFE calculations.
Expected outputs
trajectory.dcd — full production trajectory final_frame.pdb — last snapshot energy.csv — potential/kinetic/total energy over time RMSD and RMSF plots
Tips
Check the RMSD plot for the first 5–10 ns — if the protein is still drifting, allow longer equilibration before interpreting results. Use the checkpoint resume feature if a long simulation is interrupted; you do not need to restart from the beginning.