molecular-dynamics

name	molecular-dynamics
description	Molecular dynamics simulations using OpenMM with MDAgent-style automation and free energy calculations via MM-PBSA
metadata	{"openclaw":{"requires":{"env":["OPENAI_API_KEY"],"bins":["python3"],"anyBins":["gmx","cpptraj"]},"primaryEnv":"OPENAI_API_KEY"}}

Molecular Dynamics — OpenMM Simulations

Run molecular dynamics simulations on protein structures using OpenMM and calculate binding free energies via MM-PBSA.

Capabilities

• System setup: Prepare protein systems with solvent, ions, and force fields (AMBER14/CHARMM36)
• Energy minimization: Minimize system energy before production runs
• Production MD: Run NVT/NPT simulations with configurable parameters
• Free energy calculations: MM-PBSA binding free energy via AmberTools/cpptraj
• Trajectory analysis: Basic RMSD, RMSF, and contact analysis (see trajectory-analysis skill for advanced)

Scripts

scripts/run_md.py — MD simulation launcher

Wraps the MDAgent/OpenMM simulation pipeline. Accepts a PDB input, builds the system (implicit or explicit solvent), runs equilibration and production MD, and records artifacts to the Artifact DAG for provenance tracking.

python skills/molecular-dynamics/scripts/run_md.py \
  --input-pdb structure.pdb \
  --output-dir results/ \
  --forcefield amber14-all \
  --water-model tip3p \
  --temperature 300 \
  --pressure 1.0 \
  --equil-steps 10000 \
  --prod-steps 500000 \
  --solvent explicit \
  --platform CUDA

scripts/run_free_energy.py — MM-PBSA free energy calculation

Runs MM-PBSA binding free energy calculations on completed MD trajectories. Requires AmberTools (cpptraj) and a production trajectory with topology.

python skills/molecular-dynamics/scripts/run_free_energy.py \
  --sim-dir trajectory_dir/ \
  --output-dir fe_results/ \
  --n-cpus 4 \
  --amberhome /path/to/amber

Parameters

MD Simulation

• input_pdb: Path to input PDB structure
• forcefield: Force field to use (default: "amber14-all")
• water_model: Water model (default: "tip3p")
• temperature_k: Simulation temperature in Kelvin (default: 300)
• pressure_atm: Pressure in atm for NPT (default: 1.0)
• equil_steps: Equilibration steps (default: 10000)
• prod_steps: Production MD steps (default: 500000)
• timestep_fs: Integration timestep in femtoseconds (default: 2.0)
• solvent: Solvent model — "explicit" or "implicit" (default: "explicit")
• platform: OpenMM platform — "CUDA", "OpenCL", or "CPU" (default: "CUDA")

Free Energy (MM-PBSA)

• sim_dir: Directory containing production trajectory (prod.dcd + system.prmtop)
• n_cpus: Number of CPUs for parallel MM-PBSA (default: 4)
• amberhome: Path to AmberTools installation

Requirements

• OpenMM (openmm): MD simulation engine
• MDAnalysis (MDAnalysis): Trajectory I/O and atom selection
• AmberTools (cpptraj): MM-PBSA free energy calculations
• MDAgent (optional): Academy-based MD orchestration agents
• Academy (optional): Distributed agent execution (Layer 4)