Rowan Docs

Pose-Analysis MD

Python API · stjames models · API example

How it works

This workflow simulates a bound protein–ligand pose in explicit water to assess ligand motion, persistent contacts, and binding-site waters. OpenMM minimizes, warms, and equilibrates the complex before running production trajectories.

Pose analysis is also available in Protein MD, which Rowan recommends for new simulations.

Settings

  • Input: provide a prepared protein–ligand complex with explicit hydrogens and matching “Ligand SMILES.” Set “Ligand residue name” to identify the bound ligand (default LIG). The SMILES supplies bond orders and must match the ligand's heavy-atom connectivity. Complete the forcefield compatibility check.
  • Trajectories: “Num trajectories” controls the number of production replicas; “Simulation time (ns)” is the length of each. The web form starts with four 10 ns trajectories at “Temperature (K)” = 300.
  • Saved frames: “Frame save interval (ps)” defaults to 10 ps. Shorter intervals give finer time resolution and larger downloads; the interval must be an integer multiple of the integration timestep.
  • Protein restraints: “Protein restraint cutoff (Å)” defaults to 7 Å in the web form. Protein Cα atoms farther from the ligand are harmonically restrained to their initial positions; closer atoms remain flexible. Leave blank to disable restraints. This does not prune the protein.
  • Additional analysis: “Analysis interval (ps)” defaults to 100 ps in the web form and controls receptor solvent-accessible surface area and MM/GBSA interaction-energy sampling. Leave blank to disable these analyses. Ligand RMSD and contacts still use saved frames.
  • Downloads and clustering: “Save solvent?” retains solvent in exported trajectories. With this enabled, “Max solvent to save” retains the nearest waters to the ligand in each frame; leaving it blank retains all solvent. “Trajectory clustering” groups similar conformations and provides representative frames.
  • Advanced settings: defaults use Sage 2.3.0 for the ligand, ff14SB for the protein, TIP3P water, 0.5 ns equilibration, and a 4 fs timestep with constrained hydrogen bonds and 3 Da hydrogen masses. See Protein MD settings for pressure, salt, water buffer, and other controls.

Notes

Production replicas start from a common equilibrated state. Compare their behavior: agreement across replicas is more informative than a single stable trajectory.

Ligand RMSD measures heavy-atom displacement from each trajectory's first saved frame after aligning protein Cα atoms, without fitting the ligand itself. Persistent contacts are protein–ligand heavy-atom pairs separated by less than 3.5 Å in at least 20% of saved frames; occupancy is the fraction of frames satisfying that distance criterion.

Hydration-site analysis automatically combines replicas to identify recurrent water positions and water-mediated protein–ligand bridges, even when solvent is omitted from downloads. Results include trajectories, the minimized complex, ligand and protein motion metrics, and contact and hydration-site occupancies.

A stable pose supports geometric plausibility over the sampled time and under the chosen restraints. It does not establish binding affinity or prove the pose is correct. MM/GBSA interaction energies omit binding entropy and should not be treated as binding free energies.

Further reading