MD binding-energy assessment
GROMACS 2026.3 · ReadyHow to run the protein–protein MD binding-energy workflow, reopen persisted tasks, inspect artifacts, and interpret the MD + MM/PBSA estimate responsibly.
PDB/ENT upload or paste, partner assignment, production MD, PBC correction, Cα RMSD and analysis-window audit, MM/PBSA and residue decomposition, persistent history, downloads, and Mol* visualization.
MM/PBSA is an end-state estimate, not experimental affinity or rigorous absolute free energy. An RMSD plateau is a structural-stability diagnostic, not proof of thermodynamic convergence. Production comparisons still need matched protocols and independent replicas.
1. Upload or drop a PDB
Choose a .pdb/.ent file. The file is read locally and validated before it is sent.
2. Assign both partners
Review detected chain IDs, then assign receptor and ligand chains. The groups must exist and must not overlap.
3. Run MD and MM/PBSA
Modal prepares and runs MD, computes backbone-fitted Cα RMSD, records the selected stable window, then evaluates MM/PBSA on those frames.
4. Reopen the result
Select a history card to restore chain assignments, stage progress, energy components, artifacts, and the representative 3D structure.
Automatic selection
The workflow fits each frame on the complex backbone, calculates Cα RMSD after periodic-boundary correction, and scans tail windows from earlier to later. It accepts the earliest tail that is at least 250 ps long and satisfies all three gates: |slope| ≤ 0.5 Å/ns, RMSD standard deviation ≤ 1.0 Å, and early-to-late drift ≤ 0.75 Å.
Fail-safe interpretation
Runs below 1 ns are labelled too short to assess. If a longer run passes no window, the final half is used only as a non-converged diagnostic and the UI says No plateau detected. A manual start in Advanced overrides automatic selection and remains recorded. An RMSD plateau is not proof of thermodynamic or MM/PBSA convergence; consequential comparisons need matched protocols and independent replicas.
Structure file
Required for the default workflow
.pdb or .ent, up to 20 MiB. The browser checks that the file contains ATOM or HETATM coordinate records.
Receptor / ligand chains
Required
Two non-overlapping groups of PDB chain IDs. For an antibody–antigen complex, use the antibody heavy/light chains as receptor and antigen chains as ligand (or the reverse, but stay consistent across comparisons).
Simulation preset
Required · validation / screening / production
20 ps is a technical pipeline check, 1 ns is preliminary, and 10 ns is a first production estimate. Only matched protocols should be compared.
Temperature
Required · default 300 K
Accepted range is 273–330 K and the same value is used for equilibration and PBSA parameters.
Run name
Optional · Advanced
A human-readable history label. When omitted, the uploaded PDB filename is used.
Energy analysis start
Optional · Advanced
Leave blank to detect an RMSD-stable tail automatically. Enter a time in ps only when you have reviewed the trace and need a manual window; the override is recorded in the result.
Raw GROMACS mode
Advanced · diagnostic only
Switch execution mode only when you need an expert command-level diagnostic. The guided binding-energy workflow does not expose a shell command.
The guided workflow accepts standard-residue soluble protein complexes. Non-water HETATM records, non-standard residues, glycans, small molecules, metals, membranes, and covalent adducts are rejected until separately parameterized.
Use example loads a real three-chain protein structure and fills receptor A+B, ligand C, 300 K, and the validation preset. Replace every field with your own complex when running real work.
Overview
Chain assignments, live progress, RMSD trace, selected energy window, total/component energies, strongest residue contributions, limits, and errors.
Files
RMSD XVG/CSV, stability JSON, selected-window XTC, energy/residue CSV, residue-energy PDB, TPR/index, representative PDB, and workflow log.
3D structure
The first returned PDB/ENT artifact opens in Mol* for rotation, zoom, residue inspection, and snapshots.
Required review
- Biological construct, missing residues, termini, protonation and disulfides
- Potential energy, temperature, pressure and density equilibration
- Trajectory stability, PBC correction and frame independence
- At least three independent replicas for consequential comparisons
- Matched force field, preset, temperature and sampling window across candidates
Interpretation boundary
More-negative MM/PBSA values are only directionally favorable within a matched protocol. Do not convert them directly to Kd/IC50 or compare them across unrelated constructs. A single trajectory cannot establish binding, specificity, neutralization, or mechanism; confirm important calls experimentally.
The upload is rejected
Confirm the extension is .pdb or .ent, the file is below 20 MiB, and at least one line begins with ATOM or HETATM.
The task remains queued/running
A cold Modal container can take longer to start. The page refreshes automatically; leaving and reopening the task does not cancel it.
The complex is unsupported
Remove crystallographic waters if needed and inspect every non-water HETATM or non-standard residue. Ligands, glycans, metals and modified residues require explicit parameters and are intentionally blocked.
No 3D view appears
Wait for completion and confirm representative.pdb is listed in Files. Other trajectory and energy artifacts remain downloadable even if the viewer cannot initialize.