Protein-Ligand Scoring

Protein-Ligand Scoring calculates a protein-ligand interaction energy in kcal/mol from the supplied geometry. More negative values indicate stronger predicted interactions. The calculation supports rapid ranking of poses.

Supported binders include small molecules, peptides, macrocycles and stapled peptides. For peptide, macrocycle and stapled-peptide complexes, chain_id identifies the receptor and binder_chain identifies the binder.

The reported quantity is an interaction energy. It does not represent a binding free energy. The score is intended for comparative ranking. Use Absolute Binding Free Energy (ABFE) when an estimate of binding free energy is required.

When to use it

Protein-Ligand Scoring is suitable for scoring docked poses and structures relaxed with Protein-Ligand Pose Refinement. Coordinate quality directly affects the reported energy, so relaxed geometries are preferred for comparative ranking.

Inputs

Supply the complex as a structure. Small-molecule calculations require the protein and either a separate ligand file or the ligand residue name within the protein structure. Peptide, macrocycle and stapled-peptide calculations require the structure, receptor chain and binder chain.

Input	Required	What it is
`protein_file`	yes	Protein structure in PDB or CIF format.
`chain_id`	no	Receptor chain against which the binder is scored.
`ligand_file`	no	Ligand structure in SDF format.
`ligand_resname`	no	Residue name of a ligand represented as a HETATM record in the protein file.
`binder_chain`	no	Binder chain within the same structure. This field selects the peptide path and applies to peptide, macrocycle and stapled-peptide binders. Leave it unset for small molecules.
`ligand_smiles`	no	Optional SMILES definition of the binder chemistry. Supply at most one of `ligand_smiles`, `helm` and `drug_smiles`.
`helm`	no	Binder chemistry in HELM2 notation. This field is an alternative to `ligand_smiles` and is appropriate for macrocycles whose equivalent SMILES is impractical to write manually. See the HELM reference.
`drug_smiles`	no	Synonym of `ligand_smiles`, retained for forwarding this field from Docking.
`crop_radius`	no, default `10`	For a peptide binder, retains receptor residues within this distance in Angstroms of the binder. `0` disables cropping. The default cropped configuration is recommended for large receptors.
`interior_dielectric`	no, default `1`	Solute interior dielectric for the peptide path. Consult the caution below before changing this value or interpreting the sign of a charged binder’s score.
`keep_cofactors`	no	Cofactor residue names to retain, for example `ZN` or `ZN,NDP`.
`extra_chains`	no	Additional protein chain IDs to include.

How to run it

Submit the calculation through Azulene Studio, the Python SDK or the CLI. The Get started page contains installation, authentication and submission instructions.

In Azulene Studio

Open Protein-Ligand Scoring from the tools list. On the Inputs and Parameters step, upload the protein structure. For a small molecule, add a ligand file or specify the ligand residue name and chain. For a peptide binder, specify the receptor chain and binder chain. Select Review and Submit.

From the Python SDK

from azulene import jobs

result = jobs.submit(
    job_type="opal_ml_score",
    input_data={
        "protein_file": "/path/to/your/protein.pdb",
        "ligand_resname": "PJE",
        "chain_id": "C",
    },
)

From the CLI

Pass the inputs as a JSON string.

azulene jobs submit --job-type opal_ml_score \
  --input-data '{"protein_file": "/path/to/your/protein.pdb", "ligand_resname": "PJE", "chain_id": "C"}'

Reading the result

The headline values are best ΔG in kcal/mol (best_dg_kcal_mol) and the number of poses scored (n_poses_scored). best_dg_kcal_mol is the interaction energy of the strongest predicted pose. More negative values indicate stronger predicted interactions.

The ranked table contains every scored pose in order of predicted interaction strength. Each row contains a rank, numbered from 1, a pose label (trial_label) and the pose interaction energy in kcal/mol (dg_kcal_mol). A bar chart displays ΔG for each pose. The table supports sorting, score-based colouring and CSV export. For protein-to-protein inputs, additional columns report per-chain charges and atom counts. These columns are blank for protein-ligand inputs.

Notes

Relaxed structures are preferred when available because the reported interaction energy depends directly on the supplied coordinates. Protein-Ligand Pose Refinement can be applied before scoring.

For a charged peptide binder, treat the sign of the score with caution. At the default interior_dielectric of 1, a charged binder can receive a non-binding score. A higher dielectric can change both the magnitude and sign of the result. Comparisons between binders require a fixed setting. The sign of an isolated score must not be interpreted as a binary binding classification.

For an absolute binding free energy calculation on selected hits, use Absolute Binding Free Energy (ABFE).