IndiskaAI Logo
Back
Structural analysis

Molecular Docking

Candidate binding poses and interface maps, ranked and explained.

4 stages

Our workflow

  1. 01

    Structure Preparation

    Target and partner are processed, with the binding site defined where applicable.

  2. 02

    Candidate Pose Generation

    Binding orientations are sampled and evaluated with the selected docking method.

  3. 03

    Pose Selection

    Representative poses are shortlisted by score, clustering, and structural assessment.

  4. 04

    Interface Characterization

    Interface residues and predicted interactions of the selected complex are characterised.

Illustrative renders, not outputs from an actual run.

From predicted binding poses to structural insights for drug discovery.

Docking types

Six system types, each with its own sampling and scoring considerations.

  • Protein surface with a small-molecule ligand (orange) sitting inside a binding pocket

    Protein–Small Molecule

    Pocket-directed docking with interaction breakdown.

  • Two protein partners (blue and green) with their shared interface highlighted in orange

    Protein–Protein

    Association modes and interface characterisation.

  • Antibody with heavy and light chains in blue shades and an antigen (magenta) engaging the Fab tip, interface in orange

    Antibody–Antigen

    CDR-aware docking of the paratope–epitope interface.

  • Protein surface with a helical peptide (orange) bound across it

    Protein–Peptide

    Peptide binding modes and interactions, allowing for backbone flexibility.

  • Protein surface (lavender) alongside a DNA double helix

    Protein–DNA

    DNA recognition and binding-interface analysis, including groove contacts.

  • Protein surface (green) with an RNA helix (orange) bound along it

    Protein–RNA

    RNA interaction modes and interface mapping, allowing for RNA flexibility.

Illustrative renders of each system type.

What we analyse

  • Binding-Pose Prediction

    Real structure · PDB 1HSG

    Generate candidate poses for a ligand in a protein's binding pocket and rank them by docking score and key contacts. A score orders poses; it is not a binding affinity.

  • Binding Interface Characterization

    Real structure · PDB 1BRS

    Identify where two proteins meet, mapping interface residues within 4.5 Å and listing contacts whose geometry is consistent with possible hydrogen bonds or salt bridges. These are contacts in a model, not confirmed interactions.

  • Residue-Level Interaction Mapping

    Real structure · PDB 3HFM

    Resolve the interface pair by pair, from each contacting residue to its partner, contact type and distance in ångströms.

  • Comparative Candidate Analysis

    Illustrative data

    Compare docking outcomes for the parent, a reference and AI-assisted candidates side by side under one protocol. Scores are for ordering only.

Also availableHotspot analysisEpitope / paratope mappingSAR interpretation across a ligand seriesPredicted model assessment
Docking and MD together
  1. 01

    Docking

  2. 02

    Candidate poses

  3. 03

    Molecular dynamics

  4. 04

    Interaction persistence

Docking proposes poses; simulation tests whether they hold. MD does not automatically validate a pose.

From Inputs to Insights

  • A protein sequence, or a protein structure (experimental or predicted)
  • Antibody heavy and light chains, where relevant
  • A peptide, ligand or set of candidate molecules
  • Known binding-site information, if you have it
  • Any constraints that apply

Applications

Drug discoveryBiologicsAntibody engineeringPeptide therapeuticsProtein engineeringMolecular recognition
Scope & limitations

A high-scoring pose is a structural hypothesis, not a demonstration that two molecules bind, and docking scores are not affinities.

Have a structure to investigate?

Explore predicted binding poses, interfaces, and residue-level interactions with a docking workflow designed around your molecular system.