Skip to main content

Overview

This tutorial covers the first two docking modes available in Rev-Bind’s Virtual Screening Engine: Rigid Receptor Docking and Flexible Docking. Together, they form the foundation of Revilico’s tiered virtual screening workflow — from rapid, high-throughput screening all the way to structure-activity-guided hit refinement.
Static Docking and Flexible Docking Review — Watch Video

Docking Modes at a Glance

Rev-Bind offers three docking modes, each designed for a different stage of your screening campaign:
ModeProtein FlexibilityRescoringBest For
Rigid ReceptorStaticStandardRapid library-scale screening
Flexible DockingKey residues flexCNN rescoringSAR-guided hit refinement
Ensemble DockingFull MD-sampled flexStandardHigh-accuracy late-stage screening
This tutorial covers Rigid Receptor and Flexible Docking. For Ensemble Docking, see the Ensemble Docking tutorial.

Part 1: Rigid Receptor Docking

Rigid receptor docking keeps the protein structure static and samples multiple conformations of each ligand within the binding pocket. This is your fastest, highest-throughput screen — ideal for processing large compound libraries quickly.

How It Works

The algorithm holds the protein fixed, then exhaustively samples ligand conformations inside the defined binding box. Each pose is scored using an energy-based function, and the top-ranked conformations are returned for downstream analysis.

Step-by-Step Walkthrough

1

Navigate to Virtual Screening

From the Revilico OS dashboard, go to Binding Chemistry → Virtual Screening Engine. You will see three docking modes: Rigid Receptor, Flexible, and Ensemble. Select Rigid Receptor Docking.
2

Name Your Pipeline

Enter a descriptive name for your screening campaign. This name will identify the pipeline in your results history.
3

Upload Your Compound Library

Upload your compound library (SDF or SMILES format), or select demo files to follow along. In this walkthrough, we use 800 AXL inhibitor compounds.
4

Select Your Protein of Interest

Search for and select your target protein. In this walkthrough, we use AXL (no inhibitor) — the apo form of the receptor. The platform will automatically populate the 3D protein structure.
5

Define the Binding Pocket

Using structural information from literature or a prior RevPocket search, draw the docking box around your binding pocket of interest. This tells the algorithm where on the protein to compute interaction energies.
Run a RevPocket analysis first if you have not yet characterized the binding site. It will give you the coordinates and key residue information you need to set the box accurately.
6

Set Exhaustiveness

Adjust the exhaustiveness parameter to control the depth of conformational sampling. Higher values increase accuracy at the cost of compute time. See the RevScreen documentation for guidance on calibrating this parameter to your campaign needs.
7

Run the Pipeline

Click Run Pipeline → Create → Close. Your rigid receptor docking campaign is now queued. Results will populate in the RevAnalytics pane when the job completes.

Part 2: Flexible Docking

Flexible docking builds on rigid receptor docking by allowing selected binding site residues to move during the simulation. It also applies Convolutional Neural Network (CNN) rescoring to improve pose quality assessment — delivering more confident binding predictions when you already have SAR or structural context.

When to Use Flexible Docking

Use flexible docking as a secondary filter after rigid receptor screening, particularly when:
  • You have SAR data from previous campaigns or literature identifying key binding interactions
  • You know which residues engage with your ligands (e.g., hinge region residues for kinases)
  • You want CNN rescoring to re-rank your top poses with greater accuracy

How CNN Rescoring Works

After docking, the Convolutional Neural Network rescoring layer re-evaluates each pose using a 3D-aware deep learning model trained on crystallographic binding data. This step helps correct for cases where classical docking scoring functions misrank poses — especially useful for novel chemotypes or compounds with unusual binding geometries.

Step-by-Step Walkthrough

1

Navigate to Flexible Docking

From the Virtual Screening Engine, select Flexible Docking.
2

Configure Your Inputs

Enter a pipeline name. If you are continuing from a rigid receptor run, your compound library and protein will carry over automatically. Otherwise, re-upload your compound library and re-select your protein.
3

Identify Key Flexible Residues

This is the critical step that differentiates flexible from rigid docking. Using your SAR knowledge or structural literature, identify which binding site residues you want to allow to flex during the calculation.In the walkthrough example, we select Pro621 and additional hinge region residues based on known kinase binding interactions. Click on residues in the 3D viewer to toggle their flexibility.
Not all residues can be set to flex — the platform will indicate which ones are eligible. Focus on residues for which you have structural evidence of involvement in key binding interactions.
4

Define the Docking Box

Position the docking box around your binding pocket, just as in rigid receptor docking. Calibrate the exhaustiveness parameter based on your campaign requirements.
5

Run the Pipeline

Click Run Pipeline → Create. The algorithm will sample ligand conformations, allow the designated residues to flex, and apply CNN rescoring to refine pose rankings. Your results will appear in RevAnalytics when complete.

Comparing Results

After running both screens, compare the rank order from rigid vs. flexible docking. Compounds that rank highly in both are your highest-confidence hits for progression. Discordant rankings — especially where a compound improves under flexible docking — often indicate that the flexible residues you selected are critical for binding.

Next Steps

Once you have completed rigid and flexible docking screens, your top-ranked compounds are ready for:
  • Ensemble Docking — For highest-priority compounds, run ensemble docking against MD-sampled protein conformations for maximum accuracy
  • RevAnalytics — Analyze docking scores, interaction fingerprints, and binding pose quality across your screened library
  • RevFEP — Run rigorous free energy perturbation calculations on your shortlisted candidates
  • RevMD-Bind — Validate top compounds with full protein-ligand MD simulations