ProteinIQ

DiffDock-L

AI-powered molecular docking for predicting protein-ligand binding poses using diffusion models

proteinsmall moleculemolecular dockingstructure predictionartificial intelligencediffusion modelsdrug discovery
Loading...

Model overview: DiffDock-L

What is DiffDock-L?

DiffDock-L is the latest version of the state-of-the-art molecular docking tool developed at MIT that uses diffusion models to predict how small molecule ligands bind to protein targets. Building upon the original DiffDock published at ICLR 2023, DiffDock-L represents a paradigm shift in computational docking by treating the problem as a generative modeling task rather than traditional optimization.

DiffDock-L vs. DiffDock

Released in February 2024, DiffDock-L introduced several enhancements over the original DiffDock:

  • Enhanced accuracy: DiffDock-L achieves 43% success rate for ligand RMSD <2Å compared to the original DiffDock's 38%, representing a substantial improvement in prediction accuracy. On the DockGen benchmark specifically designed to test generalization, DiffDock-L improved success rate from 10% to 24%.
  • Better generalization: The model uses ECOD structure-based cluster sampling to ensure better performance on proteins with novel binding pocket architectures. This addresses a key limitation where the original DiffDock struggled with unseen protein targets.
  • Larger model & more training data: DiffDock-L benefits from increased training data, larger model sizes, and novel synthetic data generation techniques, leading to more robust predictions across diverse protein-ligand systems.
  • Optimized output: By default, DiffDock-L outputs 10 high-quality predictions compared to the previous 40, focusing on quality over quantity and reducing computation time by approximately 2x while maintaining superior accuracy.

How does DiffDock-L work?

Unlike traditional docking methods that use physics-based scoring functions and optimization algorithms, DiffDock-L applies diffusion models - the same technology behind AI image generators like DALL-E and Stable Diffusion.

The key innovation is treating molecular docking as a generative process:

  • Diffusion process: The model starts with a random ligand position and orientation, then iteratively "denoises" it through learned diffusion steps to converge on a plausible binding pose.

  • SE(3) equivariance: The model respects the symmetries of 3D space - rotations and translations don't change the physics of binding. This is achieved using specialized geometric neural networks.

  • Confidence modeling: For each generated pose, DiffDock-L produces a confidence score that estimates the likelihood of the prediction being correct.

This approach allows DiffDock-L to:

  • Handle flexible ligands more naturally than rigid docking methods
  • Generate diverse binding poses without manual parameter tuning
  • Achieve state-of-the-art accuracy on standard benchmarks

Input requirements

DiffDock-L requires two inputs:

  • Protein structure: A PDB file or PDB ID from RCSB PDB database. The protein should be prepared (protonated, cleaned) for best results (you can use tools like PDB Fixer to prepare it).
  • Ligand: SMILES string (most common for small molecules), SDF or MOL2 file format are also supported.

Understanding the results

DiffDock-L generates multiple binding poses (default: 10) ranked by confidence:

Rank: Poses are sorted by confidence score, with Rank 1 being the most confident prediction.

Confidence Score: A model-predicted score indicating confidence in the binding pose. Higher (less negative) scores indicate higher confidence. Typical ranges:

  • Score > -1.5: High confidence
  • Score -1.5 to -2.5: Moderate confidence
  • Score < -2.5: Lower confidence

Multiple Poses: Examining multiple top-ranked poses is recommended, as:

  • Different poses may represent alternative binding modes
  • The highest-ranked pose isn't always correct
  • Ensemble analysis provides more robust predictions

Best practices

Number of poses:

  • Start with 10-20 poses for general use
  • Increase to 30-40 if exploring diverse binding modes
  • More poses increase computation time proportionally

Protein preparation:

  • Remove water molecules unless they're critical for binding
  • Add missing hydrogens and residues
  • Ensure proper protonation states at physiological pH

Result validation:

  • Examine top 3-5 ranked poses, not just the top one
  • Check for reasonable protein-ligand interactions (H-bonds, hydrophobic contacts)
  • Validate critical predictions with molecular dynamics or experimental methods

Limitations:

  • Does not account for protein flexibility (rigid receptor)
  • Performance may vary for large/macrocyclic ligands
  • Metalloprotein binding requires special consideration

Comparison to traditional docking

Advantages of DiffDock-L:

  • Superior accuracy on benchmark datasets
  • Better handling of flexible ligands
  • No manual parameter tuning required
  • Generates diverse poses naturally

Traditional methods (AutoDock, Vina, GOLD):

  • Faster computation time
  • More interpretable scoring functions
  • Established validation protocols
  • Better support for specialized cases (metals, covalent binding)

For most drug discovery applications, DiffDock-L provides the best balance of accuracy and ease of use.

Cost

Using DiffDock-L through ProteinIQ costs 100 credits per docking job, regardless of the number of poses generated.

Based on: Corso, G., Stärk, H., Jing, B., Barzilay, R., & Jaakkola, T. (2023). DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking. International Conference on Learning Representations (ICLR). URL: https://openreview.net/pdf?id=kKF8_K-mBbS