nvidia/nvDock

Model Overview

Description:

All-atom DiffDock Pocket is a diffusion-based model for molecular pocket docking, designed for cases where the protein binding pocket is known in advance. Unlike the original DiffDock, which mainly represents proteins using Cα atoms, the all-atom version models the full atomic structure of the protein pocket. This allows the model to capture more detailed protein–ligand interactions and improves docking accuracy in complex binding environments.

The framework includes two components: a Score model and a Confidence model. The Score model generates multiple candidate binding poses for a ligand inside the target pocket through a reverse diffusion process, while the Confidence model ranks these poses to identify the most reliable binding configuration.

The model is built on the neural network architecture introduced in the original DiffDock and is trained on curated protein–ligand complex datasets. By using all-atom protein representations together with pocket-focused training, the model can better learn local structural and chemical interactions within the binding site while maintaining efficient inference for docking tasks.

This model is available for both commercial and non-commercial use.

License/Terms of Use:

• Use of the model is governed by the NVIDIA Open Model Agreement.
  
• Use of source code is governed by Apache License, Version 2.0.
  
• ADDITIONAL INFORMATION: MIT License.
  

Deployment Geography:

Global

Use Case:

All-atom DiffDock Pocket is used when the binding site of a protein is already known. It helps predict how a small molecule is likely to fit into that pocket. The model generates multiple binding poses and ranks them to highlight the most reasonable ones. This allows researchers to quickly evaluate candidate compounds before running expensive simulations or experiments. As a result, docking and early screening become faster and more practical.

Release Date:

GitHub 07/10/2026 via URL
Hugging Face 07/10/2026 via URL

Model Architecture:

The Score model is a 3-dimensional equivariant graph neural network that has three layers: embedding, interaction layer with 9 graph convolution layers, and output layer.

• Architecture Type: Score-Based Diffusion Model (SBDM)
  
• Network Architecture: Graph Convolution Neural Network
  
• This model was developed based on: Diffdock Pocket.
  
• Number of model parameters: 44M

The Confidence model is a 3-dimensional SO(3)-equivariant heterogeneous graph neural network that has three stages: (1) per-type node embedding with edge-degree initialization, (2) interaction layers consisting of 3 HeteroEGA transformer blocks — each containing 9 Equivariant Graph Attention (EGA) modules over a heterogeneous graph (3 intra-type + 6 cross-type edges across ligand, receptor-residue, and receptor-atom nodes) and 3 per-type feed-forward networks with pre-norm residual connections, and (3) an attention-weighted pooling readout over ligand nodes followed by an MLP confidence predictor.

• Architecture Type: Confidence Estimation Model
• Network Architecture: Heterogeneous SO(3)-Equivariant Graph Attention Network (HeteroEGA)
• This model was developed based on: DiffDock-Pocket confidence architecture, replacing the Tensor-Product Convolutions with Equiformer v3 SO(2)-equivariant graph attention while preserving the original 9-way heterogeneous message-passing topology.
• Number of model parameters: 7.4M

Input:

Input Type(s): Text (Ligand, Protein), Number (Pocket center Coordinate, Poses to Generate, Batch Size, Diffusion Steps, Diffusion Time Divisions) Binary (No Final Step Noise, Save Diffusion Trajectory, and Skip Gen Conformer)
Input Format(s): Text: String (SMILES, Structural Data Files (SDF) or Tripos molecule structure (Mol2) for Ligand), String (Protein Data Bank (PDB)), Number: Integer; Binary: Boolean
Input Parameters: Text: One-Dimensional (1D), Number: One-Dimensional (1D), Binary: One-Dimensional (1D)
Other Properties Related to Input: No max sequence

Output:

Output Type(s): Text (Ligand Molecule 3D Positions), Text (Ligand Molecule 3D Positions), Number (List of Confidence Scores)
Output Format: Text: Structural Data Files (SDF), Text: Protein Data Bank (PDB), Number: Array of Floating Point 32
Output Parameters: docked_ligand (.sdf file), visualizations_files (.sdf files), pose_confidence (number)
Other Properties Related to Output: Output includes ranked binding poses with associated confidence scores. Higher confidence scores indicate more reliable predictions.

Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA's hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.

Software Integration:

Runtime Engine(s):

• PyTorch
  

Supported Hardware Microarchitecture Compatibility:

• NVIDIA Ampere
  
• NVIDIA Grace Hopper
  
• NVIDIA Hopper
  
• NVIDIA Lovelace
  

Supported Operating System(s):

• Linux
  

The integration of foundation and fine-tuned models into AI systems requires additional testing using use-case-specific data to ensure safe and effective deployment. Following the V-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment.

Model Version(s):

nvDock version 1.0

Training & Evaluation Dataset:

Data Modality:

Other: 3D Molecular Structures (protein-ligand complexes)

Training Data Size:

975,000 protein-ligand complexes (100,000 from PLINDER-time + 875,000 selected from SAIR)

Training:

Link: PLINDER

Link: SAIR

Data Collection Method by dataset:

• Manually-Collected
  

Labeling Method by dataset:

• Hybrid: Manually-Collected, Automated
  

Properties (Quantity, Dataset Descriptions, Sensor(s)): 975,000 protein-ligand complexes (100,000 from PLINDER-time automatically curated using the PDB database + 875,000 selected from SAIR). For more information, see Technical Paper.

Evaluation:

Link: PoseBusters benchmark (PDB) set

Data Collection Method by dataset:

• Manually-Collected
  

Labeling Method by dataset:

• Hybrid: Manually-Collected, Automated
  

Properties (Quantity, Dataset Descriptions, Sensor(s)): 306 complexes from PoseBusters. For more information, see Technical Paper.

Evaluation result:

Method	Type	Top-1 (%)	Oracle (%)
EqBind	Blind	2.0	–
TankBind	Blind	16.0	–
DiffDock	Blind/Conf	38.0	–
ArtiDock	Pocket-AA	78.0	–
SigmaDock	Pocket-AA / Physics	80.5	92.0
DiffDock-RL	Pocket-AA / RL / Conf	69.0	84.8
DiffDock-RL++	Pocket-AA/ RL / Physics	80.2	88.5
nvDock	Pocket-AA / Conf	81.85	94.51

Inference:

Engine: PyTorch
Test Hardware:

• NVIDIA Hopper A10G
• NVIDIA Hopper A100
• NVIDIA Hopper RTX6000-Ada,
• NVIDIA Hopper H100
• NVIDIA Hopper L40
• NVIDIA Hopper L40S

Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

Users are responsible for ensuring the physical properties of model-generated molecules are appropriately evaluated and comply with applicable safety regulations and ethical standards.

For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards.

Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns here.

Citations:

• Diffdock: Link
• Diffdock-Pocket: Link
• PLINDER: Link
• SAIR: Link

Get Help

Enterprise Support

Get access to knowledge base articles and support cases or submit a ticket.