DynamicFlow: Integrating Protein Dynamics into Drug Design

Paper Information

Citation: Zhou, X., Xiao, Y., Lin, H., He, X., Guan, J., Wang, Y., Liu, Q., Zhou, F., Wang, L., & Ma, J. (2025). Integrating Protein Dynamics into Structure-Based Drug Design via Full-Atom Stochastic Flows. International Conference on Learning Representations (ICLR). https://openreview.net/forum?id=uMAujpVi9m

Publication: ICLR 2025

Additional Resources:

• OpenReview Page

What kind of paper is this?

This is primarily a Methodological Paper ($\Psi_{\text{Method}}$) with a strong Resource ($\Psi_{\text{Resource}}$) component.

• Method: It proposes DynamicFlow, a novel multiscale architecture combining atom-level SE(3)-equivariant GNNs and residue-level Transformers within a flow matching framework to model the joint distribution of ligand generation and protein conformational change.
• Resource: It curates a significant dataset derived from MISATO, pairing AlphaFold2-predicted apo structures with multiple MD-simulated holo states, specifically filtered for this task.

What is the motivation?

Traditional Structure-Based Drug Design (SBDD) methods typically assume the protein target is rigid, which limits their applicability because proteins are dynamic and undergo conformational changes (induced fit) upon ligand binding.

• Biological Reality: Proteins exist as ensembles of states; binding often involves transitions from “apo” (unbound) to “holo” (bound) conformations, sometimes revealing cryptic pockets.
• Computational Bottleneck: While Molecular Dynamics (MD) can simulate these changes, it is computationally expensive due to energy barriers.
• Gap: Existing generative models for SBDD mostly condition on a fixed pocket structure, ignoring the co-adaptation of the protein and ligand.

What is the novelty here?

The core novelty is the simultaneous modeling of ligand generation and protein conformational dynamics using a unified flow matching framework.

• DynamicFlow Architecture: A multiscale model that treats the protein as both full-atom (for interaction) and residue-level frames (for large-scale dynamics), utilizing separate flow matching objectives for backbone frames, side-chain torsions, and ligand atoms.
• Stochastic Flow (SDE): Introduction of a stochastic variant (DynamicFlow-SDE) that improves robustness and diversity compared to the deterministic ODE flow.
• Coupled Generation: Instead of just docking into a rigid pocket, the model learns to transport the apo pocket distribution to the holo pocket distribution while simultaneously denoising the ligand.

What experiments were performed?

The authors validated the method on a curated dataset of 5,692 protein-ligand complexes.

• Baselines: Compared against rigid-pocket SBDD methods: Pocket2Mol, TargetDiff, and IPDiff (adapted as TargetDiff* and IPDiff* for fair comparison of atom numbers). Also compared against conformation sampling baselines (Str2Str).
• Metrics:
  • Ligand Quality: Vina Score (binding affinity), QED (drug-likeness), SA (synthesizability), Lipinski’s rule of 5.
  • Pocket Quality: RMSD between generated and ground-truth holo pockets, Cover Ratio (percentage of holo states successfully retrieved), and Pocket Volume distributions.
  • Interaction: Protein-Ligand Interaction Profiler (PLIP) to measure specific non-covalent interactions.
• Ablations: Tested the impact of the interaction loss, residue-level Transformer, and SDE vs. ODE formulations.

What were the outcomes and conclusions drawn?

• Improved Affinity: DynamicFlow-SDE achieved the best (lowest) Vina scores ($-7.65$) compared to baselines like TargetDiff ($-5.09$) and Pocket2Mol ($-5.50$).
• Realistic Dynamics: The model successfully generated holo-like pocket conformations with volume distributions and interaction profiles closer to ground-truth MD simulations than the initial apo structures.
• Enhancing Rigid Methods: Holo pockets generated by DynamicFlow served as better inputs for rigid-SBDD baselines (e.g., TargetDiff performance improved from $-5.09$ to $-9.00$ when using “Our Pocket”), suggesting the method can act as a “pocket refiner”.
• Robustness: The stochastic (SDE) version outperformed the ODE version in “Cover Ratio” (diversity of states captured) and binding affinity.

Reproducibility Details

Data

The dataset is derived from MISATO, which contains MD trajectories for PDBbind complexes.

Preprocessing:

• Clustering: Holo-ligand conformations clustered with RMSD threshold 1.0 Å; top 10 clusters kept per complex.
• Pocket Definition: Residues within 7Å of the ligand.
• Alignment: AlphaFold predicted structures (apo) aligned to MISATO holo structures using sequence alignment (Smith-Waterman) to identify pocket residues.

Algorithms

Flow Matching Framework:

• Continuous Variables (Pocket translation/rotation/torsions, Ligand positions): Modeled using Conditional Flow Matching (CFM).
  • Prior: Apo state for pocket; Normal distribution for ligand positions.
  • Target: Holo state from MD; Ground truth ligand.
  • Interpolant: Linear interpolation for Euclidean variables; Geodesic for rotations ($SO(3)$); Wrapped linear interpolation for torsions (Torus).
• Discrete Variables (Ligand atom/bond types): Modeled using Discrete Flow Matching based on Continuous-Time Markov Chains (CTMC).
  • Rate Matrix: Interpolates between mask token and data distribution.
• Loss Function: Weighted sum of 7 losses:
  1. Translation CFM (Eq 5)
  2. Rotation CFM (Eq 7)
  3. Torsion CFM (Eq 11)
  4. Ligand Position CFM
  5. Ligand Atom Type CTMC (Eq 14)
  6. Ligand Bond Type CTMC
  7. Interaction Loss (Eq 18): Explicitly penalizes deviations in pairwise distances between protein and ligand atoms for pairs $< 3.5\text{Å}$.

Models

Architecture: DynamicFlow is a multiscale model with 15.9M parameters.

1. Atom-Level SE(3)-Equivariant GNN:
   • Input: Complex graph (k-NN, $k=32$) and Ligand graph (fully connected).
   • Layers: 6 EGNN blocks modified to maintain node and edge hidden states.
   • Function: Updates ligand positions and predicts ligand atom/bond types.
2. Residue-Level Transformer:
   • Input: Aggregated atom features from the GNN + Residue frames/torsions.
   • Layers: 4 Transformer blocks with Invariant Point Attention (IPA).
   • Function: Updates protein residue frames (translation/rotation) and predicts side-chain torsions.

Evaluation

Metrics:

• Vina Score: vina_minimize mode used for binding affinity.
• RMSD: Minimum RMSD between generated pocket and ground-truth holo conformations.
• Cover Ratio: % of ground-truth holo conformations covered by at least one generated sample (threshold 1.42 Å).
• POVME 3: For pocket volume calculation.

Hardware

• Inference Benchmark: 1x Tesla V100-SXM2-32GB.
• Speed: Generates 10 ligands in ~35-36 seconds (100 NFE), significantly faster than diffusion baselines like Pocket2Mol (980s) or TargetDiff (156s).

Citation

@inproceedings{zhouIntegratingProteinDynamics2025,
  title = {Integrating Protein Dynamics into Structure-Based Drug Design via Full-Atom Stochastic Flows},
  author = {Zhou, Xiangxin and Xiao, Yi and Lin, Haowei and He, Xinheng and Guan, Jiaqi and Wang, Yang and Liu, Qiang and Zhou, Feng and Wang, Liang and Ma, Jianzhu},
  booktitle = {International Conference on Learning Representations},
  year = {2025},
  url = {https://openreview.net/forum?id=uMAujpVi9m}
}