Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FuseDiff: Symmetry-Preserving Joint Diffusion for Dual-Target Structure-Based Drug Design

Jianliang Wu, Anjie Qiao, Zhen Wang, Zhewei Wei, Sheng Chen

TL;DR

FuseDiff, an end-to-end diffusion model that jointly generates a ligand molecular graph and two pocket-specific binding poses conditioned on both pockets, achieves state-of-the-art docking performance and enables the first systematic assessment of dual-target pose quality prior to docking-based pose search.

Abstract

Dual-target structure-based drug design aims to generate a single ligand together with two pocket-specific binding poses, each compatible with a corresponding target pocket, enabling polypharmacological therapies with improved efficacy and reduced resistance. Existing approaches typically rely on staged pipelines, which either decouple the two poses via conditional-independence assumptions or enforce overly rigid correlations, and therefore fail to jointly generate two target-specific binding modes. To address this, we propose FuseDiff, an end-to-end diffusion model that jointly generates a ligand molecular graph and two pocket-specific binding poses conditioned on both pockets. FuseDiff features a message-passing backbone with Dual-target Local Context Fusion (DLCF), which fuses each ligand atom's local context from both pockets to enable expressive joint modeling while preserving the desired symmetries. Together with explicit bond generation, FuseDiff enforces topological consistency across the two poses under a shared graph while allowing target-specific geometric adaptation in each pocket. To support principled training and evaluation, we derive a dual-target training set and use an independent held-out test set for evaluation. Experiments on the benchmark and a real-world dual-target system show that FuseDiff achieves state-of-the-art docking performance and enables the first systematic assessment of dual-target pose quality prior to docking-based pose search.

FuseDiff: Symmetry-Preserving Joint Diffusion for Dual-Target Structure-Based Drug Design

TL;DR

FuseDiff, an end-to-end diffusion model that jointly generates a ligand molecular graph and two pocket-specific binding poses conditioned on both pockets, achieves state-of-the-art docking performance and enables the first systematic assessment of dual-target pose quality prior to docking-based pose search.

Abstract

Dual-target structure-based drug design aims to generate a single ligand together with two pocket-specific binding poses, each compatible with a corresponding target pocket, enabling polypharmacological therapies with improved efficacy and reduced resistance. Existing approaches typically rely on staged pipelines, which either decouple the two poses via conditional-independence assumptions or enforce overly rigid correlations, and therefore fail to jointly generate two target-specific binding modes. To address this, we propose FuseDiff, an end-to-end diffusion model that jointly generates a ligand molecular graph and two pocket-specific binding poses conditioned on both pockets. FuseDiff features a message-passing backbone with Dual-target Local Context Fusion (DLCF), which fuses each ligand atom's local context from both pockets to enable expressive joint modeling while preserving the desired symmetries. Together with explicit bond generation, FuseDiff enforces topological consistency across the two poses under a shared graph while allowing target-specific geometric adaptation in each pocket. To support principled training and evaluation, we derive a dual-target training set and use an independent held-out test set for evaluation. Experiments on the benchmark and a real-world dual-target system show that FuseDiff achieves state-of-the-art docking performance and enables the first systematic assessment of dual-target pose quality prior to docking-based pose search.
Paper Structure (50 sections, 1 theorem, 9 equations, 6 figures, 5 tables)

This paper contains 50 sections, 1 theorem, 9 equations, 6 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Structure-Based Drug Design
  4. Protein–ligand complex datasets
  5. Dual-target ligand generation
  6. Methodology
  7. Problem Definition
  8. Dual-target Datasets
  9. Training set BN2-DT
  10. Test set DDF
  11. Diffusion Models for dual-target SBDD
  12. Overview of FuseDiff.
  13. Diffusion Process.
  14. Parameterization of the denoising process.
  15. Optimization.
  16. ...and 35 more sections

Key Result

proposition 1

FuseDiff satisfies R1--R4.

Figures (6)

  • Figure 1: Overview of FuseDiff.FuseDiff defines a conditional diffusion model over $M=(G,X_1,X_2)$ given a pocket pair $P=(P_1,P_2)$, where the two target-specific complexes share the same molecular graph $G$ while maintaining separate poses in $P_1$ and $P_2$. A fixed forward (diffusion) process $q$ (red arrows) corrupts $(G^{(0)},X_1^{(0)},X_2^{(0)})$ into $(G^{(T)},X_1^{(T)},X_2^{(T)})$ for training. For generation, the learned reverse (denoising) process $p_\theta$ (green arrows) starts from $M^{(T)}\!\sim\! p_{\text{base}}(M^{(T)}\mid P)$ and iteratively samples $M^{(t-1)}$ using a denoiser $f_\theta$. At step $t$, $f_\theta$ takes $(G^{(t)},X_1^{(t)},X_2^{(t)})$, $t$, and $P$ as input and predicts $(\widehat{X}_1^{(0)},\widehat{X}_2^{(0)},\widehat{V}^{(0)},\widehat{B}^{(0)})$ to parameterize the reverse transition.
  • Figure 2: Dual-target Local Context Fusion (DLCF) within one denoising layer. (a) Two pocket--ligand point-cloud graphs for $P_1$ and $P_2$, each containing a fully-connected ligand subgraph. (b) The augmented graph induced from the two point-cloud graphs, on which dual-target message passing updates node/edge embeddings. (c) Coordinate update block that separately updates ligand node coordinates in the two point-cloud graphs (with pocket atoms fixed).
  • Figure 3: QED--SA density of filtered molecules ($\mathrm{LPSK}\!=\!5$, $\mathrm{logP}\!\in\![-0.4,5.6]$). KDE heatmaps show $\mathrm{QED}$ (x-axis) versus $\mathrm{SA}$ (y-axis) for FuseDiff (left) and DualDiff (right). The upper-right region corresponds to DL ($\mathrm{QED}\!>\!0.6$, $\mathrm{SA}\!>\!0.7$), where FuseDiff shows higher density, consistent with its larger DL yield (78 vs. 15).
  • Figure 4: Vina Score and Vina Dock distributions on GSK3$\beta$ and JNK3.Left:FuseDiffVina Score on generated poses. Right:Vina Dock after Vina conformational search for FuseDiff and DualDiff.
  • Figure 5: Qualitative visualization in two pockets. Top: GSK3$\beta$; bottom: JNK3. Left: a reference dual-target ligand with poses obtained by Vina pose search. Middle/Right: two FuseDiff samples.
  • ...and 1 more figures

Theorems & Definitions (2)

  • definition 1: Requirements R1--R4
  • proposition 1