Abstract:Structure-based drug design (SBDD) aims to generate 3D ligand molecules that bind to specific protein targets. Existing 3D deep generative models including diffusion models have shown great promise for SBDD. However, it is complex to capture the essential protein-ligand interactions exactly in 3D space for molecular generation. To address this problem, we propose a novel framework, namely Binding-Adaptive Diffusion Models (BindDM). In BindDM, we adaptively extract subcomplex, the essential part of binding sites responsible for protein-ligand interactions. Then the selected protein-ligand subcomplex is processed with SE(3)-equivariant neural networks, and transmitted back to each atom of the complex for augmenting the target-aware 3D molecule diffusion generation with binding interaction information. We iterate this hierarchical complex-subcomplex process with cross-hierarchy interaction node for adequately fusing global binding context between the complex and its corresponding subcomplex. Empirical studies on the CrossDocked2020 dataset show BindDM can generate molecules with more realistic 3D structures and higher binding affinities towards the protein targets, with up to -5.92 Avg. Vina Score, while maintaining proper molecular properties. Our code is available at

What problem does this paper attempt to address?

### What problems does this paper attempt to solve? This paper aims to solve the key challenges in Structure - Based Drug Design (SBDD). Specifically, the goal of SBDD is to generate 3D ligand molecules that can bind to specific protein targets and regulate their functions. Although existing 3D deep - generation models (such as diffusion models) have shown great potential in SBDD tasks, accurately capturing the complexity of protein - ligand interactions in 3D space remains a difficult problem. #### Main problems 1. **Complexity of protein - ligand interactions**: It is very difficult to accurately capture the essential interactions between proteins and ligands in 3D space. 2. **Limitations of existing methods**: - **Autoregressive models (ARMs)**: They are prone to error accumulation and it is difficult to find the optimal generation sequence. - **Existing diffusion models**: Although they perform excellently, they pay insufficient attention to the sub - structures of specific binding sites in protein - ligand complexes, and these sub - structures are crucial for generating high - affinity molecules. #### Proposed solutions To solve these problems, the authors propose a new framework - **Binding - Adaptive Diffusion Models (BINDDM)**. The main innovations of BINDDM include: 1. **Adaptive extraction of binding sub - complexes**: By learning an adjustable structural pooling method, directly extract the key sub - complexes responsible for binding from protein - ligand complexes. 2. **SE(3) - equivariant neural network processing**: Use SE(3) - equivariant neural networks to process the selected sub - complexes and pass the information back to the entire complex to enhance target - aware 3D molecular diffusion generation. 3. **Cross - level interaction nodes**: By designing cross - level interaction nodes, iteratively fuse the global binding context to ensure sufficient information exchange between the complex and its sub - complexes. #### Experimental results Experiments show that BINDDM can generate molecules with more realistic 3D structures and higher binding affinities on the CrossDocked2020 dataset while maintaining appropriate molecular properties. For example, BINDDM has an average Vina score of - 5.92, which is significantly better than other benchmark methods. ### Summary BINDDM effectively solves the complexity problem of protein - ligand interactions in SBDD by introducing binding - adaptive sub - complex extraction and cross - level interaction nodes, and improves the stability and binding affinity of generated molecules.

Binding-Adaptive Diffusion Models for Structure-Based Drug Design

AUTODIFF: Autoregressive Diffusion Modeling for Structure-based Drug Design

General Binding Affinity Guidance for Diffusion Models in Structure-Based Drug Design

Structure-based Drug Design with Equivariant Diffusion Models

BInD: Bond and Interaction-generating Diffusion Model for Multi-objective Structure-based Drug Design

Interaction-based Retrieval-augmented Diffusion Models for Protein-specific 3D Molecule Generation

A Unified Conditional Diffusion Framework for Dual Protein Targets Based Bioactive Molecule Generation

Protein-Ligand Interaction Prior for Binding-aware 3D Molecule Diffusion Models

DiffBP: Generative Diffusion of 3D Molecules for Target Protein Binding

Rectified Flow For Structure Based Drug Design

DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Design

DynamicBind: predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model

DiffBindFR: An SE(3) Equivariant Network for Flexible Protein-Ligand Docking

Structure-based Drug Design Benchmark: Do 3D Methods Really Dominate?

Manifold-Constrained Nucleus-Level Denoising Diffusion Model for Structure-Based Drug Design

Rethinking Specificity in SBDD: Leveraging Delta Score and Energy-Guided Diffusion

FlexSBDD: Structure-Based Drug Design with Flexible Protein Modeling

Delta Score: Improving the Binding Assessment of Structure-Based Drug Design Methods

Reinforced Genetic Algorithm for Structure-based Drug Design

CBGBench: Fill in the Blank of Protein-Molecule Complex Binding Graph