mlsanityData to 7 Oct 2026

BoKDiff: Best-of-K Diffusion Alignment for Target-Specific 3D Molecule Generation

arXiv 2501.15631 ↗First posted 26 Jan 2025q-bio.BM cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Structure-based drug design (SBDD) leverages the 3D structure of biomolecular targets to guide the creation of new therapeutic agents. Recent advances in generative models, including diffusion models and geometric deep learning, have demonstrated promise in optimizing ligand generation. However, the scarcity of high-quality protein-ligand complex data and the inherent challenges in aligning generated ligands with target proteins limit the effectiveness of these methods. We propose BoKDiff, a novel framework that enhances ligand generation by combining multi-objective optimization and Best-of-K alignment methodologies. Built upon the DecompDiff model, BoKDiff generates diverse candidates and ranks them using a weighted evaluation of molecular properties such as QED, SA, and docking scores. To address alignment challenges, we introduce a method that relocates the center of mass of generated ligands to their docking poses, enabling accurate sub-component extraction. Additionally, we integrate a Best-of-N (BoN) sampling approach, which selects the optimal ligand from multiple generated candidates without requiring fine-tuning. BoN achieves exceptional results, with QED values exceeding 0.6, SA scores above 0.75, and a success rate surpassing 35%, demonstrating its efficiency and practicality. BoKDiff achieves state-of-the-art results on the CrossDocked2020 dataset, including a -8.58 average Vina docking score and a 26% success rate in molecule generation. This study is the first to apply Best-of-K alignment and Best-of-N sampling to SBDD, highlighting their potential to bridge generative modeling with practical drug discovery requirements. The code is provided at https://github.com/khodabandeh-ali/BoKDiff.git.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 24 of its 39 references that are in the map. First posted on arXiv 26 Jan 2025.

Builds on

24 of 39 references.

References · 24

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  4. 2017Deep reinforcement learning from human preferencescited by 3,846
  5. 2019Fine-Tuning Language Models from Human Preferencescited by 2,124
  6. 2023Training Diffusion Models with Reinforcement Learningcited by 939
  7. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 767
  8. 2023RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignmentcited by 672
  9. 2022Structure-based Drug Design with Equivariant Diffusion Modelscited by 213
  10. 2022Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pocketscited by 182
  11. 20233D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Predictioncited by 176
  12. 2022A 3D Generative Model for Structure-Based Drug Designcited by 148
12 more
  1. 2024BoNBoN Alignment for Large Language Models and the Sweetness of Best-of-n Samplingcited by 120
  2. 2024Theoretical guarantees on the best-of-n alignment policycited by 116
  3. 2022Generating 3D Molecules for Target Protein Bindingcited by 104
  4. 2024DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Designcited by 87
  5. 2024Large-scale Reinforcement Learning for Diffusion Modelscited by 78
  6. 2021Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Modelscited by 72
  7. 2022DiffBP: Generative Diffusion of 3D Molecules for Target Protein Bindingcited by 58
  8. 2022Reinforced Genetic Algorithm for Structure-based Drug Designcited by 46
  9. 2024Antigen-Specific Antibody Design via Direct Energy-based Preference Optimizationcited by 30
  10. 2024DecompOpt: Controllable and Decomposed Diffusion Models for Structure-based Molecular Optimizationcited by 27
  11. 2023Learning Subpocket Prototypes for Generalizable Structure-based Drug Designcited by 25
  12. 2024Fair Bilevel Neural Network (FairBiNN): On Balancing fairness and accuracy via Stackelberg Equilibriumcited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 1.

3 papers.

  1. 2025Uncertainty-Aware Multi-Objective Reinforcement Learning-Guided Diffusion Models for 3D De Novo Molecular Designcited by 2
  2. 2025Diffusion Models at the Drug Discovery Frontier: A Review on Generating Small Molecules versus Therapeutic Peptidescited by 0
  3. 2026A Systematic Evaluation of Molecule Generation Models for De Novo Drug Design: From Benchmarks to Practical Insightscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Aligning Target-Aware Molecule Diffusion Models with Exact Energy Optimization
  2. 2022DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
  3. 2024AUTODIFF: Autoregressive Diffusion Modeling for Structure-based Drug Design
  4. 2025Molecule Generation for Target Protein Binding with Hierarchical Consistency Diffusion Model
  5. 2026Generating Developable 3D Molecules via Pocket-Conditioned Diffusion and Property-Aware Optimization
  6. 2025READ: A Retrieval-Alignment Diffusion Framework for Structure-based Drug Design
  7. 2024Bridging the Gap between Learning and Inference for Diffusion-Based Molecule Generation
  8. 2024Diffusion Models in textitDe Novo Drug Design