mlsanityData to 14 Oct 2026

3DMolFormer: A Dual-channel Framework for Structure-based Drug Discovery

arXiv 2502.05107 ↗First posted 7 Feb 2025cs.CE cs.LGcited by 4

Not on the map. Why?

Abstract (arXiv)

Structure-based drug discovery, encompassing the tasks of protein-ligand docking and pocket-aware 3D drug design, represents a core challenge in drug discovery. However, no existing work can deal with both tasks to effectively leverage the duality between them, and current methods for each task are hindered by challenges in modeling 3D information and the limitations of available data. To address these issues, we propose 3DMolFormer, a unified dual-channel transformer-based framework applicable to both docking and 3D drug design tasks, which exploits their duality by utilizing docking functionalities within the drug design process. Specifically, we represent 3D pocket-ligand complexes using parallel sequences of discrete tokens and continuous numbers, and we design a corresponding dual-channel transformer model to handle this format, thereby overcoming the challenges of 3D information modeling. Additionally, we alleviate data limitations through large-scale pre-training on a mixed dataset, followed by supervised and reinforcement learning fine-tuning techniques respectively tailored for the two tasks. Experimental results demonstrate that 3DMolFormer outperforms previous approaches in both protein-ligand docking and pocket-aware 3D drug design, highlighting its promising application in structure-based drug discovery. The code is available at: https://github.com/HXYfighter/3DMolFormer .

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 47 of its 94 references that are in the map. First posted on arXiv 7 Feb 2025.

Builds on

47 of 94 references.

References · 47

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2016On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minimacited by 1,352
  4. 2018Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generationcited by 364
  5. 2022DiffDock: Diffusion Steps, Twists, and Turns for Molecular Dockingcited by 313
  6. 2017Molecular De Novo Design through Deep Reinforcement Learningcited by 267
  7. 2018Molecular Transformer - A Model for Uncertainty-Calibrated Chemical Reaction Predictioncited by 229
  8. 2022Structure-based Drug Design with Equivariant Diffusion Modelscited by 217
  9. 2021ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Predictioncited by 204
  10. 2022Pocket2Mol: Efficient Molecular Sampling Based on 3D Protein Pocketscited by 183
  11. 20233D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Predictioncited by 177
  12. 2021Motif-based Graph Self-Supervised Learning for Molecular Property Predictioncited by 159
35 more
  1. 2022A 3D Generative Model for Structure-Based Drug Designcited by 149
  2. 2022EquiBind: Geometric Deep Learning for Drug Binding Structure Predictioncited by 144
  3. 2021Geometric Deep Learning on Molecular Representationscited by 113
  4. 2022Generating 3D Molecules for Target Protein Bindingcited by 104
  5. 2024DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Designcited by 88
  6. 2021Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Modelscited by 73
  7. 2021Hit and Lead Discovery with Explorative RL and Fragment-based Molecule Generationcited by 68
  8. 2022MolGenSurvey: A Systematic Survey in Machine Learning Models for Molecule Designcited by 66
  9. 2023Language models can generate molecules, materials, and protein binding sites directly in three dimensions as XYZ, CIF, and PDB filescited by 59
  10. 2020Reinforcement Learning for Molecular Design Guided by Quantum Mechanicscited by 51
  11. 2020Guiding Deep Molecular Optimization with Genetic Explorationcited by 51
  12. 2022Reinforced Genetic Algorithm for Structure-based Drug Designcited by 46
  13. 2022Unified 2D and 3D Pre-Training of Molecular Representationscited by 45
  14. 2021Property-Aware Relation Networks for Few-Shot Molecular Property Predictioncited by 41
  15. 2022LIMO: Latent Inceptionism for Targeted Molecule Generationcited by 40
  16. 2022Structure-based drug design with geometric deep learningcited by 40
  17. 2022E3Bind: An End-to-End Equivariant Network for Protein-Ligand Dockingcited by 35
  18. 2022A Systematic Survey of Chemical Pre-trained Modelscited by 34
  19. 2022Structure-aware Protein Self-supervised Learningcited by 33
  20. 2024Deep Confident Steps to New Pockets: Strategies for Docking Generalizationcited by 27
  21. 2022Zero-Shot 3D Drug Design by Sketching and Generatingcited by 24
  22. 2021Deep Molecular Representation Learning via Fusing Physical and Chemical Informationcited by 21
  23. 2022The Role of AI in Drug Discovery: Challenges, Opportunities, and Strategiescited by 20
  24. 2023Drug Discovery with Dynamic Goal-aware Fragmentscited by 19
  25. 2023De novo Drug Design using Reinforcement Learning with Multiple GPT Agentscited by 19
  26. 2023FABind: Fast and Accurate Protein-Ligand Bindingcited by 17
  27. 2023Protein-ligand binding representation learning from fine-grained interactionscited by 14
  28. 2023Generation of 3D Molecules in Pockets via Language Modelcited by 13
  29. 2024BindGPT: A Scalable Framework for 3D Molecular Design via Language Modeling and Reinforcement Learningcited by 12
  30. 2024Uni-Mol Docking V2: Towards Realistic and Accurate Binding Pose Predictioncited by 10
  31. 2024Rethinking Specificity in SBDD: Leveraging Delta Score and Energy-Guided Diffusioncited by 9
  32. 2023Do Deep Learning Methods Really Perform Better in Molecular Conformation Generation?cited by 8
  33. 2023Utilizing Reinforcement Learning for de novo Drug Designcited by 7
  34. 2024Structure-based Drug Design Benchmark: Do 3D Methods Really Dominate?cited by 7
  35. 2024Token-Mol 1.0: Tokenized drug design with large language modelcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 1.

4 papers.

  1. 2025MolChord: Structure-Sequence Alignment for Protein-Guided Drug Designcited by 2
  2. 2025QE-Catalytic: A Graph-Language Multimodal Base Model for Relaxed-Energy Prediction in Catalytic Adsorptioncited by 1
  3. 2025TransDiffSBDD: Causality-Aware Multi-Modal Structure-Based Drug Designcited by 0
  4. 2026Fine-tuning Pocket-Aware Diffusion Models via Denoising Policy Optimizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond Affinity: A Benchmark of 1D, 2D, and 3D Methods Reveals Critical Trade-offs in Structure-Based Drug Design
  2. 2025FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction
  3. 2025Apo2Mol: 3D Molecule Generation via Dynamic Pocket-Aware Diffusion Models
  4. 2024ShEPhERD: Diffusing shape, electrostatics, and pharmacophores for bioisosteric drug design
  5. 2026Generating Developable 3D Molecules via Pocket-Conditioned Diffusion and Property-Aware Optimization
  6. 2026Beyond SBDD: Geometric Deep Learning in Polypharmacology and Multi-target Drug Design
  7. 2026SynLaD: Latent Diffusion for Generating Synthesizable Molecules Conditioned on 3D Pharmacophore Profiles
  8. 2026FuseDiff: Symmetry-Preserving Joint Diffusion for Dual-Target Structure-Based Drug Design