Research map / Probabilistic and scientific modeling / Molecules and proteins
Protein interface and docking prediction
Predicts interfaces, ligand docking and conformations with equivariant 3D structure networks.
72 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2019–2026) · an approach in Molecules and proteins, keyword cluster “proteins · binding · sequence”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
18 papers in 2023–24 and 47 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 3.7% to 5.1%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2019 · End-to-End Learning on 3D Protein Structure for Interface Prediction (NeurIPS 2019 · 41 citations)
- 2020 · CASTER: Predicting Drug Interactions with Chemical Substructure Representation (AAAI 2020 · 153 citations)
- 2021 · Fast End-to-End Learning on Protein Surfaces (CVPR 2021 · 86 citations)
- 2022 · TANKBind: Trigonometry-Aware Neural NetworKs for Drug-Protein Binding Structure Prediction (NeurIPS 2022 · 171 citations)
- 2023 · Rotamer Density Estimator is an Unsupervised Learner of the Effect of Mutations on Protein-Protein Interaction (ICLR 2023 · 29 citations)
- 2024 · PSC-CPI: Multi-Scale Protein Sequence-Structure Contrasting for Efficient and Generalizable Compound-Protein Interaction Prediction (AAAI 2024 · 13 citations)
- 2025 · NeuralPLexer3: Accurate Biomolecular Complex Structure Prediction with Flow Models (NeurIPS 2025 · 8 citations)
- 2026 · SAIR: Enabling Deep Learning for Protein-Ligand Interactions with a Synthetic Structural Dataset (ICLR 2026 · 11 citations)
Earliest papers here
- End-to-End Learning on 3D Protein Structure for Interface Prediction (NeurIPS 2019 · 41 citations)
- CASTER: Predicting Drug Interactions with Chemical Substructure Representation (AAAI 2020 · 153 citations)
- Energy-based models for atomic-resolution protein conformations (ICLR 2020 · 13 citations)
What has been tried
36 papers here report 360 measured changes (a component added, removed or swapped, compared with a baseline): helped 77%, no clear effect 8%, hurt 14%, mixed 2%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add intra-modality contrastive learning | Karimi, four test splits · AUPRC | 25.36 vs 25.42; 24.47 vs 24.67; 10.46 vs 11.03; 11.30 vs 11.65 | Helped | PSC-CPI: Multi-Scale Protein Sequence-Structure Contrasting for Efficient and Generalizable Compound-Protein Interaction Prediction (AAAI 2024 · 13 citations) |
| Add weighted interface-based training data sampling and cropping | PoseBusters · accuracy | Helped | NeuralPLexer3: Accurate Biomolecular Complex Structure Prediction with Flow Models (NeurIPS 2025 · 8 citations) | |
| Replace the harmonic prior with an isotropic Gaussian prior | PDBBind sequence similarity, Radius-Pocket · %<2 RMSD | 20.5 -> 17.0 | Hurt | Harmonic Self-Conditioned Flow Matching for joint Multi-Ligand Docking and Binding Site Design (ICML 2024 · 7 citations) |
| Add motion alignment | ATLAS (R2) · Cα-RMSE | 1.40 -> 1.30 | Helped | 4D Diffusion for Dynamic Protein Structure Prediction with Reference and Motion Guidance (AAAI 2025 · 6 citations) |
| Use RDE-predicted side chains instead of Rosetta-packed side chains | PPDBench peptide generation · IMP%-S | Helped | PPFLOW: Target-Aware Peptide Design with Torsional Flow Matching (ICML 2024 · 6 citations) | |
| add sidechain torsion sampling at inference | PDBBind test, flexible redocking · %<2 Top-1 RMSD | 39.8 -> 53.9 | Helped | Re-Dock: Towards Flexible and Realistic Molecular Docking with Diffusion Bridge (ICML 2024 · 6 citations) |
| Add L_fit | DB5.5 · IRMSD median | 13.02 -> 11.13 | Helped | Rigid Protein-Protein Docking via Equivariant Elliptic-Paraboloid Interface Prediction (ICLR 2024 · 5 citations) |
| add data augmentation | CrossDocked · Vina Dock (Avg.) | -8.76 -> -9.12 | Helped | FlexSBDD: Structure-Based Drug Design with Flexible Protein Modeling (NeurIPS 2024 · 5 citations) |
| add folding-stability fine-tuning | SKEMPIv2.0 test split, overall · Spearman | 0.43 -> 0.45 | Mixed | Predicting mutational effects on protein binding from folding energy (ICML 2025 · 2 citations) |
| remove CPLA; use DSDP-predicted pockets | PDBbind, RMSD <2Å · success rate (%) | 47.4 -> 41.2 | Hurt | DeltaDock: A Unified Framework for Accurate, Efficient, and Physically Reliable Molecular Docking (NeurIPS 2024 · 2 citations) |
| use task-aware readout instead of sum readout | Protein-MT EC · Fmax | 0.722 -> 0.810 | Helped | HeMeNet: Heterogeneous Multichannel Equivariant Network for Protein Multi-task Learning (AAAI 2025 · 1 citation) |
| add energy frustration and 3D structure to the enzyme encoder | Supervised, 69 enzyme families · PR-AUC (%) | 75.3±1.9 -> 79.3±1.2 | Helped | UniZyme: A Unified Protein Cleavage Site Predictor Enhanced with Enzyme Active-Site Knowledge (NeurIPS 2025 · 1 citation) |
What reviewers will expect
Datasets and baselines named in the 31 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- PDBBind v2020 · 3 of 31 papers
- ATLAS · 2 of 31 papers
- UniProt · 2 of 31 papers
- AlphaFoldDB · 2 of 31 papers
Baselines
- TankBind · 4 of 31 papers
- DiffDock · 3 of 31 papers
- AlphaFold 3 · 2 of 31 papers
- Vina · 2 of 31 papers
- EquiBind · 2 of 31 papers
- Gnina · 2 of 31 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 65 papers; from OpenAlex and the papers' affiliations).
- Tsinghua University · 8 papers
- University of Science and Technology of China · 5 papers
- Westlake University · 4 papers
- University of Chinese Academy of Sciences · 4 papers
- Zhejiang University · 3 papers
- Renmin University of China · 3 papers
- Hong Kong University of Science and Technology · 3 papers
- Nanjing University · 2 papers
- University of Illinois, Urbana Champaign · 2 papers
- Mila · 2 papers
Use it
Download BibTeX of 32 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Generative protein structure design (142 papers)
- Protein fitness and variant effect prediction (71 papers)
- Protein representation learning (40 papers)
- Biological sequence transformers (34 papers)
Elsewhere in the topic
- Prediction-powered and conformal inference (129 papers)
- Statistical analysis of ML practice (122 papers)
- Neuro-symbolic probabilistic layers (120 papers)
- Structure-based 3D molecule generation (107 papers)
- Sparse graphical and tensor models (97 papers)
- Flow matching generative modeling (97 papers)
- Machine learning for physical simulation (96 papers)
- Multimodal trajectory prediction (95 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).