mlsanityData to 7 Oct 2026

Multi-Scale Representation Learning on Proteins

arXiv 2204.02337 ↗First posted 4 Apr 2022cs.LG cs.AI q-bio.BMcited by 52

Not on the map. Why?

Abstract (arXiv)

Proteins are fundamental biological entities mediating key roles in cellular function and disease. This paper introduces a multi-scale graph construction of a protein -- HoloProt -- connecting surface to structure and sequence. The surface captures coarser details of the protein, while sequence as primary component and structure -- comprising secondary and tertiary components -- capture finer details. Our graph encoder then learns a multi-scale representation by allowing each level to integrate the encoding from level(s) below with the graph at that level. We test the learned representation on different tasks, (i.) ligand binding affinity (regression), and (ii.) protein function prediction (classification). On the regression task, contrary to previous methods, our model performs consistently and reliably across different dataset splits, outperforming all baselines on most splits. On the classification task, it achieves a performance close to the top-performing model while using 10x fewer parameters. To improve the memory efficiency of our construction, we segment the multiplex protein surface manifold into molecular superpixels and substitute the surface with these superpixels at little to no performance loss.

Key facts. Cited by 52 papers in the map (data through 7 Oct 2026). Builds on 16 of its 69 references that are in the map. First posted on arXiv 4 Apr 2022.

Builds on

16 of 69 references.

References · 16

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  3. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  4. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  5. 2019Evaluating Protein Transfer Learning with TAPEcited by 229
  6. 2020Hierarchical Generation of Molecular Graphs using Structural Motifscited by 204
  7. 2018DeepDTA: Deep Drug-Target Binding Affinity Predictioncited by 171
  8. 2020ProtTrans: Towards Cracking the Language of Life's Code Through Self-Supervised Deep Learning and High Performance Computingcited by 162
  9. 2020ATOM3D: Tasks On Molecules in Three Dimensionscited by 73
  10. 2016Protein-Ligand Scoring with Convolutional Neural Networkscited by 57
  11. 2016Superpixels: An Evaluation of the State-of-the-Artcited by 52
  12. 2019Learning protein sequence embeddings using information from structurecited by 50
4 more
  1. 2017DeepSF: deep convolutional neural network for mapping protein sequences to foldscited by 46
  2. 2018End-to-End Learning on 3D Protein Structure for Interface Predictioncited by 32
  3. 2017Deriving Neural Architectures from Sequence and Graph Kernelscited by 23
  4. 2019Improving detection of protein-ligand binding sites with 3D segmentationcited by 15

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 12; 2023: 17; 2024: 9; 2025: 6; 2026: 8.

52 papers.

Sort:

By area:

  1. 2022DiffDock: Diffusion Steps, Twists, and Turns for Molecular Dockingcited by 307
  2. 20233D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Predictioncited by 176
  3. 2022Learning Hierarchical Protein Representations via Complete 3D Graph Networkscited by 54
  4. 2022Multimodal learning with graphscited by 54
  5. 2022Structure-based drug design with geometric deep learningcited by 40
  6. 2022Structure-aware Protein Self-supervised Learningcited by 33
  7. 2022Contrastive Representation Learning for 3D Protein Structurescited by 30
  8. 2023DiffPack: A Torsional Diffusion Model for Autoregressive Protein Side-Chain Packingcited by 24
  9. 2023Unsupervised Protein-Ligand Binding Energy Prediction via Neural Euler's Rotation Equationcited by 18
  10. 2023Generalist Equivariant Transformer Towards 3D Molecular Interaction Learningcited by 18
  11. 2024Evaluating representation learning on the protein structure universecited by 16
  12. 2022Antibody-Antigen Docking and Design via Hierarchical Equivariant Refinementcited by 15
38 more
  1. 2022Geometry-Complete Perceptron Networks for 3D Molecular Graphscited by 15
  2. 2022Integration of Pre-trained Protein Language Models into Geometric Deep Learning Networkscited by 14
  3. 2024Pre-training with Fractional Denoising to Enhance Molecular Property Predictioncited by 14
  4. 2023Pre-Training Protein Encoder via Siamese Sequence-Structure Diffusion Trajectory Predictioncited by 13
  5. 2023Causal Optimal Transport of Abstractionscited by 13
  6. 2023ProFSA: Self-supervised Pocket Pretraining via Protein Fragment-Surroundings Alignmentcited by 12
  7. 2023Protein-ligand binding representation learning from fine-grained interactionscited by 12
  8. 2023Learning Harmonic Molecular Representations on Riemannian Manifoldcited by 8
  9. 2023Knowledge-augmented Graph Machine Learning for Drug Discovery: A Surveycited by 6
  10. 2023Bi-level Contrastive Learning for Knowledge-Enhanced Molecule Representationscited by 6
  11. 2023AtomSurf : Surface Representation for Learning on Protein Structurescited by 5
  12. 2024An Equivariant Pretrained Transformer for Unified 3D Molecular Representation Learningcited by 5
  13. 2024A Cross-Field Fusion Strategy for Drug-Target Interaction Predictioncited by 5
  14. 2025GLProtein: Global-and-Local Structure Aware Protein Representation Learningcited by 5
  15. 2023Modeling Molecular Structures with Intrinsic Diffusion Modelscited by 4
  16. 2022EGR: Equivariant Graph Refinement and Assessment of 3D Protein Complex Structurescited by 3
  17. 2022Holographic-(V)AE: an end-to-end SO(3)-Equivariant (Variational) Autoencoder in Fourier Spacecited by 3
  18. 2022Structure-based drug discovery with deep learningcited by 3
  19. 2024Binding Affinity Prediction: From Conventional to Machine Learning-Based Approachescited by 3
  20. 2026Surface-based Molecular Design with Multi-modal Flow Matchingcited by 3
  21. 2026Self-Conditioned Denoising for Atomistic Representation Learningcited by 3
  22. 2023Geometric Deep Learning for Structure-Based Drug Design: A Surveycited by 2
  23. 2025Tokenizing Electron Cloud in Protein-Ligand Interaction Learningcited by 2
  24. 2023Graph Representation Learning for Interactive Biomolecule Systemscited by 1
  25. 2024QuickBind: A Light-Weight And Interpretable Molecular Docking Modelcited by 1
  26. 2025Bidirectional Hierarchical Protein Multi-Modal Representation Learningcited by 1
  27. 2023Mixup-Augmented Meta-Learning for Sample-Efficient Fine-Tuning of Protein Simulatorscited by 0
  28. 2023Target-Free Compound Activity Prediction via Few-Shot Learningcited by 0
  29. 2024Exploiting Hierarchical Interactions for Protein Surface Learningcited by 0
  30. 2024Geometric Self-Supervised Pretraining on 3D Protein Structures using Subgraphscited by 0
  31. 2024Advanced atom-level representations for protein flexibility prediction utilizing graph neural networkscited by 0
  32. 2025Demystify Protein Generation with Hierarchical Conditional Diffusion Modelscited by 0
  33. 2025Topological Feature Compression for Molecular Graph Neural Networkscited by 0
  34. 2025Flow-Based Fragment Identification via Binding Site-Specific Latent Representationscited by 0
  35. 2026Towards Multiscale Graph-based Protein Learning with Geometric Secondary Structural Motifscited by 0
  36. 2026GDEGAN: Gaussian Dynamic Equivariant Graph Attention Network for Ligand Binding Site Predictioncited by 0
  37. 2026h-MINT: Modeling Pocket-Ligand Binding with Hierarchical Molecular Interaction Networkcited by 0
  38. 2026SurfDesign: Effective Protein Design on Molecular Surfacescited by 0

50 most cited of 52 All 52 (JSON)