mlsanityData to 7 Oct 2026

Improving Molecular Representation Learning with Metric Learning-enhanced Optimal Transport

arXiv 2202.06208 ↗First posted 13 Feb 2022cs.LG cs.AI q-bio.QMcited by 9

Not on the map. Why?

Abstract (arXiv)

Training data are usually limited or heterogeneous in many chemical and biological applications. Existing machine learning models for chemistry and materials science fail to consider generalizing beyond training domains. In this article, we develop a novel optimal transport-based algorithm termed MROT to enhance their generalization capability for molecular regression problems. MROT learns a continuous label of the data by measuring a new metric of domain distances and a posterior variance regularization over the transport plan to bridge the chemical domain gap. Among downstream tasks, we consider basic chemical regression tasks in unsupervised and semi-supervised settings, including chemical property prediction and materials adsorption selection. Extensive experiments show that MROT significantly outperforms state-of-the-art models, showing promising potential in accelerating the discovery of new substances with desired properties.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 38 of its 84 references that are in the map. First posted on arXiv 13 Feb 2022.

Builds on

38 of 84 references.

References · 38

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Domain-Adversarial Training of Neural Networkscited by 3,023
  3. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,362
  4. 2019Invariant Risk Minimizationcited by 1,832
  5. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  6. 2021E(n) Equivariant Graph Neural Networkscited by 951
  7. 2021Generalized Out-of-Distribution Detection: A Surveycited by 796
  8. 2016Automatic chemical design using a data-driven continuous representation of moleculescited by 762
  9. 2017Learning to Generalize: Meta-Learning for Domain Generalizationcited by 762
  10. 2020Out-of-Distribution Generalization via Risk Extrapolation (REx)cited by 737
  11. 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 682
  12. 2015Optimal Transport for Domain Adaptationcited by 457
26 more
  1. 2020Generalized ODIN: Detecting Out-of-distribution Image without Learning from Out-of-distribution Datacited by 408
  2. 2019Domain Generalization via Model-Agnostic Learning of Semantic Featurescited by 405
  3. 2015Return of Frustratingly Easy Domain Adaptationcited by 389
  4. 2019Analyzing Learned Molecular Representations for Property Predictioncited by 373
  5. 2019Semi-supervised Domain Adaptation via Minimax Entropycited by 333
  6. 2017Joint Distribution Optimal Transportation for Domain Adaptationcited by 280
  7. 2019Evaluating Protein Transfer Learning with TAPEcited by 229
  8. 2018A DIRT-T Approach to Unsupervised Domain Adaptationcited by 215
  9. 2018DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptationcited by 212
  10. 2021ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Predictioncited by 204
  11. 2016Correlation Alignment for Unsupervised Domain Adaptationcited by 146
  12. 2020ProGen: Language Modeling for Protein Generationcited by 127
  13. 2021Unbalanced minibatch Optimal Transport; applications to Domain Adaptationcited by 123
  14. 2017Large-Scale Optimal Transport and Mapping Estimationcited by 120
  15. 2018Deep Metric Learning with Hierarchical Triplet Losscited by 112
  16. 2016Theoretical Analysis of Domain Adaptation with Optimal Transportcited by 105
  17. 2021Towards a Theoretical Framework of Out-of-Distribution Generalizationcited by 94
  18. 2021Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Dockingcited by 81
  19. 2020ATOM3D: Tasks On Molecules in Three Dimensionscited by 73
  20. 2021Minibatch optimal transport distances; analysis and applicationscited by 62
  21. 2017DeepSF: deep convolutional neural network for mapping protein sequences to foldscited by 46
  22. 2018PotentialNet for Molecular Property Predictioncited by 46
  23. 2017Deep Learning for Computational Chemistrycited by 38
  24. 2021Reliable Graph Neural Networks for Drug Discovery Under Distributional Shiftcited by 22
  25. 2021Deep Molecular Representation Learning via Fusing Physical and Chemical Informationcited by 21
  26. 2015Generalized conditional gradient: analysis of convergence and applicationscited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 2; 2024: 2; 2025: 2; 2026: 2.

9 papers.

  1. 2022Unbalanced Optimal Transport, from Theory to Numericscited by 45
  2. 2024Wasserstein Wormhole: Scalable Optimal Transport Distance with Transformerscited by 14
  3. 2023A Hierarchical Training Paradigm for Antibody Structure-sequence Co-designcited by 13
  4. 2023Instructor-inspired Machine Learning for Robust Molecular Property Predictioncited by 8
  5. 2025Graph Data Selection for Domain Adaptation: A Model-Free Approachcited by 7
  6. 2025Fast Estimation of Wasserstein Distances via Regression on Sliced Wasserstein Distancescited by 5
  7. 2026Amortized Optimal Transport from Sliced Potentialscited by 2
  8. 2024Linear Partial Gromov-Wasserstein Embeddingcited by 1
  9. 2026SurfDesign: Effective Protein Design on Molecular Surfacescited by 0