Research map / Probabilistic and scientific modeling / Molecules and proteins
ATOM3D: Tasks on Molecules in Three Dimensions
Raphael John Lamarre Townshend, Martin Vögele, Patricia Adriana Suriana, Alexander Derry, Alexander Powers, Yianni Laloudakis, Sidhika Balachandar, Bowen Jing, Brandon M. Anderson, Stephan Eismann, Risi Kondor, Russ Altman and 1 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add 3D coordinates to the 2D molecular representation | ENN · SMP, μ | MAE: 0.496 -> 0.052 | Improved |
| add 3D coordinates to the 2D molecular representation | ENN · SMP, ε_gap | MAE: 0.154 -> 0.095 | Improved |
| add 3D coordinates to the 2D molecular representation | ENN · SMP, U₀ᵃᵗ | MAE: 0.182 -> 0.025 | Improved |
| use 3D atomistic geometry instead of sequence-only inputs | PIP · DB5 | AUROC: 0.841 -> 0.844 | No clear effect |
| use 3D atomistic geometry instead of protein sequence | RES · RES | accuracy: 0.30 -> 0.451 | Improved |
| use 3D atomistic geometry instead of protein sequence | GNN · MSP | AUROC: 0.554 -> 0.609 | Improved |
| use 3D atomistic geometry rather than a 1D joint representation | 3DCNN · LBA | RMSE: 1.565 -> 1.416 | Improved |
| use 3D atomistic geometry rather than a 1D joint representation | 3DCNN · LBA | glob. R_P: 0.573 -> 0.550 | Hurt |
| use 3D atomistic geometry rather than a 1D joint representation | GNN · LEP | AUROC: 0.696 -> 0.681 | Hurt |
| select a 3DCNN rather than a GNN | PIP · DB5 | AUROC: 0.669 -> 0.844 | Improved |
| select a 3DCNN rather than a GNN | RES · RES | accuracy: 0.082 -> 0.451 | Improved |
| select a GNN rather than a 3DCNN | MSP · MSP | AUROC: 0.574 -> 0.609 | Improved |
Closest papers by meaning
- Enhancing Molecular Representation Learning Through the Combination of 3D and 2D Graph Machine Learning (Student Abstract) (AAAI 2025 · similarity 0.89)
- Beyond Atoms: Enhancing Molecular Pretrained Representations with 3D Space Modeling (ICML 2025 · similarity 0.89)
- Empowering Active Learning for 3D Molecular Graphs with Geometric Graph Isomorphism (NeurIPS 2024 · similarity 0.88)
- AtomSurf: Surface Representation for Learning on Protein Structures (ICLR 2025 · similarity 0.88)
- Score-based 3D molecule generation with neural fields (NeurIPS 2024 · similarity 0.88)
- M²N: A Progressive Macro-to-Micro 3D Modeling Scheme for Unveiling Drug-Target Affinity (AAAI 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
