Research map / Graph learning / Graph self-supervised learning
Context-Aware Safe Medication Recommendations with Molecular Graph and DDI Graph Embedding
Qianyu Chen, Xin Li, Kunnan Geng, Mingzhong Wang
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 medication context from diagnosis, procedure, and co-occurrence information to molecular GNN encoding (ablation) | Carmen medication representations · MIMIC-III | Jaccard (%): 51.20±00.14 -> 53.13±00.11 | Improved |
| Add medication context from diagnosis, procedure, and co-occurrence information to molecular GNN encoding (ablation) | Carmen medication representations · MIMIC-IV | Jaccard (%): 48.74±00.27 -> 50.27±00.09 | Improved |
| Add non-transitive DDI message-dispelling encoding (ablation) | Carmen · MIMIC-III | Jaccard (%): 53.13±00.11 -> 52.67±00.21 | Hurt |
| Add non-transitive DDI message-dispelling encoding (ablation) | Carmen · MIMIC-IV | Jaccard (%): 50.27±00.09 -> 50.06±00.12 | Hurt |
| Use message dispelling instead of conventional message passing for the DDI graph (ablation) | Carmen · MIMIC-III | Jaccard (%): 53.15±00.09 -> 52.67±00.21 | No clear effect |
| Use message dispelling instead of conventional message passing for the DDI graph (ablation) | Carmen · MIMIC-IV | Jaccard (%): 50.33±00.11 -> 50.06±00.12 | No clear effect |
| Use DDI message-dispelling encoding instead of an additional DDI loss (ablation) | Carmen · MIMIC-III | Jaccard (%): 53.23±00.18 -> 52.67±00.21 | Hurt |
| Use DDI message-dispelling encoding instead of an additional DDI loss (ablation) | Carmen · MIMIC-IV | Jaccard (%): 50.49±00.09 -> 50.06±00.12 | Hurt |
| Apply Carmen's DDI encoding rather than no DDI encoding (ablation) | Carmen recommendations · MIMIC-III and MIMIC-IV test recommendations | unsafe drug-pair occurrences over 10 runs | Improved |
| Apply a DDI loss instead of DDI encoding (ablation) | Carmen recommendations · MIMIC-III and MIMIC-IV test recommendations | unsafe drug-pair occurrences over 10 runs | Mixed |
| Use Carmen rather than GAMENet | Medication recommendation · MIMIC-III | Jaccard (%): 51.59±00.09 -> 52.67±00.21 | Improved |
| Use Carmen rather than SafeDrug | Medication recommendation · MIMIC-IV | Jaccard (%): 48.57±00.62 -> 50.06±00.12 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GAMENet — GAMENet: Graph Augmented MEmory Networks for Recommending Medication Combination (AAAI 2019) · better on 2 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Dual-Channel Learning Framework for Drug-Drug Interaction Prediction via Relation-Aware Heterogeneous Graph Transformer (AAAI 2024 · similarity 0.90)
- Biomedical Knowledge Graph Embedding with Householder Projection (Student Abstract) (AAAI 2024 · similarity 0.88)
- Customized Subgraph Selection and Encoding for Drug-drug Interaction Prediction (NeurIPS 2024 · similarity 0.88)
- Motif-Oriented Representation Learning with Topology Refinement for Drug-Drug Interaction Prediction (AAAI 2025 · similarity 0.87)
- Knowledge Enhanced Representation Learning for Drug Discovery (AAAI 2024 · similarity 0.87)
- On the Scalability of GNNs for Molecular Graphs (NeurIPS 2024 · similarity 0.87)
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.
