Research map / Graph learning / Knowledge and temporal graphs
Extracting Biomolecular Interactions Using Semantic Parsing of Biomedical Text
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use AMR rather than SDG parses for sentence-level inference | Biomolecular interaction extraction · PubMed45-ERN, PubMed45, and AIMed | F1: 0.25±0.16 -> 0.33±0.16; 0.32±0.18 -> 0.45±0.25; 0.27±0.12 -> 0.39±0.05 | Helped |
| Use AMR GDK instead of AMR sentence-level inference | AMR-based extraction · PubMed45-ERN, PubMed45, and AIMed | F1: 0.33±0.16 -> 0.35±0.16; 0.45±0.25 -> 0.51±0.23; 0.39±0.05 -> 0.51±0.11 | Helped |
| Use SDG GDK instead of SDG sentence-level inference | SDG-based extraction · PubMed45-ERN, PubMed45, and AIMed | F1: 0.25±0.16 -> 0.25±0.16; 0.32±0.18 -> 0.38±0.15; 0.27±0.12 -> 0.47±0.08 | Mixed |
| Use GDK rather than MSI for AMR-based document inference | AMR-based extraction · PubMed45-ERN, PubMed45, and AIMed | F1: 0.32±0.14 -> 0.35±0.16; 0.45±0.24 -> 0.51±0.23; 0.51±0.11 -> 0.51±0.11 | No clear effect |
| Jointly use AMR and SDG parses with GDK instead of AMR GDK alone | Document-level biomolecular interaction extraction · PubMed45-ERN and PubMed45 | F1: 0.35 to 0.38; 0.51 to 0.57 | Helped |
| Jointly use AMR and SDG parses with GDK instead of AMR GDK alone | Document-level biomolecular interaction extraction · AIMed | F1: 0.51±0.11 -> 0.52±0.09 | No clear effect |
| Use joint AMR-SDG MSI instead of joint AMR-SDG GDK | Joint parse document inference · AIMed | F1: 0.52±0.09 -> 0.55±0.09 | Helped |
| Replace MMD with KL-D in GDK (ablation) | SDG-based extraction · PubMed45-ERN | F1: 0.25±0.16 -> 0.21±0.17 | Hurt |
| Replace MMD with cross kernels in GDK (ablation) | SDG-based extraction · PubMed45-ERN | F1: 0.25±0.16 -> 0.26±0.13 | No clear effect |
| Replace MMD with KL-D in GDK (ablation) | AMR-based extraction · PubMed45-ERN | F1: 0.35±0.16 -> 0.37±0.17 | No clear effect |
| Replace MMD with cross kernels in GDK (ablation) | AMR-based extraction · PubMed45-ERN | F1: 0.35±0.16 -> 0.29±0.13 | Hurt |
Closest papers by meaning
- Knowledge Enhanced Representation Learning for Drug Discovery (AAAI 2024 · similarity 0.84)
- Prot2Text-V2: Protein Function Prediction with Multimodal Contrastive Alignment (NeurIPS 2025 · similarity 0.83)
- BioBridge: Bridging Biomedical Foundation Models via Knowledge Graphs (ICLR 2024 · similarity 0.83)
- Contextualizing biological perturbation experiments through language (ICLR 2025 · similarity 0.83)
- Retrieval-Augmented Language Model for Knowledge-aware Protein Encoding (ICML 2025 · similarity 0.83)
- MMSite: A Multi-modal Framework for the Identification of Active Sites in Proteins (NeurIPS 2024 · similarity 0.83)
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.