Research map / Graph learning / Graph self-supervised learning
Supervised Graph Contrastive Learning for Gene Regulatory Networks
Sho Oshima, Yuji Okamoto, Taisei Tosaki, Ryosuke Kojima
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 knockdown-experiment teacher GRNs as supervision (SupGCL) versus no pretraining | Graph Transformer · Hazard prediction (breast) | C-index: 0.601 -> 0.650 | Improved |
| add supervised augmentation term with tau_a=0.25 instead of tau_a=infinity (ablation) | SupGCL · Subtype classification (breast) | accuracy: 0.832 -> 0.847 | Improved |
| add supervised augmentation term with tau_a=0.25 instead of tau_a=infinity (ablation) | SupGCL · Cancer relation Rel (breast) | accuracy: 0.573 -> 0.600 | Improved |
| add supervised augmentation term with tau_a=0.25 instead of tau_a=infinity (ablation) | SupGCL · CC (breast) | subset accuracy: 0.259 -> 0.291 | Improved |
| lower tau_a to 0.10 instead of default 0.25 (ablation) | SupGCL · Hazard (breast) | C-index: 0.650 -> 0.670 | No clear effect |
| tau_a infinity equals unsupervised node-level GCL (ablation) | SupGCL · breast GRN tasks | downstream performance vs GRACE | No clear effect |
| use existing graph representation learning pretraining (GAE, GraphCL, GRACE, SGRL) for node-level tasks | Graph Transformer · BP, CC, Rel node-level tasks | accuracy | No clear effect |
| pretrain with GraphCL | Graph Transformer · Subtype classification (breast) | accuracy: 0.804 -> 0.719 | Hurt |
| replace Graph Transformer with GAT encoder in GRACE (ablation) | GRACE · breast GRN hazard and BP | C-index / subset accuracy: Hazard 0.642 -> 0.632; BP 0.230 -> 0.241 | Mixed |
| SupGCL embeddings versus GAE, GraphCL, GRACE, SGRL | breast GRN embeddings · disease subtype clustering | NMI and ARI | Improved |
Closest papers by meaning
- SOLA-GCL: Subgraph-Oriented Learnable Augmentation Method for Graph Contrastive Learning (AAAI 2025 · similarity 0.90)
- Does GCL Need a Large Number of Negative Samples? Enhancing Graph Contrastive Learning with Effective and Efficient Negative Sampling (AAAI 2025 · similarity 0.89)
- Enhancing Graph Contrastive Learning for Protein Graphs from Perspective of Invariance (ICML 2025 · similarity 0.88)
- TopoGCL: Topological Graph Contrastive Learning (AAAI 2024 · similarity 0.87)
- ReGCL: Rethinking Message Passing in Graph Contrastive Learning (AAAI 2024 · similarity 0.87)
- Beyond Homophily: Graph Contrastive Learning with Macro-Micro Message Passing (AAAI 2025 · 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.
