Research map / Graph learning / Graph self-supervised learning
GATCL: An Adaptive Contrastive Learning Framework Based on MHGAT for Spatial Domain Identification in Spatial Transcriptomics
Shilin Zhang, Weiliang Huo, Qingchen Zhang, Xiulong Liu
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 multi-head graph attention encoder instead of GCN (ablation) | GATCL · DLPFC (average ARI) | ARI: 0.5414 -> 0.5746 | Improved |
| add attention pooling instead of average pooling (ablation) | GATCL · DLPFC (average ARI) | ARI: 0.5548 -> 0.5746 | Improved |
| add adaptive graph contrastive learning module (ablation) | GATCL · DLPFC (average ARI) | ARI: 0.5123 -> 0.5746 | Improved |
| add MHGAT instead of GCN (ablation) | GATCL · Breast cancer | ARI: 0.5314 -> 0.6182 | Improved |
| add attention pooling instead of average pooling (ablation) | GATCL · Breast cancer | ARI: 0.5791 -> 0.6182 | Improved |
| add adaptive graph contrastive learning module (ablation) | GATCL · Breast cancer | ARI: 0.5201 -> 0.6182 | Improved |
| add MHGAT instead of GCN (ablation) | GATCL · Mouse somatosensory cortex | ARI: 0.5440 -> 0.6496 | Improved |
| add attention pooling instead of average pooling (ablation) | GATCL · Mouse somatosensory cortex | ARI: 0.5896 -> 0.6496 | Improved |
| add adaptive graph contrastive learning module (ablation) | GATCL · Mouse somatosensory cortex | ARI: 0.5136 -> 0.6496 | Improved |
| use GATCL instead of STAGATE | spatial domain identification · DLPFC (average ARI) | ARI: 0.5151 -> 0.5746 | Improved |
| use GATCL instead of baselines on individual DLPFC samples | spatial domain identification · DLPFC samples 151507, 151670 | ARI: 151507: 0.5468 (STAGATE) vs 0.4363; 151670: 0.5203 (GraphST) vs 0.4456 | Mixed |
| add adaptive graph contrastive learning on sample 151670 (ablation) | GATCL · DLPFC sample 151670 | ARI: 0.4565 -> 0.4456 | No clear effect |
Closest papers by meaning
- Global Context-aware Representation Learning for Spatially Resolved Transcriptomics (ICML 2025 · similarity 0.88)
- DUSTED: Dual-Attention Enhanced Spatial Transcriptomics Denoiser (AAAI 2025 · similarity 0.87)
- MERGE: Multi-faceted Hierarchical Graph-based GNN for Gene Expression Prediction from Whole Slide Histopathology Images (CVPR 2025 · similarity 0.87)
- Multi-modal Topology-embedded Graph Learning for Spatially Resolved Genes Prediction from Pathology Images with Prior Gene Similarity Information (CVPR 2025 · similarity 0.87)
- SUICA: Learning Super-high Dimensional Sparse Implicit Neural Representations for Spatial Transcriptomics (ICML 2025 · similarity 0.87)
- Preserving DEG Rankings for Gene Discovery in Histology-Based Spatial Gene Expression Prediction (NeurIPS 2026 · similarity 0.86)
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.
