Research map / Architectures and efficiency / Attention and transformers
HEXST: Hexagonal Shifted-Window Transformer for Spatial Transcriptomics Gene Expression Prediction
Keunho Byeon, Jin Tae Kwak
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 |
|---|---|---|---|
| replace square window with hexagonal window (2D RoPE kept) (ablation) | HEXST · SpaRED (7 datasets) | PCC-F: 0.3690 -> 0.3993 | Improved |
| replace Cartesian 2D RoPE with HexRoPE (ablation) | HEXST · SpaRED (7 datasets) | PCC-F: 0.3993 -> 0.4227 | Improved |
| add Pearson correlation loss (ablation) | HEXST · SpaRED (7 datasets) | PCC-F: 0.3541 -> 0.4227 | Improved |
| add MSE loss (ablation) | HEXST · SpaRED (7 datasets) | PCC-S: 0.1194 -> 0.7697 | Improved |
| add transcriptomic feature alignment loss (ablation) | HEXST · SpaRED (7 datasets) | MI-F / AUC 0vNZ: PCC-F 0.3818 -> 0.4227; MI-F 0.1615 -> 0.1540; AUC0vNZ 0.6933 -> 0.6820 | Mixed |
| add deviation-matching loss (ablation) | HEXST · SpaRED (7 datasets) | MI-F / AUC 0vNZ: PCC-F 0.3796 -> 0.4227; MI-F 0.1611 -> 0.1540; AUC0vNZ 0.6892 -> 0.6820 | Mixed |
| add both deviation-matching and transcriptomic alignment losses (ablation) | HEXST · SpaRED (7 datasets) | PCC-S: 0.6962 -> 0.7697 | Improved |
| use HEXST instead of PEKA for gene expression prediction | spatial gene expression prediction · SpaRED (7 datasets) | PCC-F: 0.3863 -> 0.4227 | Improved |
| add HEXST-predicted transcriptomics-guided spot representations to image features | MIL slide classifier · TCGA-PRAD Gleason grading | accuracy (%): 45.73 -> 48.55 | Improved |
| add HEXST-predicted transcriptomics-guided spot representations to image features | MIL survival model · TCGA-LUAD overall survival | C-index: 0.5845 -> 0.5970 | Improved |
Closest papers by meaning
- HEST-1k: A Dataset For Spatial Transcriptomics and Histology Image Analysis (NeurIPS 2024 · similarity 0.91)
- Diffusion Generative Modeling for Spatially Resolved Gene Expression Inference from Histology Images (ICLR 2025 · similarity 0.90)
- Preserving DEG Rankings for Gene Discovery in Histology-Based Spatial Gene Expression Prediction (NeurIPS 2026 · similarity 0.89)
- Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching (ICML 2025 · similarity 0.88)
- Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images (NeurIPS 2026 · similarity 0.87)
- MERGE: Multi-faceted Hierarchical Graph-based GNN for Gene Expression Prediction from Whole Slide Histopathology Images (CVPR 2025 · 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.
