Research map / Graph learning / Graph self-supervised learning
SpaEF: Spatially Resolved Transcriptomics Data Element-Wise Denoising Framework Powered by Large Models
Zekuan Shang, Xiaosong Han, Liupu Wang, Wei Du, Peng Zhao, Yuanshu Li, Yubin Xiao, Xuan Wu, You Zhou
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 |
|---|---|---|---|
| use OmiCLIP text encoder for spot features instead of PCA (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.531 -> 0.611 | Improved |
| use OmiCLIP instead of scGPT for spot features (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.569 -> 0.611 | Improved |
| use OmiCLIP instead of Geneformer for spot features (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.496 -> 0.611 | Improved |
| add GenePT relationships to the co-expression gene graph (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.540 -> 0.611 | Improved |
| use GenePT instead of scGPT for gene graph (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.548 -> 0.611 | Improved |
| use GenePT instead of Geneformer for gene graph (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.543 -> 0.611 | Improved |
| use GAT-based AGAE instead of GCN-based GAE (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.559 -> 0.611 | Improved |
| add the gene graph to the spot graph (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.539 -> 0.611 | Improved |
| use element-wise weighting addition instead of matrix-wise weighting (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.573 -> 0.611 | Improved |
| use element-wise weighting addition instead of cross-attention fusion (ablation) | SpaEF · HGBM PCNA gene-protein correlation | PCC: 0.555 -> 0.611 | Improved |
| use SpaEF instead of STAGATE on extremely sparse data | SpaEF · HDHBC | PCC: 40.54% improvement in PCC | Improved |
| use SpaEF instead of DeepGFT | SpaEF · HOCWT masking | RMSE: 0.3931 -> 0.3664 | 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.
- DUSTED — DUSTED: Dual-Attention Enhanced Spatial Transcriptomics Denoiser (AAAI 2025) · better on 3 of 4 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
- Global Context-aware Representation Learning for Spatially Resolved Transcriptomics (ICML 2025 · similarity 0.87)
- SUICA: Learning Super-high Dimensional Sparse Implicit Neural Representations for Spatial Transcriptomics (ICML 2025 · similarity 0.86)
- DUSTED: Dual-Attention Enhanced Spatial Transcriptomics Denoiser (AAAI 2025 · similarity 0.86)
- STINR: Deciphering Spatial Transcriptomics via Implicit Neural Representation (CVPR 2025 · similarity 0.85)
- SToFM: a Multi-scale Foundation Model for Spatial Transcriptomics (ICML 2025 · similarity 0.85)
- SpotDiff: Spatial Gene Expression Imputation Diffusion with Single-Cell RNA Sequencing Data Integration (AAAI 2025 · similarity 0.85)
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.
