Related Experiment Video
Updated: May 12, 2026
10:16
Mining Spatial Transcriptomics Datasets using DeepSpaceDB
Published on: September 5, 2025
Combining Spatial Multi-Omics Data to Decipher Spatial Domains and Elucidate Cell Heterogeneity Based on
Yuejing Lu1,2, Rui Qiao1, Ying Li3
1School of Mathematics and Statistics, Henan University of Science and Technology, Luoyang, China.
Summary
SOTMGF, a novel framework, enhances spatial multi-omics analysis by fusing diverse data types. It improves spatial domain identification and reveals molecular insights for biomarker discovery.
Area of Science:
- Spatial biology
- Multi-omics data integration
- Computational biology
Background:
- Spatial multi-omics technologies offer in situ molecular profiling.
- Integrating multi-modal data for spatial domain identification and cell heterogeneity analysis remains challenging.
Purpose of the Study:
- To develop SOTMGF, a self-supervised, goal-directed multi-view graph fusion framework for spatial multi-omics data.
- To enhance spatial domain identification, data denoising, and detection of spatially variable molecular features.
Main Methods:
- SOTMGF framework with five modules: pre-clustering, sparse feature processing, multi-view feature extraction and fusion, and multi-modality integration.
- Iterative optimization of self-training and graph embedding within a unified framework.
- Joint analysis of spatial transcriptomics (ST) and proteomics (SP), computational generation of spatial ATAC-seq, and reconstruction of spatial pseudo-expression.
Main Results:
- SOTMGF outperformed existing methods in spatial domain identification and denoising.
- Identified spatial dark genes/proteins (SDGs/SDPs) and revealed mRNA-protein discordance.
- Predicted key transcription factors and aided biomarker and therapeutic target discovery.
Conclusions:
- SOTMGF advances spatial biology research by enabling comprehensive multi-omics data integration.
- The framework facilitates a deeper understanding of molecular regulatory mechanisms.
- SOTMGF supports biomarker and therapeutic target discovery for improved health outcomes.
