stFormer integrates spatial ligand signaling into a foundation model for spatial transcriptomics
- 1Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Key Laboratory of System Control and Information Processing, Ministry of Education of China, Shanghai 200240, China
- 2State Key Laboratory of Biopharmaceutical Preparation and Delivery, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China
- 3Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Key Laboratory of System Control and Information Processing, Ministry of Education of China, Shanghai 200240, China; State Key Laboratory of Biopharmaceutical Preparation and Delivery, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China. Electronic address: [email protected]
Abstract
Recent foundation models for single-cell transcriptomic data generate informative, context-aware gene and cell representations. Spatial transcriptomic (ST) data offer extra positional insights but were not considered by these single-cell models. We introduce stFormer, a transformer model tailored for ST data, which employs the cross-attention module to incorporate spatial ligand genes. To unify different ST technologies with a trade-off between resolution and gene coverage, we propose a biased cross-attention method that enables the model to do learning with single-cell resolution on whole-transcriptome but low-resolution Visium data. We assembled a pretraining corpus comprising ∼4.1 million spatial samples from public human Visium datasets. After pretraining, stFormer improved upon the state-of-the-art single-cell foundation model, scFoundation, across diverse tasks, including cell clustering, batch effect correction, cell type prediction, and gene function prediction. stFormer also revealed intercellular ligand-receptor signaling responses via in silico perturbation. Finally, we demonstrated stFormer's utility in yielding biological findings on two case studies.
Related
- Comments on "SpatialDG: a novel spatial domain identification method for spatially resolved transcriptomics data based on dual-graph neural network".
- Spatial transcriptomics reveals microenvironmental heterogeneity in osteoarthritic subchondral bone
- Applying Spatial Transcriptomics to the Human Heart: A Practical Guide for Translational Studies.
- Highly resolved tumor architecture via matched spatial and nucleus transcriptomics from a single tissue section
- hoodscanR: profiling single-cell neighborhoods in spatial transcriptomics data
BibTeX
@article{cao2026stformer,
title = {{stFormer integrates spatial ligand signaling into a foundation model for spatial transcriptomics}},
author = {Shenghao Cao and Kaiyuan Yang and Jiabei Cheng and Jiachen Li and Hong-Bin Shen and Xiaoyong Pan and Ye Yuan},
journal = {Cell Reports Methods},
year = {2026},
month = sep,
doi = {10.1016/j.crmeth.2026.101612},
url = {https://doi.org/10.1016/j.crmeth.2026.101612},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - stFormer integrates spatial ligand signaling into a foundation model for spatial transcriptomics
AU - Shenghao Cao
AU - Kaiyuan Yang
AU - Jiabei Cheng
AU - Jiachen Li
AU - Hong-Bin Shen
AU - Xiaoyong Pan
AU - Ye Yuan
PY - 2026
DA - 2026/09/01
JO - Cell Reports Methods
DO - 10.1016/j.crmeth.2026.101612
UR - https://doi.org/10.1016/j.crmeth.2026.101612
AB - Recent foundation models for single-cell transcriptomic data generate informative, context-aware gene and cell representations. Spatial transcriptomic (ST) data offer extra positional insights but were not considered by these single-cell models. We introduce stFormer, a transformer model tailored for ST data, which employs the cross-attention module to incorporate spatial ligand genes. To unify different ST technologies with a trade-off between resolution and gene coverage, we propose a biased cross-attention method that enables the model to do learning with single-cell resolution on whole-transcriptome but low-resolution Visium data. We assembled a pretraining corpus comprising ∼4.1 million spatial samples from public human Visium datasets. After pretraining, stFormer improved upon the state-of-the-art single-cell foundation model, scFoundation, across diverse tasks, including cell clustering, batch effect correction, cell type prediction, and gene function prediction. stFormer also revealed intercellular ligand-receptor signaling responses via in silico perturbation. Finally, we demonstrated stFormer's utility in yielding biological findings on two case studies.
SN - 2667-2375
ER -