# SToFM: a Multi-scale Foundation Model for Spatial Transcriptomics

arXiv: https://arxiv.org/abs/2507.11588 · HTML: https://mlsanity.com/paper/2507.11588/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 78 references that are in the map.
- First posted on arXiv 15 Jul 2025.

## Builds on (most cited first, 16 of 16)

- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [How Powerful are Graph Neural Networks?](https://mlsanity.com/paper/1810.00826/) (2018)
- [SimCSE: Simple Contrastive Learning of Sentence Embeddings](https://mlsanity.com/paper/2104.08821/) (2021)
- [Hierarchical Graph Representation Learning with Differentiable Pooling](https://mlsanity.com/paper/1806.08804/) (2018)
- [From Louvain to Leiden: guaranteeing well-connected communities](https://mlsanity.com/paper/1810.08473/) (2018)
- [GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond](https://mlsanity.com/paper/1904.11492/) (2019)
- [HEST-1k: A Dataset for Spatial Transcriptomics and Histology Image Analysis](https://mlsanity.com/paper/2406.16192/) (2024)
- [Why Do Pretrained Language Models Help in Downstream Tasks? An Analysis of Head and Prompt Tuning](https://mlsanity.com/paper/2106.09226/) (2021)
- [Rotation Invariant Graph Neural Networks using Spin Convolutions](https://mlsanity.com/paper/2106.09575/) (2021)
- [Causal machine learning for single-cell genomics](https://mlsanity.com/paper/2310.14935/) (2023)
- [STimage-1K4M: A histopathology image-gene expression dataset for spatial transcriptomics](https://mlsanity.com/paper/2406.06393/) (2024)
- [LangCell: Language-Cell Pre-training for Cell Identity Understanding](https://mlsanity.com/paper/2405.06708/) (2024)
- [xTrimoGene: An Efficient and Scalable Representation Learner for Single-Cell RNA-Seq Data](https://mlsanity.com/paper/2311.15156/) (2023)
- [ST-Align: A Multimodal Foundation Model for Image-Gene Alignment in Spatial Transcriptomics](https://mlsanity.com/paper/2411.16793/) (2024)
- [Large-Scale Cell Representation Learning via Divide-and-Conquer Contrastive Learning](https://mlsanity.com/paper/2306.04371/) (2023)
- [Cell-ontology guided transcriptome foundation model](https://mlsanity.com/paper/2408.12373/) (2024)

## Built on by (most cited first, 8 of 8)

- [Adapting a Pre-trained Single-Cell Foundation Model to Spatial Gene Expression Generation from Histology Images](https://mlsanity.com/paper/2603.19766/) (2026)
- [Artificial Intelligence Virtual Cells: From Measurements to Decisions across Modality, Scale, Dynamics, and Evaluation](https://mlsanity.com/paper/2510.12498/) (2025)
- [SAGE-FM: A lightweight and interpretable spatial transcriptomics foundation model](https://mlsanity.com/paper/2601.15504/) (2026)
- [Spatial Transcriptomics as Images for Large-Scale Pretraining](https://mlsanity.com/paper/2603.13432/) (2026)
- [Building artificial intelligence virtual tissue (AIVT) for tissue state representation, feature prediction, and dynamic simulation](https://mlsanity.com/paper/2606.29883/) (2026)
- [CellWorld: From Gene-Level Reconstruction to Latent Cell Prediction in Spatial Transcriptomics Foundation Models](https://mlsanity.com/paper/2608.06659/) (2026)
- [Towards Scalable Context-Aware Single-Cell Spatial Transcriptomics Prediction from Histology Images](https://mlsanity.com/paper/2609.36429/) (2026)
- [CellMSA: Context Modeling for Single-Cell Representation Learning](https://mlsanity.com/paper/2609.38908/) (2026)

## Abstract (arXiv)

Spatial Transcriptomics (ST) technologies provide biologists with rich insights into single-cell biology by preserving spatial context of cells. Building foundational models for ST can significantly enhance the analysis of vast and complex data sources, unlocking new perspectives on the intricacies of biological tissues. However, modeling ST data is inherently challenging due to the need to extract multi-scale information from tissue slices containing vast numbers of cells. This process requires integrating macro-scale tissue morphology, micro-scale cellular microenvironment, and gene-scale gene expression profile. To address this challenge, we propose SToFM, a multi-scale Spatial Transcriptomics Foundation Model. SToFM first performs multi-scale information extraction on each ST slice, to construct a set of ST sub-slices that aggregate macro-, micro- and gene-scale information. Then an SE(2) Transformer is used to obtain high-quality cell representations from the sub-slices. Additionally, we construct SToCorpus-88M, the largest high-resolution spatial transcriptomics corpus for pretraining. SToFM achieves outstanding performance on a variety of downstream tasks, such as tissue region semantic segmentation and cell type annotation, demonstrating its comprehensive understanding of ST data through capturing and integrating multi-scale information.