A Multimodal Foundation Model of Spatial Transcriptomics and Histology for Biological Discovery and Clinical Prediction

TL;DR

STORM is a multimodal foundation model that integrates spatial transcriptomics and histology to improve biological discovery and clinical prediction across multiple organs and platforms.

Contribution

The paper introduces STORM, a hierarchical model trained on extensive data that bridges imaging and omics, enhancing spatial analysis and clinical outcome prediction.

Findings

Improves spatial domain discovery with biologically coherent tissue maps.

Outperforms existing methods in predicting gene expression from H&E images.

Enhances immunotherapy response prediction and prognostication across diverse cohorts.

Abstract

Spatial transcriptomics (ST) enables gene expression mapping within anatomical context but remains costly and low-throughput. Hematoxylin and eosin (H\&E) staining offers rich morphology yet lacks molecular resolution. We present \textbf{\ours} (\textbf{S}patial \textbf{T}ranscriptomics and hist\textbf{O}logy \textbf{R}epresentation \textbf{M}odel), a foundation model trained on 1.2 million spatially resolved transcriptomic profiles with matched histology across 18 organs. Using a hierarchical architecture integrating morphological features, gene expression, and spatial context, STORM bridges imaging and omics through robust molecular--morphological representations. STORM enhances spatial domain discovery, producing biologically coherent tissue maps, and outperforms existing methods in predicting spatial gene expression from H\&E images across 11 tumor types. The model is platform-agnostic, performing consistently across Visium, Xenium, Visium HD, and CosMx. Applied to 23 independent cohorts comprising 7,245 patients, STORM significantly improves immunotherapy response prediction and prognostication over established biomarkers, providing a scalable framework for spatially informed discovery and clinical precision medicine.