Abstract:The rise of spatial transcriptomics (ST) is transforming our understanding of tumor heterogeneity by enabling high-resolution, location-specific mapping of gene expression across the tumor microenvironment. However, the translational potential of spatial transcriptomics is still limited by its high cost, hindering the assembly of large patient cohorts needed for robust biomarker discovery. Here we present Path2Space, a deep learning model trained on spatial transcriptomics data to predict spatial gene expression directly from histopathology slides. Studying breast cancer, the model was first trained on one ST breast cancer patients cohort and further independently tested and validated on two other ST cohorts. Path2Space identifies 4,500 genes whose spatial expression is robustly predicted in all three cohorts, markedly more accurate than existing ST predictors. Second, we show that based on the inferred ST, Path2Space accurately infers cell-type abundances in the tumor microenvironment (TME). Thirdly, Applying Path2Space to the TCGA breast cancer cohort we identify three TME spatially grounded subsets of patients that have markedly different survival rates. Fourth, analyzing two large breast cancer patients cohorts treated with trastuzumab, we show that the heterogeneity of the inferred HER2 levels across the tumor microenvironment is predictive of response, concordant with the bystander hypothesis. Finally, we developed a spatially interpretable model for predicting trastuzumab response directly from the H&E-slides. This model's performance surpasses that of existing response prediction models that use measured bulk multi-omics data. Path2Space is the first approach capable of robustly delineating the tumor microenvironment on an unprecedented scale directly from the tumor slides. It heralds the upcoming development of robust spatially grounded biomarkers of patients survival and treatment that promises to transform precision oncology in both the developed and developing worlds.

Investigating spatial dynamics in spatial omics data with StarTrail

Revealing Spatial Multimodal Heterogeneity in Tissues with SpaTrio

Unveiling Tissue Structure and Tumor Microenvironment from Spatially Resolved Transcriptomics by Hypergraph Learning

Spatial transcriptomics in human biomedical research and clinical application

spatialGE is a User-Friendly Web Application that Facilitates Spatial Transcriptomics Data Analysis

Spatial transcriptomics in development and disease

spatialGE: A user-friendly web application to democratize spatial transcriptomics analysis

Scalable spatial single-cell transcriptomics and translatomics in 3D thick tissue blocks

Spatial Transcriptomics Technology in Cancer Research.

Spatial transcriptomics: a new frontier in cancer research

StarmapVis: an interactive and narrative visualisation tool for single-cell and spatial data

Path2Space: An AI Approach for Cancer Biomarker Discovery Via Histopathology Inferred Spatial Transcriptomics

Spatially aware dimension reduction for spatial transcriptomics

Inferring histology-associated gene expression gradients in spatial transcriptomic studies

Inferring super-resolution tissue architecture by integrating spatial transcriptomics with histology

Accurately deciphering spatial domains for spatially resolved transcriptomics with stCluster

Interpretable Spatial Gradient Analysis for Spatial Transcriptomics Data

Data enhancement in the age of spatial biology

Differential gene expression analysis of spatial transcriptomic experiments using spatial mixed models

Enhancing Spatial Transcriptomics Analysis by Integrating Image-Aware Deep Learning Methods

Integrating spatial and single-cell transcriptomics data using deep generative models with SpatialScope