Abstract
The application of spatial transcriptomics (ST) technologies is booming and has already yielded important insights across many different tissues and disease models. In nephrology, ST technologies have helped to decipher the cellular and molecular mechanisms in kidney diseases and have allowed the recent creation of spatially anchored human kidney atlases of healthy and diseased kidney tissues. During ST data analysis, the computationally annotated clusters are often superimposed on a histologic image without their initial identification being based on the morphologic and/or spatial analyses of the tissues and lesions. Herein, histopathologic ST data from a human kidney sample were modeled to correspond as closely as possible to the kidney biopsy sample in a health care or research context. This study shows the feasibility of a morphology-based approach to interpreting ST data, helping to improve our understanding of the lesion phenomena at work in chronic kidney disease at both the cellular and the molecular level. Finally, the newly identified pathology-based clusters could be accurately projected onto other slides from nephrectomy or needle biopsy samples. Thus, they serve as a reference for analyzing other kidney tissues, paving the way for the future of molecular microscopy and precision pathology.