Digital Spatial Pathway Mapping Reveals Prognostic Tumor States in Head and Neck Cancer

Citation: J. Hense*, M. Jamshidi Idaji*, L. Ciernik, J. Dippel, F. Ersan, M. Knebel, A. Pusztai, A. Sendelhofert, O. Buchstab, S.Fröhling, S. Otto, J. Hess, P. Liokatis, F. Klauschen, K.-R. Müller, A. Mock, 2025. “Digital Spatial Pathway Mapping Reveals Prognostic Tumor States in Head and Neck Cancer”, bioRxiv, DOI: 10.1101/2025.11.24.689710, *=equal contribution

• The paper is accessible through bioRxiv

• You can download all the codes via Github.

Summary: In this work, we infer transcriptome-derived signaling pathway activities directly from routine H&E slides. We propose to use the MIL heatmaps for stratifying the patients. In this regard, we propose a spatial tumor pathway activity score (TAPAS) for quantifying spatially resolved pathway activity based on XAI heatmaps. In this repository, we share the codes for this manuscript. We have also shared data of an exemplar patient at Zenodo. We have prepared the pipelines in a way that they work smoothly with the shared data.