Charting a structural and biochemical ECM niche in solid tumours using multi-matrisomics (Dr. Hannah Williams) Group Zlobec, Williams The extracellular matrix (ECM) forms part of the triad of the tumour ecosystem but is understudied in terms of its contribution to tumour biology. The Williams group utilises spatially resolved multi-matrisomics (spatial transcriptomics, proteomics and digital image analysis) in 2D and 3D to deeply characterise the structural and biochemical manifestations of the matrisome and the association of this to tumour biology and its clinical relevance for patient outcome. We work across a variety of solid tumour types including colorectal cancer and pancreatic ductal adenocarcinoma. Our current interests include: structural and biochemical phenotyping of the ECM across 2D and 3D modalities, the ECM and epithelial identity and CAF mechanisms of desmoplasia. Multi-modal assessment of extracellular matrix in solid tumours
3Dhist: Evolution of colorectal cancer metastases using multimodal 2D and 3D imaging Group Zlobec, Williams The goal of this SNF-funded project is to elucidate metastatic pathways, from lymph node metastases (LNM) to tumour deposits (TD), using 2D and 3D analyses. As the clinical partner in this multi-institutional collaboration, our group is responsible for the morphological and spatial mapping of LNM and TD in 2D to assess their metastatic potential and generate insights that may refine current clinical and biological paradigms. Our 2D observations will inform subsequent 3D analysis using microCT, nanoCT, and 3D light-sheet fluorescence microscopy (LSFM), enabling 2D observations to be visualised in a volumetric context. Through this integrated framework, the project aims to provide a more comprehensive understanding of CRC metastatic evolution. Computational analysis reveals that the presence of tumor budding and complex tumor morphology in LNMs is linked to worse prognosis
Building tools for computer-assisted diagnostics Group Zlobec, Williams In addition to exploratory tissue analysis, our team develops, tests, and validates in-house, open-source, and commercially available algorithms for potential diagnostic use and seamless workflow integration. We are currently generating a pan-lymph node metastasis algorithm (MetAssist) using state-of-the-art deep learning methods. We optimise the entire process, from laboratory workflows to data analysis, and support the visualisation of results and interaction between our algorithms and pathologists’ assessments. By integrating pathologists feedback, our goal is to enhance the performance of MetAssist and deliver a reliable nodal screening tool to assist pathologists in routine clinical practice on several tissue types. We work closely with our expert pathology colleagues for this purpose. Computational Analysis of Metastases in Multi-cancer Lymph Nodes
Unravelling the spatially resolved transcriptional landscape of colorectal cancer biology Group Zlobec, Williams Our group leverages single-cell spatial transcriptomics (GeoMx, CosMx) to decode the biology of solid tumors in unprecedented detail. We focus heavily on technical rigor, benchmarking data acquisition and analysis pipelines to ensure robust biological interpretation. By integrating tissue morphology with underlying molecular states, we have identified novel independent prognostic biomarkers, such as a tumor-interaction score linking cancer-stromal engagement to inflammation and extracellular remodeling. Furthermore, we investigate predictive markers for radiotherapy response, connecting tumor shrinkage to glutamine-related gene transcription. Ultimately, our work combines deep biological analysis with clinically relevant insights to improve patient stratification. Morpho-molecular analysis. (a) PanCK-stained TMA core. (b) Single-cell segmentation showing the tumor-interaction score (epithelial) vs. TME (grey). (c) Visualization of tissue acquisition and downstream single-cell phenotyping within a selected field of view
