SPATIAL OMICS FOR DEEP PHENOTYPING IN COLORECTAL CANCER Group Zlobec, Williams Williams group are using state-of-the-art spatial transcriptomic methodologies to characterize transcriptional subtype heterogeneity and decipher the biological processes underpinning epithelial identity in CRC. Zlobec group, together with Lunaphore Technologies (Innosuisse) are establishing a high-dimensional protein panel to study tumour budding and its microenvironment under native and treatment scenarios. Data integration from distinct spatial modalities of the Williams and Zlobec groups using well-documented patient collectives and ngTMA® will provide high dimensional, multi-omics perspective of colorectal cancer. Tumor microenvironment in colorectal cancer at 20x magnification: a, Panck (red) and Vimentin (green); b, CD20 (pink) and CD3 (yellow); c, E-cadherin (green) and CDX2 (red).
Digital pathology & AI to gain novel insghts into colorectal cancer Group Zlobec Our Sinergia project uses AI to harness the power of histopathology images, genomics (focusing on STRs), and pharmacoscopy to gain novel insights into colorectal cancer biology and understand their impact on clinical outcomes. We investigate morphomolecular relationships, including the CMS classification, and intratumoral heterogeneity in order to learn new interpretable & clinically important features from histopathology images. We use various computational methods, including graphs and deep learning) to evaluate the structural and spatial patterns at the tumor invasion front in neoadjuvantly treated patients. We’ve extended our scope to understanding CMS using spatial transcriptomic and protein expression analysis. The tumor microenvironment, with its complex stromal patterns and immune contexture are important focus points. Collaborators on this project include M. Rodriguez (IBM Research), M. Anisimova (ZHAW), B. Snijder (ETH Zürich), A. Fischer (HES-SO & UniFribourg) and V. Koelzer (UniZürich). Epithelial cell and lymphocyte graphs in colorectal cancer.
Building tools for computer-assisted diagnostics Group Zlobec In addition to exploratory tissue analysis, our team builds, tests and validates in-house, open-source and commercially available algorithms for potential diagnostic use and workflow integration. We are currently running a comparative study on the impact of scanners and performance of different software for Ki-67 detection and quantification. We use deep learning methods for segmentation and metastatic detection in lymph nodes, and streamline processes lab and data analysis processes, for e.g from scanning to construction of “next-generation Tissue Microarrays®” (www.ngtma.com) to visual presentation of results and analysis. We use graphs and geometric deep learning to learn about tumor budding and lymphocytes, and as part of our collaboration with the International Budding Consortium, generate hot-spot detection and tumor budding quantification algorithms in early stage pT1 cancers. Computational Analysis of Colorectal Cancer Metastases in Lymph Nodes.