Project Summary – Tech Core 2 Spatial tumor heterogeneity plays a critical role in multiple stages of cancer progression and metastasis including the venous invasion that contributes to increased risk of cancer cell dissemination. This process involves spatially distinct interaction between cancer cells and the surrounding microenvironment at multiple sites. In Tech1, PI Wirtz will develop a new 3D multiscale tumor cell mapping method, CODA, which can create a 3D large-scale tumor cell anatomy at single cell level via tissue histology image integration and trained deep-learning semantic algorithms. What is highly desired to further add to this 3D tumor cell anatomic atlas is genome-wide molecular information such as mRNAs and a large panel of proteins for unbiased discovery of cell subtype, state, and interaction, and potentially to infer new mechanisms or targets for therapeutic intervention. TECH2 PI Fan recently developed a novel technology called DBiT-seq for high-spatial-resolution multi-omics mapping via deterministic barcoding in tissue at cellular level (~10µm), whole transcriptome scale (>22,000 genes), high coverage (>2,000 genes per 10µm pixel), and multi-omics profiling (co-mapping of ~300 protein markers), which can be readily applied to FFPE tissue sections and integrated with CODA. In Tech2, we propose the following two aims: AIM 1. A high-throughput, low cost, high quality/coverage, multi-omic mapping method (DBiT-seq) with full compatibility with human PFA and FFPE tissue samples. AIM 2. Integrating CODA and DBiT-seq for 3D multi-omic tumor imaging. Successful completion of these two aims will lead to the first genome-wide multi- omics 3D view of vascular or lymphovascular invasion of human tumors and a powerful technology platform for the consortium to investigate spatial tissue heterogeneity in other human cancers.