Abstract 3765: Single cell spatial proteomics analysis and computational evaluation pipeline

Abstract Resolving tissue and proteomic heterogeneity is critical to decoding the structure and function of tumor-immune microenvironment (TIME). Such understanding requires profiling of tumor and immune cell proteomic features with spatial resolution at the single-cell level. Although such spatially resolved methods and data sets are becoming increasingly available, analytical and computational methods that can extract the highly complex features and interactions within TIME are lacking. To address that problem, we have developed a computational pipeline we call the Spatial Proteomics Analysis and Computational Evaluation Pipeline (SPACE). The SPACE pipeline is composed of many analysis modules for processing and mining highly multiplexed imaging-based data types to explore TIME composition, organization, and heterogeneity. The pipeline generates and interprets biomarker expression and positional information from multiplexed images using algorithms for image indexing, image registration, quality control, segmentation, identification and removal of non-specific signals, data normalization, automatic identification of missing data, and adjustment for left-over signals. The accurate intensity measurements at single cell level are then used to calculate the proposed spatial features that represent cellular interactions in TIME. A hierarchical decision tree of cell markers is used to annotate types and identities for individual cells. The SPACE enables statistical and differential analyses of complex spatial features as well as cell types/identities with respect to clinical annotations and genomic alterations. The visualization of spatial and imaging data is made possible through an open-source OMERO image repository and spatial maps that integrate diverse markers in a single representation. Here, we demonstrate the applications of our pipeline in diverse gastrointestinal tumor types (e.g., small bowel adenocarcinoma) and validate the importance of integrating tissue heterogeneity at spatial and single-cell level. The framework is applicable to nearly all highly multiplexed imaging data platforms, including but not limited to, CycIF, CODEX, and imaging mass cytometry. Citation Format: Behnaz Bozorgui, Zeynep Dereli, Guillaume Thibault, John N. Weinstein, Anil Korkut. Single cell spatial proteomics analysis and computational evaluation pipeline [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3765.