A0584
Title: Spatial transcriptomics profiling of the tumor microenvironment with dependent random partitions
Authors: Yunshan Duan - University of Texas at Austin (United States) [presenting]
Shuai Guo - MD Anderson Cancer Center (United States)
Wenyi Wang - The University of Texas MD Anderson Cancer Center (United States)
Peter Mueller - UT Austin (United States)
Abstract: Young-onset colorectal cancer (YOCRC), diagnosed at ages 18-50 years, is an increasingly urgent global health problem that remains poorly understood. Colorectal cancer has high intra-tumor heterogeneity, especially in the tissue architecture. We propose to develop state-of-the-art single-cell and spatial molecular profiling strategies to investigate the tumor microenvironment (TME) of YOCRC, filling a gap in current literature. We will develop model-based inference for spatial partitioning of samples using spatial transcriptomics and single-cell data for YOCRC and late-onset colorectal cancer (LOCRC) for comparison. Spot-level inference under existing methods cannot provide inference on cell-level behavior and immune profiles. The innovative features of the proposed work are (i) spatial random partition of imputed single cells with preference for spatially connected clusters; and (ii) dependent random partition of immune and non-immune cells. The proposed analysis aims to investigate spatial segmentation and cell interactions. Downstream tasks, including differential gene expression and immune lineage analysis, will be conducted. Our long-term goal is to fill the gaps in understanding of the distinct nature of TME in YOCRC, to gain biologic insight that may help YOCRC treatment and enable reducing the burden of this disease.
