A0357
Title: Statistical identification of cell type-specific spatially variable genes in spatial transcriptomics
Authors: Lulu Shang - MD Anderson (United States) [presenting]
Abstract: An essential task in spatial transcriptomics involves identifying genes with spatial expression patterns, known as spatially variable genes (SVGs). Importantly, a subset of SVGs displays diverse spatial expression patterns within a given cell type, thus representing key transcriptomic signatures underlying cellular heterogeneity. Celina, a statistical method, is presented for systematically detecting this subset of cell type-specific SVGs (ct-SVGs). Celina utilizes a spatially varying coefficient model to accurately capture each gene's spatial expression pattern in relation to the distribution of cell types across tissue locations, ensuring effective type I error control and high statistical power. The performance of Celina is evaluated through comprehensive simulations and applications to five real datasets, where existing methods are also adapted and examined, originating from other analytic settings to detect ct-SVGs. Celina proves powerful compared to these ad hoc method adaptations in single-cell resolution spatial transcriptomics and stands as the only effective solution for spot-resolution spatial transcriptomics. The ct-SVGs detected by Celina also enable novel biologically informed downstream analyses, unveiling functional cellular heterogeneity at an unprecedented scale.
