Niche Differential Gene Expression Analysis in Spatial Transcriptomics Data Identifies Context-Dependent Cell-Cell Interactions
Kaishu Mason, University of Pennsylvania
Spatial transcriptomic technologies have enabled scientists to better understand the complex relationships between different cell types in tissue. One popular line of inquiry is to identify which genes show global spatial expression patterns in tissue, known as spatially variable gene analysis. However, this mode of analysis is sample specific and does not generalize well across multiple studies. Additionally, spatially variable gene analysis is ill suited to recognize signals in gene expression due to local interactions between cells. To address these limitations, we introduce a new concept known as niche gene analysis which identifies genes in each cell type that show expression patterns based on the neighborhood profile, or niche. We present a method for identifying these niche genes called De-NicheR which provides robust type 1 error control for identifying niche enriched genes while also adjusting for technical effects such as spot swapping. We use De-NicheR to explore niche patterns in tumor associated macrophages in the colorectal cancer TME, including the identification of potential marker genes between Kupffer cells and tumor associated macrophages.
Abstract Author(s): Kaishu Mason, Dr.Nancy Zhang, University of Pennsylvania