Identifying genes with location dependent noise variance in spatial transcriptomics data

Abstract

Spatial transcriptomics (ST) holds the promise to identify the existence and extent of spatial variation of gene expression in complex tissues. Such analyses could help identify gene expression signatures that distinguish between physiology and disease. Existing tools to detect spatially variable genes assume a constant noise variance across location (homoscedastic). This assumption might miss important biological signals when the variance could change across locations, e.g., in the tumor microenvironment. As an alternative, we propose NoVaTeST, a novel method to identify genes with location-dependent noise variance in ST data. NoVaTeST models gene expression as a function of location with a heteroscedastic noise. It then compares the model to one with homoscedastic noise to detect genes that show significant spatial variation in noise. Our results show genes detected by NoVaTeST provide complimentary information to existing tools while providing important biological insights.