Decomposing spatial heterogeneity of cell trajectories with Paella

ABSTRACT Spatial transcriptomics provides a unique opportunity to study continuous biological processes in a spatial context. We developed Paella, a computational method to decompose a cell trajectory into multiple spatial sub-trajectories and identify genes with differential temporal patterns across spatial sub-trajectories. Applied to spatial transcriptomics datasets of cancer, Paella identified spatially varying genes associated with tumor progression, providing insights into the spatial heterogeneity of cancer development.