A1011
Title: Bayesian multi-sample and multi-scale clustering with feature selection for spatial transcriptomics data
Authors: Alvin Sheng - University of Minnesota, Twin Cities (United States) [presenting]
Thierry Chekouo - University of Minnesota (United States)
Sandra Safo - University of Minnesota (United States)
Abstract: Recent advances in spatial transcriptomics have enabled researchers to measure gene expression at the single-cell spatial resolution, often for multiple tissue sections in a single study. The aim is to present a Bayesian method that simultaneously performs factor analysis and spatial clustering on multiple samples, where the clustering is done at both the single-cell and tissue regional scale. Therefore, the method simultaneously sorts the cells into cell types and partitions the cells into spatial domains of tissue. To increase interpretability, a feature selection mechanism is employed within the estimation of the sparse factor loadings matrix, which detects genes that discriminate between cell-type clusters. The advantages of the method are illustrated over alternative state-of-the-art approaches through simulation studies and three real data applications.
