A New Statistical Approach for the Simultaneous Clustering of Genes and Cells in Spatial Transcriptomic Experiments

• Activity Number: 484 - Applied Bayesian Methodology
• Type: Contributed
• Date/Time: Wednesday, August 10, 2022 : 2:00 PM to 3:50 PM
• Sponsor: International Society for Bayesian Analysis (ISBA)
• Author(s): Andrea Sottosanti* and Davide Risso
• Companies: University of Padova and University of Padova
• Keywords: Co-clustering; EM algorithm; integrated complete log likelihood; Model-based clustering; Spatial transcriptomics

Abstract

Spatial transcriptomics is a modern sequencing technology that allows the measurement of the activity of thousands of genes in a tissue sample and maps where the activity is occurring. This technology has enabled the study of spatially-expressed genes, i.e., genes which exhibit spatial variation across the tissue. Comprehending gene functions and interactions in different areas of the tissue is of great scientific interest, as it might lead to a deeper understanding of several key biological mechanisms. However, adequate statistical tools that exploit the previously unavailable spatial information are still lacking. We introduce a new model that clusters the spatial expression profiles of the genes according to the tissue areas. This is accomplished by performing a co-clustering, i.e., inferring the latent block structure of the data and inducing a simultaneous clustering of the genes and of the image. Our proposed methodology is validated with a series of simulations and its usefulness in responding to specific biological questions is illustrated with an application to a human brain tissue sample processed with the 10X-Visium protocol.