Co-clustering Models for Spatial Transciptomics: Analysis of a Human Brain Tissue Sample

In the last few years, we have witnessed a substantial improvement in the efficiency of DNA sequencing techniques with the advent of 10X-Visium, a new technology that is capable of providing the expression of tens of thousands of genes inside thousands of cells from a tissue sample. From a statistical perspective, this technology represents an astonishing step forward in the analysis of single cell data, as it gives access to a huge amount of information inaccessible to us until now. In this paper, we apply some innovative statistical methods that cluster both the rows and the columns of a data matrix to a human brain tissue sample processed with 10X-Visium. This operation is known as co-clustering and aims to detect groups of genes whose expression activity is similar in some specific areas of the brain tissue.