This course covers the analysis of 10X Visium data using Seurat v4. We usually run it as a three-hour module after the single-cell RNA-seq course, because part of the theory is the same. You will learn how to

• create a Seurat v4 object
• perform QC and filter out low quality spots (damaged tissue)
• normalize gene expression values and detect highly variable genes with SCTransform
• reduce dimensions with PCA using the highly variable genes
• use the PCs to cluster spots with graph based clustering
• visualize clusters with UMAP and overlay with the tissue image
• detect spatially variable genes
• visualise gene expression on the tissue image
• integrate samples using the mutual nearest neighbor approach (anchors)
• subset anatomical regions
• predict cell type composition in spots: Integrate with scRNA-seq data

Course material:

• slides
• lecture videos
• exercises. The data are available on Chipster server in the example sessions listed in the exercise sheet.