Fig. 3: Visium and Xenium co-registration.
From: Giotto Suite: a multiscale and technology-agnostic spatial multiomics analysis ecosystem
a, Pictogram showing multimodal co-registration from serial tissue sections. b, Aligned Visium H&E, post-Xenium HER2 immunofluorescence (cyan), Xenium (blue) and Visium (orange) spatial datasets from a single breast cancer sample (10x Genomics). c, Overlay of HER2 protein (grayscale), 10x cell segmentations (cyan) and ERBB2 transcripts (points, red); inset shows zoomed-in region. d, CD20 intensity data (far left), MS4A1 transcript counts (middle left), HER2 intensity data (middle right) and ERBB2 transcript counts (far right) hexbinned at the 8-μm center-to-center level. e, Bivariate Moran’s I of MS4A1 versus CD20 (red) and ERBB2 versus HER2 (blue) calculated for hexbinned values at the 8-μm, 16-μm, 32-μm, 64-μm, 128-μm, 256-μm and 512-μm center-to-center levels. f, Rank of correlation scores between overlapping genes expressed in Xenium and registered Visium data. g, LOESS regression (solid curves) of mean log expression against ranked correlation scores from f. h,i, Spatial feature expression plots for registered Xenium and Visium data for genes FASN (h) and HDC (i). IF, immunofluorescence.
