Extended Data Fig. 5: MAGIC-seq enables spatial transcriptomic profiling of diverse tissue types at high throughput.

a. Fluorescence intensity of barcoded arrays in the nine-grid chip. Cy3-labeled linker was used to mark the position of all spots and the average fluorescence of each spot was calculated and summarized for each region. Boxes, interquartile range. Center lines, median. Whiskers, values within 1.5× interquartile range of the top and bottom quartiles. b. Evaluation of MAGIC-seq’s performance among different regions on the same slide. Three types of tissues (olfactory bulb, cerebellum and colon) were sectioned at a thickness of 10 μm and mounted to the same nine-grid chip. The sections on the left or right side for each tissue type were selected as replicates for downstream sample preparation and sequencing analysis. Unsupervised clustering and spatial mapping of all spots from two replicates of each tissue type were performed. The correlation between two replicates was calculated based on gene expression of all spots. c. Correlation analysis of gene expression among four brain sections, related to Fig. 1h. R, Spearman correlation coefficient. d. UMAP plots of spots from four brain sections. The colored dots represented spots from different sections (left) and clusters (right), respectively, related to Fig. 1h. e. Spatial distribution of clusters across different brain sections. f. Distribution of the number of UMI (left) and gene (right) counts per spot for the nine tissue types. Boxes, interquartile range. Center lines, median. Whiskers, values within 1.5× interquartile range of the top and bottom quartiles. g. Unsupervised clustering of spots from nine tissue types based on gene expression. h. Mean expression of marker genes for each cluster in the nine tissue types. i. Spatial mapping of representative marker genes for each cluster in the placenta. Exp., expression.