Extended Data Fig. 1: scRNAseq reveals cell heterogeneity in PAD.
a, Representative images of gastrocnemius muscle from non-ischemic (n = 8) and PAD (n = 11) stained for nuclei (Hoechst, blue) and WGA (white) (Scale bar: 50 µm) and quantification of percentage of centrally nucleated fibers over total fibers. b, Representative images of gastrocnemius muscle from non-ischemic (n = 8) and PAD (n = 11) stained for αSMA (magenta), CD31 (green), nuclei (Hoechst, blue) and WGA (white) (Scale bar: 50 µm) and quantification of number of arterioles per mm2. c, TSNE plot (Fig. 1g) separated by condition, color indicates the condition. d, Schematic representation of the pipeline used for generating the spatial transcriptomics dataset (see Methods). e, Schematic representation of the pipeline used for transferring cluster labels from scRNAseq to spatial data (see Methods). f, Stacked bar plots showing cluster percentage in each condition in the spatial dataset from non-ischemic (n = 3) and PAD (n = 3) patients, color-coded by cluster. Each stack represents mean - SEM. g, Bar plots showing Normalized Enrichment Score of significant (adjusted p-value < 0.05) “Abnormality of muscle physiology” pathways (Human Phenotype Ontology) from GSEA analysis on the pseudobulk dataset, color indicates the condition. Each dot represents a single patient in panels a and b. Student’s t test (two-tailed, unpaired, parametric, ns > 0.05) was used in a and b. Adaptive multi-level Monte-Carlo scheme (as implemented in fgsea package) was used in g. P = 0.0695 (a), P = 0.0019 (b).
