Extended Data Fig. 6: Mapping of GWAS susceptibility genes and identification of MS-blood vessels in WM samples.

a) Spatial plots mapping MS susceptibility genes onto spatial clusters. Colour scale depicts bins of the number of MS GWAS genes¹² expressed per cluster; border colours depict WM tissue subgroup. MS GWAS genes could not be mapped on sample C2, as only one cluster was identified (that is, the relative expression between clusters could not be computed). b) Overview of blood vessel (BV) analysis in WM lesion samples. (i) Representative spatial plot of a WM lesion sample with BV cluster 3 in red. (ii) Identification of 12 MS-BV signature genes by overlapping BV cluster marker genes from 4 WM lesion samples. c) Representative spatial plots of the BV gene module in active and act/inact lesion samples. Red spots indicate positive MS-BV spots. d) Scatter plot depicting the percentage of positive MS-BV spots per subgroup for active and act/inact lesions. Horizontal lines depict mean expression per subgroup. Within samples, statistical significance between subgroups was determined for active lesions (n = 4 samples) and for act/inact lesions (n = 6 samples) by repeated measures ANOVA followed by two-tailed dependent-samples t-test with Bonferroni correction, *P < 0.05. Between samples, differences between subgroups were tested using a linear mixed model with two-tailed t-tests based on Satterthwaite’s method with Bonferroni correction, using the act/inact lesion core as reference. #P < 0.05, ###P < 0.001, for exact p-values see Supplementary Table 10.