Fig. 2: Proteomic data and summary models that differentiate untreated multiple sclerosis (MS) cohort from healthy controls (HC) and that evolve with MS natural history.

a Volcano plot of 270 somamers with cell-specificity to oligodendrocytes, monocytes, macrophages, and plasma cells that passed false discovery rate (FDR)-adjusted p < 0.05 in differentiating MS from HC. Differences were tested with two-sided Wilcoxon rank-sum test. Insets show protein PIGR (decreased in untreated MS, n = 634 vs HC, n = 69) and TNFRSF13B (increased in MS), evaluated by two-sided t-test (unadjusted p values shown). Green lines mark HC means. Boxplots display medians (black line), quartiles, whiskers (1.5× interquartile range [IQR]), and outliers. One-sided chi-squared tests revealed over-representation of oligodendrocyte markers among proteins decreased in MS (blue circles) and immune-specific markers among proteins increased in MS (red circles). b Ingenuity Pathway Analysis (IPA) functions identified from proteins significantly differentiating MS from HC. Red bars denote activated processes, blue bars denote inhibited processes (based on z-scores). Green dots indicate significance level (unadjusted –log10 p-values). c IPA-derived summary model illustrating biological processes and pathways associated with proteins distinguishing MS from HC. d IPA summary model of proteins showing significant non-zero yearly change in longitudinal untreated MS cohort, reflecting disease evolution.