Characterization of inflammatory gene profiles in platelets. (A-B) RNA-Seq data from ET (N = 7) or healthy control (N = 3) platelets³⁵  were filtered by using Hallmark⁴⁰  inflammatory response (A, n = 200) or IFN response (B, n = 224) genes, followed by unsupervised hierarchical clustering by Euclidean distance (differentially expressed genes are shown). Note clear segregation of phenotypes in both panels. (C) In silico expression studies of aggregated IFN response genes (n = 224) by phenotype (normal, n = 3; ET, n = 7); data are expressed as mean ± standard error of the mean (SEM) of reads per kilobase per million (RPKM) normalized and scaled to mean 0. (D) Platelet quantitative polymerase chain reaction of select IRGs or platelet-specific genes (VWF and PF4) from a secondary cohort of healthy controls (n = 4) or ET (n = 8), expressed as mean ± SEM normalized to actin; box plots (panels C-D) represent the within-group interquartile range encompassing 50% of the values, whereas the 95% CIs and outliers are depicted. The horizontal bar within each box represents the group median. P values (t test) are shown. NS, not significant.