Fig. 2

RA-associated chromatin dysregulation in peripheral monocytes represents an RA signature. a Barplot showing the numbers differentially accessible chromatin regions for each of the four examined immune cell types, binned according to the magnitude of the detected OA and RA vs. HD, coloured according to the fold change. RA vs. HD (left), OA vs. HD (middle), and RA vs. OA (right). b Venn diagram for the overlap of differentially accessible chromatin regions in the four examined immune cell types from the RA vs. OA analysis (|log₂ fold change | > 1, p < 0.001 and FDR < 0.1). c Heatmap of 8836 peaks obtained by comparing the ATAC-seq profiles of monocytes from healthy donors, OA patients, and RA patients (|log₂ fold change | > 1, p < 0.001 and FDR < 0.1). Each column is a sample; each row is an accessible region detected by ATAC-seq; peaks are organized based on unsupervised clustering. Samples from the same group are marked with the same colour. Unsupervised clustering of differential peaks is performed using K-means algorithm with the k value of 3. d Diagram for the average peak intensity signals of three clusters (C1, C2, and C3) for healthy donors, OA patients, and RA patients. e GREAT was used to analyse C3 peaks for enrichment of disease ontology categories and MsigDB pathways. f Principal component analysis based on the C3 peaks for all samples. Each dot is a sample. g Barplot of normalized signal intensity of C3 peaks (RAAS) in healthy donors, OA patients and RA patients. The values of p were calculated with an unpaired t-test. ****p < 0.0001. Error bar represent standard deviation (SD). HD, healthy donor; OA, osteoarthritis; RA, rheumatoid arthritis; C1, cluster 1; C2, cluster 2; C3, cluster 3