Fig. 4: Aging-induced changes to the myeloid/macrophage immune cells.

A UMAP plot of subcluster analysis for the original myeloid clusters (6,14, 19, and 20) from Fig. 1B. B Dot plot shows expression of myeloid markers used for cell-type classification. C Changes in the proportion of the myeloid subclusters with age are shown, as determined using permutation testing. Error bars indicate the 95 percent confidence interval for the magnitude of the difference between ages as determined by bootstrapping. Dashed lines indicate absolute log₂-fold changes greater than 0.58, and false discovery rate less than 0.05. Gray points indicate no significant difference between ages. D Bar plot highlights the relative signaling for all signaling pathways identified as significantly altered in myeloid cells (>20% change in strength, p-value <= 0.01) between young and old datasets. E Heatmaps show the relative contribution of each myeloid subtype to the total pathway signaling, shown for incoming and outgoing signaling in both age groups. Colors represent relative contribution for each cell type (values scaled row-wise with row max = 1). Dendrograms represent the results of hierarchical clustering of both rows and columns. F Bar plot shows the top signaling pathways from stromal fibroblasts to myeloid cells in young and old datasets. Chord plots show collagen signaling from stromal fibroblasts to myeloid cells (G), TGFβ signaling from myeloid cells to stromal fibroblasts (H), MHC-I signaling from myeloid cells to non-B cell lymphoid (NBL) cells (I), CXCL signaling from myeloid cells to NBLs cells (J), and CCL signaling from NBLs to myeloid cells in young (top) and old (bottom) datasets. Chord plot signaling is depicted by lines connecting subtypes, and the strength of signaling is depicted through the thickness of lines in young (top) and old (bottom) datasets. Source cell types are bold black, and target cell types are bold green.