Fig. 3

Open and refined search of validation cohorts. A Heatmap depicting the levels of global modifications among different ages. The graphs show the relative intensity of young and middle-old group. Color bar represents the relative intensity of identified proteins from −4 to 4. B Hierarchical cluster analysis of 17 oxidations which are all belong to post-translational and chemical derivative modifications to the exclusion of artefact modifications. Color bar represents the relative intensity of identified proteins from −4 to 4. C The scatter plot histogram graphs showing 17 oxidations fluctuation change. Comparison of oxidation differences among young, middle-age, and old group. D The scatter plot histogram graphs showing seven differentially expressed oxidations that are potential diagnostic markers for aging. The horizontal line in the middle of the data represents the median, the upper and lower color lines represent the data quartile range, and the upper part of the data represents the significance between groups. C and D data are presented as mean ± SD. As the number of labels increases, the significance increases. (*p <0.05, **p <0.01, ***p <0.005, ****p <0.001). E The ROC plot of 7oxidation and an AA substitution. F The comparison of differentially expressed oxidations in discovery and validation cohorts. Each histogram represents the proportions of each modification, showing how the same modification compares at different ages. G STRING protein–protein interaction (PPI) network analysis of oxidative proteins (PPI enrichment p-value< 1.0e-16 ). H hierarchical cluster analysis (HCA) of these oxidative proteins distinguishes different ages well. I Heatmap of unsupervised cluster analysis depicting the landscape of all identified proteins from old to young. J The trends of 5 proteins with aging, PLAU and F2, were differentially expressed using two tailed tests, and the expression increased with aging (p value < 0.05, fold change > 2). (*p <0.05, *****p <0.00001)