Skip to main content
Fig. 4 | Human Genomics

Fig. 4

From: Data-driven identification and classification of nonlinear aging patterns reveals the landscape of associations between DNA methylation and aging

Fig. 4

Analysis of variability function for age in Dataset1F. a Outline of the variability analysis using DICNAP. b Plots of MIC (X-axis) and Pearson’s correlation coefficient (Y-axis). Each dot represents a methylation site. c QQ plot of nonlinearity index, which is defined as \(MIC- \rho ^2\). The null distribution was generated by label permutations. The red line represents y = x. d Pie charts for the genome-wide methylation sites of age-related correlation patterns: NC (non-correlated), LI (linear increase), LD (linear decrease), NL (nonlinear). e Contribution rate for PC1–PC5. f PC1 and PC2 in the FPCA analysis for 1000 randomly picked NL sites. The colors represent the groups detected by clustering. g Representative age-related functions for nonlinear sites (X-axis: age, Y-axis: scaled variability)

Back to article page

Human Genomics

ISSN: 1479-7364