Abstract
Epigenetic clocks, based on DNA methylation (DNAm) have become popular estimates of “biological age”. However, many of these clocks have few overlapping CpGs, and the underlying biology remains poorly understood. This study uses weight network analysis to characterize groups of CpGs from various aging clocks in order to identify underlying methylation dynamics in aging. We estimated the shared co-methylation of just over 3,000 CpGs from multiple tissue samples. We are able to identify CpG modules-or highly connected networks of CpGs-that show distinct methylation patterning across tissues. Finally, using these modules, we determine their associations with phenotypic aging measures, as well as chromosomal and transcriptional characterizes. In conclusion, our results suggest that many of the pre-existing “epigenetic clocks” are sampling the same underlying biological phenomenon, despite inclusion of different CpG sets. Further, investigation of higher-order “modules” reduces dimensionality, facilitating identification of precipitating and resulting biological mechanism for epigenetic aging.