Abstract
The comparative cross-species analysis is a powerful tool to resolve the mysteries of evolution and phenotypic disparities among animals. This is the first network analysis of 10,000 DNA methylome data from 176 mammalian species to identify co-methylation modules that relate to individual (age, sex, tissue type) and species characteristics (e.g. phylogenetic order, maximum lifespan, adult weight). The unexpected correlation between DNA methylation and species were sufficiently strong to allow the construction of phyloepigenetic trees that parallel the phylogenetic tree. Weighted correlation network analysis identified 55 distinct co-methylation modules, i.e. sets of highly correlated CpGs. 31 of these modules are readily interpretable in terms of their relationship to age, maximum lifespan, tissue type etc. An age-related module was perturbed by gold standard anti-aging interventions in mice such as caloric restriction or growth hormone receptor knock outs. Our module-based analysis greatly enhances our biological understanding of age-related changes in DNA methylation across many species.