DNA methylation entropy is a biomarker for aging

Abstract

The dynamic nature of epigenetic modifications has been leveraged to construct epigenetic clocks that accurately predict an individual’s age based on DNA methylation levels. Here we explore whether the accumulation of epimutations, which can be quantified by Shannon’s entropy, changes reproducibly with age. Using targeted bisulfite sequencing, we analyzed the associations between age, entropy, and methylation levels in human buccal swab samples. We find that epigenetic clocks based on the entropy of methylation states predict chronological age with similar accuracy as common approaches that are based on methylation levels of individual cytosines. Our approach suggests that across many genomic loci, methylation entropy changes reproducibly with age.