Clonal hematopoiesis of indeterminate potential (CHIP) increases rapidly in prevalence beyond age 60 and has been associated with increased risk for malignancy, heart disease and ischemic stroke. CHIP is driven by somatic mutations in hematopoietic stem and progenitor cells (HSPCs). Because mutations in HSPCs often drive leukemia, we hypothesized that HSPC fitness substantially contributes to transformation from CHIP to leukemia. HSPC fitness is defined as the proliferative advantage over cells carrying no or only neutral mutations. If mutations in different genes lead to distinct fitness advantages, this could enable patient stratification. We quantified the fitness effects of mutations over 12 years in older age using longitudinal sequencing and developed a filtering method that considers individual mutational context alongside mutation co-occurrence to quantify the growth potential of variants within individuals. We found that gene-specific fitness differences can outweigh inter-individual variation and, therefore, could form the basis for personalized clinical management.

Recent studies suggest that epigenetic rejuvenation can be achieved using drugs that mimic calorie restriction and techniques such as reprogramming‐induced rejuvenation. To effectively test rejuvenation in vivo, mouse models are the safest alternative. However, we have found that the recent epigenetic clocks developed for mouse reduced‐representation bisulphite sequencing (RRBS) data have significantly poor performance when applied to external datasets. We show that the sites captured and the coverage of key CpGs required for age prediction vary greatly between datasets, which likely contributes to the lack of transferability in RRBS clocks. To mitigate these coverage issues in RRBS‐based age prediction, we present two novel design strategies that use average methylation over large regions rather than individual CpGs, whereby regions are defined by sliding windows (e.g. 5 kb), or density‐based clustering of CpGs. We observe improved correlation and error in our regional blood clocks (RegBCs) compared to published individual‐CpG‐based techniques when applied to external datasets. The RegBCs are also more robust when applied to low coverage data and detect a negative age acceleration in mice undergoing calorie restriction. Our RegBCs offer a proof of principle that age prediction of RRBS datasets can be improved by accounting for multiple CpGs over a region, which negates the lack of read depth currently hindering individual‐CpG‐based approaches.

The emergence of epigenetic predictors was a pivotal moment in geroscience, propelling the measurement and concept of biological ageing into a quantitative era. However, while current epigenetic clocks have shown strong predictive power, they do not reflect the underlying biological mechanisms driving methylation changes with age. Consequently, biological interpretation of their estimates is limited. Furthermore, our findings suggest that clocks trained on chronological age are confounded by non-age-related phenomena.To address these limitations, we developed a probabilistic model that describes methylation transitions at the cellular level. Our approach reveals two measurable components, acceleration and bias, that directly relate to perturbations of the underlying cellular dynamics. Acceleration is the proportional increase in the speed of methylation transitions across CpG sites, whereas bias is the degree of global change in methylation affecting all CpG sites uniformly. Using data from 7,028 participants from the Generation Scotland study, we found the age acceleration parameter to be associated with physiological traits known to impact healthy ageing. Furthermore, a genome-wide association study of age acceleration identified four genomic loci previously linked with ageing.