Assessing the rate of aging to monitor aging itself

Abstract

Healthy aging is the prime goal of aging research and interventions. Healthy aging or not can be quantified by biological aging rates estimated by aging clocks. Generation and accumulation of large scale high-dimensional biological data together with maturation of artificial intelligence among other machine learning techniques, have enabled and spurred the rapid development of various aging rate estimators (aging clocks). Here we review the data sources and compare the algorithms of recent human aging clocks, and the applications of these clocks in both researches and daily life. We envision that not only more and multiscale data on cross-sectional data will add momentum to the aging clock development, new longitudinal and interventional data will further raise the aging clock development to the next level to be trained by true biological age such as morbidity and mortality age.

Keywords: Aging rate; Artificial intelligence; Bioimages; Machine learning; Omics data.