June 22nd, 2026
Any sufficiently complex set of biological data assessed in a large population of various ages can be used as a basis to create an aging clock. Machine learning techniques are used to find algorithmic combinations of measurements that map to chronological age or observed mortality risk within the reference population. That algorithm then predicts age or mortality risk when used in people outside the reference population; where a person's predicted age is higher than chronological age this is thought to represent a higher burden of damage and dysfunction, and thus a greater biological age. Aging clocks have been show to work pretty well at a population level, but it remains difficult to establish how the measured parameters are determined by mechanisms of aging, and whether a clock assessment is of any practical use for one individual in the health and medical contexts.
Nonetheless, researchers are creating new clocks at a fair pace. Most omics based clocks use immune cells from a blood sample, and there has been some discussion over the years as to how relevant this is to aging in other tissues. Another point of interest has been how to separate variations in immune function that arise from stress, infection, and other transient causes from those arising from mechanisms of aging. With this background context in mind, today's open access paper reports on the use of a single cell assessment of chromatin accessibility in many different immune cell subtypes. Chromatin is structured nuclear DNA, with different sections either spooled and compact to prevent gene expression, or unspooled and accessible for gene expression. This structure is controlled by epigenetic decorations, and determines the behavior of the cell by determining which proteins are manufactured.
