#15 · ML / AI methods

Conserved transcriptomic clocks track mammalian ageing and mortality risk

The common thread is computational reduction of high-dimensional biological measurements into interpretable readouts of age. The universal transcriptomic hallmarks study integrates more than 11,000 transcriptomes from four mammals to derive conserved age and mortality clocks, together with modular network signatures of ageing. A cell-level parallel applies TrackerSci to 21 mouse tissues across ages, linking EdU labeling to snATAC profiles to map newborn-cell dynamics associated with aging-related composition changes. What is new is the pairing of conserved, mortality-linked transcriptomic clocks with cell-resolved accounting of newborn cells. The hB-PAC clock is non-invasive and draws on behavioral and morphological variables rather than transcriptomic data, so it sits at the periphery of this thread rather than defining it.