Israeli scientists have developed a remarkably accurate method to determine a person’s age from just a tiny DNA sample, offering new insight into how time is biologically recorded in a person’s cells, Hebrew University of Jerusalem announced on Tuesday. The breakthrough uses artificial intelligence to analyze changes in DNA at single-molecule resolution — achieving precision that far surpasses existing methods, according to the Press Service of Israel (TPS-IL).
Developed by researchers Bracha Ochana and Daniel Nudelman under the guidance of Professors Tommy Kaplan, Yuval Dor, and Ruth Shemer, the tool — called MAgeNet — can predict chronological age with a median error as low as 1.36 years for individuals under 50. Unlike previous models that require dozens or hundreds of genetic markers, MAgeNet needs data from just two specific genomic regions.
“It turns out that the passage of time leaves measurable marks on our DNA,” said Kaplan. “Our model decodes those marks with astonishing precision.”
The study, published in the peer-reviewed journal Cell Reports, focuses on DNA methylation, a natural process where chemical tags known as methyl groups attach to DNA, subtly influencing gene activity. While methylation has long been known to shift with age, the Hebrew University team is the first to demonstrate that examining just two loci at single-molecule detail can yield predictions with such minimal error.
“This gives us a new window into how aging works at the cellular level,” said Dor. “It’s a powerful example of what happens when biology meets AI.”
To validate their findings, the researchers analyzed over 300 blood samples from healthy individuals, alongside longitudinal data from the Jerusalem Perinatal Study. They found that their model’s accuracy remained stable regardless of smoking status, body mass index, sex, or even signs of accelerated aging — something previous age predictors have struggled to achieve.
One of the study’s most striking insights is the discovery of distinct patterns in how methylation accumulates over time. Instead of a smooth, gradual process, the data suggest that aging may occur in coordinated bursts at specific molecular sites — pointing to the existence of multiple, synchronized biological clocks.
“It’s not just about knowing your age,” said Shemer. “It’s about understanding how your cells keep track of time, molecule by molecule.”
This research also offers a major advance in forensic science, potentially enabling investigators to estimate a person’s age from a minute DNA sample left at a crime scene—something current technologies cannot do with such accuracy or efficiency.
MAgeNet’s potential practical applications span various fields, including forensic science, personalized medicine, aging research, insurance and life sciences, and public health and epidemiology.
Said Kaplan, “We’re just beginning to understand the timeline hidden in our own DNA.”
