How can someone know exactly how many years, months, and days he or she has lived? There are numerous computer sites that instantly calculate the end when the curious person provides their birthday. But what if someone has died and is unidentified or only a body part has been left after a terror attack, or if someone committed a crime and left only DNA?
This is now possible, thanks to an exceptionally accurate method for predicting chronological age from DNA developed by researchers at the Hebrew University of Jerusalem.
The team was led by Bracha Ochana and Daniel Nudelman, under the supervision of Prof. Tommy Kaplan – who heads the computational biology program at HUJI’s School of Computer Science and Engineering – along with Prof. Yuval Dor, and Prof. Ruth Shemer.
They developed an amazingly precise method to estimate a person’s age based on just a small sample of DNA. They said it not only could revolutionize forensic science by helping experts estimate a suspect’s age from just a trace of DNA – something existing tools struggle to do – but that it’s also a breakthrough with potential for medicine and aging research.
Using cutting-edge artificial intelligence, the scientists created a tool called MAgeNet that uses a simple blood test to determine a person’s chronological age – the number of years since birth, with a margin of error as small as 1.36 years (about 16 months and 9 days) for individuals under 50.
Prediction of age gives more opportunities in clinical diagnostics
“It turns out that the passage of time leaves measurable marks on our DNA,” Kaplan told The Jerusalem Post. “Our model decodes those marks with astonishing precision. We are currently working with the Israel Police on this, but not yet with the Institute of Forensic Medicine at Abu Kabir. It’s not clear how cells, tissues, and organs age.
“Our technique is based on two short genomic regions, using deep-learning networks analyzing DNA methylation patterns at a single-molecule resolution” that “achieve age predictions with a median error as low as 1.36 years in individuals under 50,” he said. “The method is unaffected by smoking, body-mass index, and gender, and has potential applications in forensics, aging research, and personalized medicine.”
THE TEAM said that the prediction of chronological and biological age from biological samples offers vast opportunities in clinical diagnostics, monitoring, and other types of research. Chronological age, defined as the amount of time since birth, correlates strongly with health status; biological age, while harder to define, can provide more accurate information on aging and the propensity for disease, the researchers said.
The study, published in Cell Reports under the title “Time is encoded by methylation changes at clustered CpG sites,” analyzed blood samples from more than 300 healthy people, as well as data from a decade-long longitudinal analysis of the Jerusalem Perinatal Study (JPS), led by Prof. Hagit Hochner at the Faculty of Medicine. As they show, the model worked consistently across a range of variables and even different signs of biological aging.
Methylation is a biochemical process in which a methyl group (CH3) is added to a molecule such as DNA, RNA, proteins, or others to alter how these molecules function within a cell, affecting processes like gene expression, protein function, and even disease risk.
DNA methylation is a crucial epigenetic modification where methyl groups are added to DNA, thereby influencing gene expression without altering its sequence. Epigenetics refers to changes in gene expression that don’t involve changes to the underlying DNA sequence. Instead, it involves factors like our behaviors and environment that can influence how our genes are read and used by the body.
TO ASSESS the potential applicability of one’s epigenetic chronological clock, they analyzed DNA methylation from urine and saliva samples collected from donors aged 24 to 74 years. DNA was extracted from each sample and carefully sequenced. By zooming in on just two key regions of the human genome, the team was able to read these changes at the level of individual molecules, then use deep learning to translate them into accurate age predictions.
This process can either repress or activate gene transcription, depending on the location of methylation within the genome. DNA methylation is vital for various biological processes, including general development, the development of diseases, and cellular differentiation. An analysis over a decade showed that early deviations from predicted age persist throughout life, and subsequent changes faithfully record time. Amazingly, accurate chronological age predictions are possible using as few as 50 DNA molecules, suggesting that age is encoded by individual cells.
“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.” The research also uncovered new patterns in how DNA changes over time, suggesting that our cells encode age both randomly and in coordinated bursts like ticking biological clocks. “It’s not just about knowing your age,” Shemer said, “it’s about understanding how your cells keep track of time, molecule by molecule.”
AS FOR biological age, which was not studied in depth by the team, it can be very different from chronological age. It is known that there are people in their 80s and 90s who are healthier and more active than those who are considerably younger, explained Prof. Naomi Habib, a computational neuroscientist at HUJI’s Edmond and Lily Safra Center for Brain Sciences, who researches cognitive decline and resilience.
“It’s a measure of your body’s functional state, influenced by genetics, lifestyle and environment, and it can differ significantly from chronological age,” Habib said. “Biological age can show up in both motor and cognitive abilities. One way to measure biological age is to look for changes in a molecule of DNA. We’re born with it, and it supposedly stays the same all through life, but in fact it undergoes changes – mutations that can bring cancer and other diseases.”
Most methylation-based epigenetic clocks were developed to reflect chronological as well as biological age, such that deviations from the former are interpreted as a reflection of faster or slower aging. The HUJI team focused on the molecular mechanisms that encode purely chronological age to better understand the underlying biology of how elapsed time is encoded in cells and to provide tools for research and forensic applications.
Habib added that there may be a connection in biological age to the lengths of telomeres – regions of repetitive DNA sequences at the end of a chromosome that protect its ends from becoming frayed or tangled. “Telomeres shorten as one gets older and the cells divide; they’re like a time of life. But these don’t divide everywhere, as in the brain.”
She and her team are “interested in solving the secret of cognitive resilience. There’s a huge amount of information from DNA and methylization and many influences on biological age, diagnoses, and personalized medicine. We’ll be able to find out whether certain medical treatment help or hurt the patient.”