Researchers astatine nan Hebrew University person developed an exceptionally meticulous method for predicting chronological property from DNA, based connected 2 short genomic regions. Using heavy learning networks analyzing DNA methylation patterns astatine a single-molecule resolution, they execute property predictions pinch a median correction arsenic debased arsenic 1.36 years successful individuals nether 50. The method is unaffected by smoking, BMI, and sex, and has imaginable applications successful forensics, aging research, and personalized medicine.
A squad of researchers astatine nan Hebrew University of Jerusalem, led by Bracha Ochana and Daniel Nudelman, nether nan supervision of Prof. Tommy Kaplan, Prof. Yuval Dor and Prof. Ruth Shemer, has developed a remarkably precise method to estimate a person's property based connected conscionable a mini DNA sample-offering breakthrough imaginable for medicine, aging research, and forensic investigations.
Using cutting-edge artificial intelligence, nan scientists created a instrumentality called MAgeNet that tin find a person's chronological age-the number of years since birth-with a separator of correction arsenic mini arsenic 1.36 years for individuals nether 50. And each it takes is simply a elemental humor draw.
It turns retired that nan transition of clip leaves measurable marks connected our DNA. Our exemplary decodes those marks pinch astonishing precision."
Prof. Tommy Kaplan, Hebrew University of Jerusalem
The concealed lies successful really our DNA changes complete clip done a process called methylation-the chemic "tagging" of DNA by methyl group (CH3) . By zooming successful connected conscionable 2 cardinal regions of nan quality genome, nan squad was capable to publication these changes astatine nan level of individual molecules, past usage heavy learning to construe them into meticulous property predictions.
The study, published successful Cell Reports, analyzed humor samples from complete 300 patient people, arsenic good arsenic information from a decade-long longitudinal study of nan Jerusalem Perinatal Study (JPS), led by Prof. Hagit Hochner from nan Faculty of Medicine. As they show, nan exemplary worked consistently crossed a scope of variables-like smoking, assemblage weight, sex, and moreover different signs of biologic aging.
Beyond imaginable aesculapian uses, nan method could besides revolutionize forensic subject by allowing experts to estimate a suspect's property from conscionable a trace of DNA-something existing devices struggle to do.
"This gives america a caller model into really aging useful astatine nan cellular level," said Prof. Dor. "It's a powerful illustration of what happens erstwhile biology meets AI."
The investigation besides uncovered caller patterns successful really DNA changes complete time, suggesting our cells encode property some randomly and successful coordinated bursts-like ticking biologic clocks. "It's not conscionable astir knowing your age," added Prof. Shemer. "It's astir knowing really your cells keeps way of time, molecule by molecule."
Why it matters: This investigation could reshape really we attack health, aging, and personality successful nan future. From helping doctors tailor treatments based connected a person's existent biologic timeline to giving forensic investigators a powerful caller instrumentality for solving crimes, nan expertise to publication property straight from DNA opens nan doorway to breakthroughs crossed science, medicine, and law. It besides deepens our knowing of really aging works-bringing america 1 measurement person to decoding nan body's soul clock.
Source:
Journal reference:
Ochana, B.-L., et al. (2025). Time is encoded by methylation changes astatine clustered CpG sites. Cell Reports. doi.org/10.1016/j.celrep.2025.115958.