Researchers from the Hebrew University of Jerusalem have unveiled a groundbreaking method for predicting chronological age from DNA, achieving a median error of just 1.36 years for individuals under 50. Utilizing a deep learning algorithm, known as MAgeNet, the technique analyzes DNA methylation changes at single-molecule resolution across two genomic regions. This innovative approach is unaffected by variables like smoking, BMI, and sex, making it ideal for applications in forensics, aging research, and personalized medicine.
Led by Bracha Ochana and Daniel Nudelman under the guidance of renowned professors, the study draws from over 300 blood samples, including extensive data from the Jerusalem Perinatal Study. The findings, published in Cell Reports, highlight how DNA encodes age through both random and coordinated changes, illuminating the biological clock at the cellular level. This research not only enhances medical treatment efficiency but also provides forensic scientists a powerful new tool for age estimation from DNA traces, revolutionizing both fields.