Epigenetic clocks are important tools in modern aging research. Typically, they use characteristic DNA methylation patterns in the genome to precisely predict a person's age and infer conclusions about the individual’s biological aging processes. However, why this works so well and what biological mechanisms are behind it remain unclear. Researchers at the Leibniz Institute on Aging – Fritz Lipmann Institute (FLI) have now investigated the basis of established epigenetic clocks and developed a new model, TFMethyl clock, that combines predictive accuracy with improved biological interpretability. The study, published in the journal Nucleic Acids Research, provides new insights into molecular aging processes and opens new possibilities for research into age-related diseases.

Jena. A person’s biological age does not always correspond to their chronological age, because lifestyle, environmental factors, or diseases can accelerate or even slow down the aging process. To make these differences measurable, so-called epigenetic clocks have been used worldwide for several years. Many of them rely on age-related changes in DNA methylation—chemical marks on DNA that can influence how genes are regulated.

Today, epigenetic clocks are among the most important biomarkers in aging research. They are used to investigate aging processes, better understand age-related diseases, and evaluate the effectiveness of potential anti-aging interventions. However, despite their high accuracy, the crucial question of which biological processes these clocks reflect has remained unanswered until now.

Taking a look inside the black box of epigenetic clocks

To close this knowledge gap, researchers at the Leibniz Institute on Aging—Fritz Lipmann Institute (FLI) in Jena, together with international partners from the Hebrew University of Jerusalem, the University of Edinburgh, and Queen Mary University of London, systematically investigated the DNA methylation sites of established epigenetic clocks and their regulatory function. The focus was on the binding sites of transcription factors that turn genes on or off, thereby controlling central processes of gene regulation. The study results have now been published in the journal Nucleic Acids Research.

“Epigenetic clocks have been providing impressively precise chronological age predictions. Nevertheless, surprisingly little is known about the biological mechanisms underlying these predictions,” explains Tushar Patel, first author of the study. The systematic analysis of DNA methylation sites used by most clock models revealed that many methylation sites do not lie within experimentally verified transcription factor binding sites.

“We therefore asked whether biological clocks could be improved by building them from regions of the genome that directly influence gene expression,” adds Steve Hoffmann, co-corresponding author of the study.

With this approach, the researchers identified a smaller group of regulatory DNA methylation sites that are closely enriched for aging processes. These include, among others, signaling pathways involved in interleukin-1β production and fatty acid metabolism. This points to molecular signaling pathways that may directly contribute to biological aging.

“Our model is one of the most accurate developed so far. More importantly, unlike other epigenetic clocks, it offers clues about the biological mechanisms that may directly shape how we age,” emphasizes Alena van Bömmel, the study’s last co-corresponding author, who has recently been appointed to the professorship of Neuroepigenetics and Data Science at the Ludwig-Maximilians University of Munich.

New epigenetic clock combines accuracy and interpretability

Based on these findings, the team developed a new epigenetic clock, the TFMethyl Clock. Unlike conventional models, this clock specifically considers DNA methylation sites that are both age-dependent and located within transcription factor binding sites. The approach was supplemented with new data-processing methods that cluster biologically similar methylation patterns. Overall, these strategies reduced confounding factors and further improved predictive performance and robustness.

The new clock achieved predictive accuracy comparable to established models and, in some cases, outperformed them. At the same time, it provides significantly deeper insights into the biological processes underlying the aging signatures and into the molecular networks to which they are connected. For example, the researchers found evidence of the involvement of signaling pathways associated with, among other things, the production of the inflammatory mediator interleukin-1β and fatty acid metabolism. Furthermore, about three-quarters of the target genes identified by the model exhibited age-dependent changes in gene activity—an indication that the selected markers do indeed reflect functionally relevant aging processes.

From biomarkers to tools in aging research

The study thus not only provides an improved method for age determination but also opens up new possibilities for systematically investigating the biological mechanisms of aging. In the long term, more biologically interpretable epigenetic clocks could help identify the molecular causes of age-related changes more precisely and inform the development of new approaches to the prevention and treatment of age-related diseases.

“In the future, epigenetic clocks could be much more than just precise biomarkers. They can help us understand which regulatory networks are involved in the aging process—and thus provide new avenues for research on aging,” the researchers said.

Publication
Enhancing the performance and interpretability of epigenetic clocks. Tushar Patel, Robert Schwarz, Konstantin Riege, Miri Varshavsky, Anne Richmond, Riccardo E Marioni, Hans A Kestler, Tommy Kaplan, Steve Hoffmann, Alena van Bömmel. Nucleic Acids Research, Volume 54, Issue 13, 22 July 2026, gkag661, https://doi.org/10.1093/nar/gkag661
https://academic.oup.com/nar/article/54/13/gkag661/8725950

Contact
Dr. Kerstin Wagner
Press & Public Relations
Phone: 03641-656378, Email: presse@leibniz-fli.de