Method reproduces aging-clock links in human data
A computational analysis reproduced individual-level estimates for 38 epigenetic biomarkers and four traits using data from 15,677 individuals.
Our read
AI & data- Studied:
- Preprint: computational study of 15,677 individuals
- What changed:
- SARDINE reproduced individual-level estimates of aging-clock associations using summary data
- The catch:
- Validation measured agreement with individual-level associations, not the clocks’ accuracy in measuring biological aging
- Why it’s aging research:
- It builds or tests a tool for aging research.
Editor of record: Boris Djordjevic · AgingResearch.News
Data from 15,677 individuals helped validate SARDINE, a statistical method for analyzing epigenetic clocks, in a computational study. These clocks use chemical marks on DNA to estimate biological age. SARDINE accurately reproduced individual-level estimates of associations across 38 epigenetic biomarkers and four traits.
The method used summary statistics from studies linking DNA methylation, a type of chemical marking, to traits. It combined these with the clocks’ scoring rules and reference information about how methylation measurements vary together, without needing individual-level data. Applying SARDINE to diverse datasets revealed substantial differences in how biomarkers were associated with aging, diseases and lifestyle factors.
Why it matters
Epigenetic clocks are increasingly used to study biological aging, but their meaning and usefulness remain uncertain. The authors propose that existing summary data can support systematic evaluation and discovery of aging biomarkers.
Caveats
The reported validation tested agreement with individual-level associations, not whether the clocks accurately measured biological aging. The work is a preprint and has not yet been peer-reviewed.
The paper
SARDINE unlocks EWAS summary statistics for large-scale epigenetic biomarker analysis
University of Wisconsin-Madison
bioRxiv, Nov 6, 2026, Open access, CC BY-NC, Preprint, not yet peer-reviewed
Comments
Be the first to comment.