Digital twins improve forecasts of human immune changes
In two external populations, a personalized model reduced errors in estimating immune age by 32% and 42% compared with the best existing method.
Limit: Health benefits and slower aging were not measured
Edited by Boris Djordjevic · AgingResearch.News
In this modeling study, researchers used profiles of individual human immune cells to build ImmuneTwins. The system created “digital twins”: computer models personalized with each person's observations and updated as new measurements arrived.
In two external populations, it reduced errors in estimating immune age—how old the immune system appears—by 32% and 42% compared with the best existing method. During repeated influenza vaccination, the twins predicted the direction of immune-cell change 35% more accurately than the best published method. Their prediction error fell by 27% as observations accumulated. The twins also indicated that immune-cell and plasma-protein measurements moved about halfway back toward a persistent personal baseline within 90 days. Plasma is the liquid part of blood. The size of responses varied between challenges.
Why it matters
Distinguishing lasting personal immune features from temporary responses could help researchers interpret changes linked to immune aging.
Caveats
This modeling study assessed forecasting accuracy, not health benefits or slower aging. The preprint has not been peer-reviewed, and the abstract does not report participant numbers or ages.
The paper
Digital twins forecast individual immune trajectories
Peking University
bioRxiv, Oct 30, 2026, Open access, CC BY-ND, Preprint, not peer-reviewed
Comments
Be the first to comment.