Functional profiling of cellular aging for personalized immune health
Loading...
Author
Meirer, J.
Jagtap, S.
Paoletti, S.
Weder, G.
Heinemann, T.
DOI
Abstract
Understanding how the human immune system ages at the cellular level is essential for differentiating individual biological age from chronological age. The ChronoType project set out to address this challenge by developing a scalable, image-driven framework to quantify immune aging from peripheral blood samples. By combining high-content single-cell imaging with self-supervised representation learning and multiple instance learning (MIL), ChronoType enables the prediction of donor immune age and systematic evaluations of how immune cell phenotypes respond to perturbations. The project establishes an end-to-end computational pipeline that transforms raw microscopy images into donor-level age predictions by resolving single-cell populations and their morphological states underlying age-related changes in immune function. These results provide a foundation for future diagnostic and interventional applications in immune health. 1 3 Immune cell High content Single cell 2 isolation imaging analysis Attention-based identification of Drug screening novel morphotype signature High relevance Uninformative Significance Predictive modelling E.g. donor age Drug response Database from 179 healthy donor samples (21 to 73 years old) Biological explainability through joint molecular and morphological representation learning Figure 1: Overview of the single-cell imaging-based analysis pipeline. Immune cells from healthy donors are isolated and profiled by high-content imaging and single-cell analysis to identify distinct morphotypes. Attention-based models extract informative morphological signatures that enable predictive modelling of donor attributes and support drug screening, while joint molecular and morphological representations provide biological interpretability.
Publication Reference
CSEM Scientific and Technical Report 2025, p. 47–48
Year
2025