The accumulation of senescent cells is increasingly recognized as a hallmark of biological aging, correlating with various chronic medical conditions. These senescent cells exhibit a unique secretory profile known as the senescence-associated secretory phenotype (SASP), which has implications for cellular senescence propagation across tissues.
In this study, the researchers developed and validated a composite SASP Score, integrating large-scale population proteomics data with a semi-supervised deep learning framework. The analytical workflow involved several key steps:
- Selection of SASP Proteins: The study began with the identification of biologically curated SASP proteins.
- Model Development: A Guided autoencoder with Transformer (GAET) model was created using data from the UK Biobank Pharma Proteomics Project (UKB-PPP).
- Evaluation and Validation: The model underwent internal evaluation and association analyses within the UK Biobank, followed by external validation and longitudinal assessment in an independent randomized clinical trial cohort.
The findings indicated that the deep learning-based SASP Score was a robust, independent predictor of mortality risk and the onset of serious chronic medical conditions, including dementia, chronic obstructive pulmonary disease (COPD), myocardial infarction, and stroke.
Moreover, in an independent cohort, multimodal exercise was shown to significantly alter the trajectory of the SASP Score over an 18-month period. This suggests that the SASP Score could serve as a valuable biomarker for tracking systemic cellular senescence burden and evaluating the outcomes of geroscience-guided interventions.
What this study does not show
This study does not establish a direct causal relationship between changes in the SASP Score and specific health outcomes, nor does it provide evidence of clinical utility in a real-world setting.
Evidence scorecard
- Analytic validity: established
- Clinical validity: established
- Clinical utility: not established
The evidence is based on primary research involving human participants.