A multi-omics analysis of age-associated diseases to identify novel sendotypes

Abstract

The global population over the age of 60 is predicted to double by 2050, resulting in a population of 2.1 billion individuals. With more than half of the current aged population having more than two age-associated diseases, this demonstrates the need for a personalised medicine approaches for patient care, and less reliance on generalized treatment strategies.

Disease endotypes are currently used to subtype patients based on the biological mechanisms driving the disease, rather than using symptomologies. By employing the use of personalised medicine approaches alongside traditional endotypes, we can stratify patients by their current and future health outcomes. This could be done by analysing senescent based proteomics in the blood and using key signatures to predict and differentiate patients.

Several bioinformatical pipelines were developed and fitted to the proteomic profiles of age-associated disease patients with the aim being stratification and prediction of health states. In COVID-19 patients, differential expression analysis was carried out to identify senescence biomarkers and then test their predictive capabilities using machine learning (Chapter 3). For kidney disease, the same method was applied and then expanded upon by using proteomics to cluster patients into severe and mild subtypes. (Chapter 4).

This thesis has combined the use of bioinformatical tools alongside senescent protein labelling to stratify age-associated patients into senescence endotypes (sendotypes) based on current and future health risks.

Parts of the thesis are redacted due to an ongoing patent registration. The full version is embargoed until 29 February 2028

Date of Award	Feb 2026
Original language	English
Sponsors	Department for the Economy
Supervisor	Steven Watterson (Supervisor), Shu-Dong Zhang (Supervisor), Taranjit Singh Rai (Supervisor) & David Gibson (Supervisor)