Episcores across the lifecourse and their capacity to estimate cancer risk

Abstract

DNA methylation (DNAm) is a reversible biological process involving the addition of methyl
groups to the specific sites on the DNA. DNAm is causally involved in regulatory processes
throughout the lifecourse such as gene control, as well as implicated in disease progression,
and can itself also be affected by environmental and genetic factors. Making use of the role
of DNAm in regulating gene control, previous research has established the capacity of
DNAm to predict levels of circulating protein in the blood. These DNAm based models of
proteins (termed episcores) allow for the instrumentation of proteins in the wealth of datasets
where DNAm is available and proteomic measurements are not, providing promise for
biomarker discovery.
This thesis aims to first establish if previously published episcores which were created in
older populations are transferable to younger populations, in which there is significantly less
environmentally driven DNAm variance. Following this, I use longitudinal datasets to
discover phenotype-epsicore associations which may be useful as biomarkers in young
cohorts, where the phenotypes are physical health risk factors measured multiple years after
the episcore. Finally, I carry out a similar analysis in older cohorts to establish the predictive
capacity of episcores in early cancer diagnosis.
Overall, this thesis strengthens the use case of episcores as a biomarker discovery tool
complementary to genetic analyses, whilst also demonstrating that care should be taken
when choosing target populations for use of episcores due to limited transferability of
episcores between populations of different ages.

Date of Award	9 Dec 2025
Original language	English
Awarding Institution	University of Bristol
Supervisor	Paul D Yousefi (Supervisor), Ben Brumpton (Supervisor), Caroline Relton (Supervisor) & Matthew J Suderman (Supervisor)