This project aims to exploring and validating previously identified molecular targets of senomorphic compound and multimodal signatures of bone cancer on larger human cohort by integrating multi-modal datasets of UK biobank, using advanced computational approaches for discovery. Our previous work identified potential protein targets, plasma proteomic alteration and mice organ transcriptomic change responsive to bisphosphonates, a class of drugs that may act as senomorphic agents. Rather than focusing solely on bisphosphonates, we aim to use these as a gateway to systematically explore the molecular impact of senomorphic targets in non-skeletal ageing-related diseases through Olink proteomics, clinical records, and genetic data of UKB. This will inform future therapeutic repurposing and precision intervention strategies. In parallel, we will focus on multiple myeloma and prostate cancer as a model of blood and bone cancer, drawing from our spatial multi-omics studies that highlight subclone evolution and tumor-microenvironment crosstalk. Using UK Biobank, we will evaluate whether our previously identified markers of aggressive clones and immune evasion are detectable before diagnosis, and whether they contribute to poorer outcomes or relapse. In addition, whether those multi-modal features are related to other co-morbidities.
Leveraging machine learning, neural networks, and large language models, we will integrate high-dimensional data to:
1. Identify population validated circulating proteins and genetic variants associated with current senomorphic targets and effects, building evidence to boost drug repurposing for ageing related disease and bone cancer.
2. Providing evidence for biomarkers and multi-modal features for early detection of blood and bone cancer.
In summary, this project will support early detection, targeted prevention and precision medicine on cancer and ageing.
