Cancer development is influenced by complex interactions among genetic susceptibility, metabolic status, inflammatory responses, and lifestyle factors. Multi-omics biomarkers, including blood biochemistry, haematological traits, metabolic indicators, inflammatory markers, and polygenic risk scores, have shown strong associations with multiple cancer types. However, integrated analyses combining these biomarker layers in large prospective cohorts remain limited.
This project aims to identify biological markers associated with cancer incidence and to develop multi-omics predictive models for cancer risk using the UK Biobank resource. First, we will evaluate single-omics biomarkers-including haematological indices, biochemical measurements, metabolic traits, and genetic risk factors-to identify features linked to major cancers such as colorectal, breast, lung, and liver cancer. Next, we will integrate biomarkers across different omics domains using statistical modelling and machine-learning approaches to construct risk-prediction algorithms.
We will further assess interactions between biomarkers and lifestyle factors (e.g., smoking, alcohol use, diet, and physical activity) and perform subgroup analyses by age, sex, and metabolic status. The overarching objective is to improve early identification of high-risk individuals and to provide insights into biological pathways involved in tumour initiation and progression. The findings may inform the development of minimally invasive biomarkers and personalised cancer-risk stratification strategies.