ICEP research

Epidemiological studies have reported associations between hundreds of exposures and cancer, and rapid advances in high-throughput ‘omics technologies will vastly expand the number of putative links. However, certain limitations of conventional epidemiology preclude confident causal inference. A major challenge now is to distinguish causal effects, of importance for understanding disease pathogenesis and prevention, from non-causal associations, which in-turn may inform risk prediction.

ICEP is organised in three work-packages with three cross-cutting strands.

Work package 1: Mendelian randomization (MR)

Leads: Lewis, Martin; Collaborators: Johansson, Muir, Ness

  • Exploits large scale international cancer GWAS consortia to determine causality of modifiable exposures (that are observationally associated with cancer risk).

Work package 2: Metabolomics & epigenetics

Leads: Relton, Davey Smith; Collaborators: Crosby, Donovan, Hamdy, Johansson, Maple, Ness

  • Large scale ‘omic profiling to investigate
    • associations of cancer risk and progression with epigenetic & metabolomic phenotypes
    • ‘omic signatures as exposure biomarkers
  • MR to investigate causality & mediation
    • new GWAS of DNA methylation
    • replication (Genetics of DNA Methylation group)
  • Biomarker demendelization

Work package 3: Mechanistic basis for potential interventions

Leads: Higgins, Timpson, Holly; Collaborators: Gaunt, Hamdy, Lewis, Ness, Perks

  • Systematic reviews of mechanistic studies
    • Automated mechanisms discovery
    • Reviews of specific mechanistic questions following established procedures
  • Recall by genotype studies
    • detailed phenotyping targeting groups defined by genotyping


  • Bioinformatics (Lead: Gaunt)
    • Laboratory information management
    • Integration with public datasets
    • Web-based data exploration utility
  • Training (Leads: Relton & Martin)
    • 2 x 4 year PhD studentships
    • X1 post-doc training fellowship
  • Knowledge transfer (Lead: Lane)

Prof Richard MartinICEP applies innovative causal analysis methods to robustly discriminate exposures that are causal cancer risk factors and possible targets for intervention, from non-causal biomarkers with potential predictive utility.

Professor Richard Martin

Prof Caroline ReltonOur main focus is on modifiable nutritional and lifestyle exposures, related intermediate phenotypes, and molecular biomarkers, in cancers of the prostate, lung, kidneys, head and neck, breast and ovaries.

Professor Caroline Relton