As one’s DNA coding is fixed, identifying environmental and lifestyle risk factors that modify genetic risk of colorectal cancer is an important preventive strategy. However, a major limitation for studying gene-environment interactions has been overcoming the multiple comparison issue due to availability of large datasets of harmonized and detailed phenotype data. 

Utilizing well-characterized studies with harmonized phenotype and genotype data from over 100,000 participants across GECCO, CORECT, and CCFR, we use innovative statistical methods to conduct genome-wide interaction analyses leading to the discovery of many novel findings. 

Colorectal cancer is particularly well-suited for testing gene-environment interactions because it is determined by several modifiable lifestyle, dietary, and pharmaceutical factors. To improve statistical power, we incorporate functional genomic data, as well as gene expression in normal colon tissue biopsies and colon organoids to inform testing and follow up on novel findings.

Our work on gene-NSAIDs interactions (PMID: 25781442) was highlighted by Dr. Francis Collins in his NIH Director's Blog as an example of precision medicine that he hopes will be substantially expanded.