Genome-wide investigation of gene-environment interactions using a case-only design for inflammatory bowel diseases

Postgraduate Support

Principal Investigator

Associated Principal Investigator

Background and current state of research

Gene-environmental (G×E) interactions are frequently cited to explain the gap between the identified genetic variants and the proportion of heritability they actually explain. Chronic inflammatory bowel diseases are only one example for which many genetic risk factors have been identified, but the interplay with environmental risk factors is hypothesized as important for actual disease manifestation. However, little progress has been made to identify such interactions in the past. A possible reason for this lack of identification is an insufficient power of previous case-control or cohort studies, given a presumably moderate size of genetic effects and their interactions with environmental factors. A powerful design for analyzing G×E interactions in epidemiological studies is the case-only design, where only information on diseased cases, but not on controls is required, and which is considerably more powerful than the standard case-control approach. In the past, case-only studies have been conducted following a candidate gene approach, but their usefulness has yet to be fully explored for genome-wide applications.

Figure 1: Number of publications in Pubmed (as of Jan 27, 2013) with keywords “gene environment interaction” and  “case only”

Number of publications in Pubmed (as of Jan 27, 2013) with keywords “gene environment interaction” and  “case only”

Our goals

We want to

  • develop and implement the case-only study design approach for genome-wide studies,
  • investigate gene-environment interactions for inflammatory diseases on a genome-wide scale using available data,
  • implement a pipeline facilitating the application of this analytical strategy to other diseases.

How to get there

In the methodological part (year one), the case-only approach will be generalized so as to allow use of genetic risk models other than the dominant and recessive ones employed so far. In the context of genetic studies, there are also a couple of methodological issues that need to be clarified. What are the effects of population stratification upon the validity of G´E interactions inferred from cases only? Can family-based tests like the TDT adapted to G´E analysis? If SNPs are only proxy for causal mutations, how does linkage disequilibrium affect the power to detect G´E? How do genotyping and environment-typing errors affect the validity of G´E interaction?

 In the medical research part (year two), the established case-only approach will be used to screen for gene-environmental interactions in IBD at a genome-wide scale. This will be done exploiting the extensive IBD database acquired by the popgen biobank since 2003. Several hundred cases with IBD provided biological samples in the past and have been genotyped. Study participants also completed questionnaires, which included selected environmental factors. This proof-of-principle application will be used to assess the feasibility and power of the case-only approach in a real-world data set.

In the technical part (year three), a bioinformatics pipeline will be implemented to facilitate the adoption of the analytical strategy in other disease studies. The mere scale of the data sets requires software tools to be developed for efficient and reliable analysis. These bioinformatics tools also will have to be extensively tested, validated, and documented.



Figure 1: Number of publications in Pubmed (as of Jan 27, 2013) with keywords “gene environment interaction” and  “case only”