Recently, Genome-Wide Association Studies (GWAS) have led to the discovery of numerous genetic polymorphisms involved in complex human diseases. However, these polymorphisms contribute only a little to the overall genetic variability of these diseases, suggesting the need for new kind of investigations in order to disentangle the so-called "missing heritability". The purpose of my PhD project was to investigate how different research strategies relying on statistical and biological considerations could help in determining whether part of this missing heritability could reside in interaction phenomena between genetic polymorphisms. Firstly, we applied different statistical methodologies and looked for interactions between polymorphisms that could influence the risk of venous thrombosis (VT). Even though this study was based on two large GWAS datasets, we were not able to identify pairwise interactions that survive multiple testing. This work suggests that strong interactive phenomena between common SNPs are unlikely to contribute much to the risk of VT. Second, by adopting a hypothesis-driven approach relying on biological arguments, we sought for interactions between microRNA related polymorphisms that could alter genetic expression. Using two large GWAS datasets in which genome-wide monocyte expression was also available, we were able to demonstrate the existence of two pairwise interaction phenomena on monocyte expression involving miRNAs polymorphisms: 1/ the expression of HLA-DPB1 was modulated by a polymorphism in its 3'UTR region with a polymorphism in the hsa-mir-219-1 microRNA sequence; 2/ similarly, the expression of H1F0 was influenced by a polymorphism in its 3'UTR region interacting with a polymorphism in the microRNA hsa-mir-659. Altogether, this project supports for the role of gene x gene interactions in the interindividual variability of biological processes but their identifications remain a tedious task requiring large samples and the development of new research strategies and methodologies.
