Genome-wide association interaction analysis for Alzheimer’s disease.

Identification of epistasis is a challenging task that when successful gives new clues to systems-level genetics where the complexity of underling biology of human disease can be better understood. Though many novel methods for detecting epistasis have been proposed and many studies for epistasis detection have been conducted, so far few studies can demonstrate replicable epistasis. In the present work, we propose a minimal protocol for exhaustive genome-wide association interaction (GWAI) analysis that involves screening for epistasis over large-scale genomic data combining strengths of different methods and statistical tools. The different steps of this protocol are illustrated on a real-life data application for Alzheimer’s disease (a large cohort of 2259 patients and 6017 controls from France). Using this protocol, we identified AD-associated interacting SNPs-pair from chromosome 6q11.1 (rs6455128, the KHDRBS2 gene) and 13q12.11 (rs7989332, the CRYL1 gene) and male-specific epistasis between SNPs from chromosome 5q34 (rs729149 and rs3733980, the WWC1 gene) and 15q22.2 (rs9806612, rs9302230 and rs7175766, the TLN2 gene). The transcriptome analysis revealed negative correlation between expression levels of KHDRBS2 and CRYL1 in both the temporal cortex and cerebellum brain regions and positive correlation between the expression levels of CRYL1 and WWC1 in the temporal cortex brain region. A replication analysis strategy and a meta-analysis approach in independent data confirmed effects of some of the discovered interactions.