Genome-wide epistasis analysis reveals significant epistatic signals associated with Parkinson's disease risk

Genome-wide association studies have increased our understanding of Parkinson's disease genetics by identifying common disease-associated variants. However, much of the heritability remains unaccounted for, and we hypothesized that this could be explained, in part, by epistasis, the statistical interaction between two or more genetic variants. Here, we developed a genome-wide non-exhaustive epistasis screening pipeline called variant-variant interaction through variable thresholds (VARI3) and applied it to diverse Parkinson's disease genome-wide association study cohorts. We used 14 cohorts of European ancestry (14 671 cases and 17 667 controls) as a discovery stage, identifying 14 significant candidate variant-variant interactions. We then used four independent cohorts (13 377 cases and 413 789 controls) as the replication stage, successfully replicating three epistasis signals located near to SNCA and within MAPT and WNT3. Admixture analysis showed that the epistatic effect on Parkinson's disease of those variants at these loci was observed in individuals with both European ancestry and Native American ancestry. We assessed the functional impact of the epistasis signals across a range of functional/-omics datasets, identifying significant single-variant expression quantitative trait loci across brain tissues, epistasis expression quantitative trait locus signals in whole blood, Parkinson's disease-relevant pathways and ontologies, and chromatin interactions between the regions of the interacting single nucleotide polymorphisms. In conclusion, we identified and replicated novel epistatic signals associated with Parkinson's disease risk across multiple cohorts with diverse genetic ancestry, highlighting their enrichment in pathways relevant to Parkinson's disease.