View article

We performed the largest genetic study of Parkinson’s disease to date, involving analysis of 11.4M SNPs in 37.7K cases, 18.6K ‘proxy-cases’ and 1.4M controls, discovering 39 novel risk loci. In total, we identified 92 putative independent genome-wide significant signals including 53 at previously published loci. Next, we dissected risk within these loci, identifying 22 candidate independent risk variants in close proximity to one another representing multiple risk signals in one locus (20 variants proximal to known risk loci). We then employed tests of causality within a Mendelian randomization framework to infer functional genomic consequences for genes within loci of interest in concert with protein-centric network analyses to nominate likely candidates for follow-up investigation. This report also shows expression network signatures of PD loci to be heavily brain enriched and different in comparison to Alzheimer’s disease. We also used risk scoring methods to improve genetic predictions of disease risk, and show that GWAS signals explain 11-15% of the heritable risk of PD at thresholds below genome-wide significance. Additionally, these data also suggest genetic correlations relating to risk overlapping with brain morphology, smoking status and educational attainment. Further analyses of smoking initiation and cognitive performance relating to PD risk in more comprehensive datasets show complex etiological links between PD risk and these traits. These data in sum provide the most comprehensive understanding of the genetic architecture of PD to date, revealing a large number of additional loci, and demonstrating that there remains a …