Stratification by a polygenic risk score of common variation aids in Alzheimer's disease rare variant discovery
Oluwatosin Olayinka, John J. Farrell, Congcong Zhu, Zainab Khurshid, , Eden R. Martin, William S. Bush, Margaret A. Pericak‐Vance, Li‐San Wang, Gerard D. Schellenberg, Jonathan L. Haines, Kathryn L. Lunetta, Xiaoling Zhang, Lindsay A. FarrerAbstract
INTRODUCTION
We utilized an Alzheimer's disease (AD) polygenic risk score (PRS) to discover associations with novel rare variants (RVs).
METHODS
PRSs for European ancestry (EA) participants of the Alzheimer's Disease Sequencing Project were calculated using summary statistics from a large genome‐wide association study. Participants were classified into high ( n = 5738) and low ( n = 5324) PRS groups based on the median PRS and on the lower and upper 35% of the PRS distribution.
RESULTS
Risk variants were disproportionately enriched in the low‐PRS group, while protective ones were disproportionately enriched in the high‐PRS group. Genome‐wide significant (GWS) associations for increased AD risk were identified with RVs spanning a 3.5‐Mb region on chromosome 14. GWS protective variants in ALDH9A1 , BICC1 , and PAN3 were identified in the upper 35% PRS group.
CONCLUSION
Our findings provide unique opportunities to study RVs whose effects are opposite to the risk conferred by the genetic background.