Risk prediction of late-onset Alzheimer’s disease implies an oligogenic architecture
Qian Zhang, Julia Sidorenko, Baptiste Couvy‐Duchesne, Riccardo E. Marioni, Margaret J. Wright, Alison Goate, Edoardo Marcora, Kuan‐lin Huang, Tenielle Porter, Simon M. Laws, Australian Imaging Biomarkers and Lifestyle (AIBL) Study, Colin L. Masters, Ashley I. Bush, Christopher Fowler, David Darby, Kelly Pertile, Carolina Restrepo, Blaine R. Roberts, Jo Robertson, Rebecca Rumble, Tim Ryan, Steven Collins, Christine Thai, Brett Trounson, Kate Lennon, Qiao‐Xin Li, Fernanda Yevenes Ugarte, Irene Volitakis, Michael Vovos, Rob Williams, Jenalle E. Baker, Alyce Russell, Madeline Peretti, Lidija Milicic, Lucy Lim, Mark Rodrigues, Kevin Taddei, Tania Taddei, Eugene Hone, Florence Lim, Shane Fernandez, Stephanie R. Rainey‐Smith, Steve Pedrini, Ralph N. Martins, James D. Doecke, Pierrick Bourgeat, Jürgen Fripp, Simon Gibson, Hugo Leroux, David G. Hanson, Vincent Doré, Ping Zhang, Samantha C. Burnham, Christopher C. Rowe, Victor L. Villemagne, Paul Yates, Sveltana Bozin Pejoska, Gareth T. Jones, David Ames, Elizabeth Cyarto, Nicola T. Lautenschlager, Kevin J. Barnham, Lesley Cheng, Andy Hill, Neil Killeen, Paul Maruff, Brendan Silbert, Belinda M. Brown, Hamid R. Sohrabi, Greg Savage, Michaël Vacher, Perminder S. Sachdev, Karen A. Mather, Nicola J. Armstrong, Anbupalam Thalamuthu, Henry Brodaty, Loïc Yengo, Jian Yang, Naomi R. Wray, Allan F. McRae, Peter M. Visscher
Abstract
Abstract Genetic association studies have identified 44 common genome-wide significant risk loci for late-onset Alzheimer’s disease (LOAD). However, LOAD genetic architecture and prediction are unclear. Here we estimate the optimal P -threshold ( P optimal ) of a genetic risk score (GRS) for prediction of LOAD in three independent datasets comprising 676 cases and 35,675 family history proxy cases. We show that the discriminative ability of GRS in LOAD prediction is maximised when selecting a small number of SNPs. Both simulation results and direct estimation indicate that the number of causal common SNPs for LOAD may be less than 100, suggesting LOAD is more oligogenic than polygenic. The best GRS explains approximately 75% of SNP-heritability, and individuals in the top decile of GRS have ten-fold increased odds when compared to those in the bottom decile. In addition, 14 variants are identified that contribute to both LOAD risk and age at onset of LOAD.