Pleiotropic meta-analysis of Alzheimer's disease and cognitive performance.

Abstract

A major challenge in the discovery of genetic risk loci for many complex diseases is reaching a sufficient sample size to adequately power the genetic analysis to detect modest effect sizes. With the increased number of publicly available genome-wide association studies (GWAS) summary statistics data over the last decade, a number of approaches has received increasing attention for exploiting the pleiotropy between correlated traits to improve statistical power. Pleiotropy is the phenomenon of one locus influencing multiple traits and it can aid to the discovery of novel genetic associations and point to underlying shared biological pathways. In the current study we aimed to explore the use of the pleiotropic statistical methods for estimating the genetic overlap among Alzheimer's disease (AD) and Cognitive performance (CP) and to discover novel AD associated loci. We utilised publicly available GWAS summary statistics for each trait; the AD summary statistics data by Kunkle et al. (2019) (N=63,926) and the CP data by Lee et al. (2018) (N=257,828). We estimated the genetic correlation between the two traits using the LD Score regression (LDSC) and jointly analysed the traits by using a modification of the meta-analysis method; the multi-trait analysis of GWAS (MTAG). LDSC identified a significant correlation between AD and CP (rg =-0.28, p=1.4e-05). Using MTAG, results showed a power increase of 35% to detect AD-specific risk loci despite the moderate rg between the two traits. A novel association was identified with rs7208590 (pMTAG =2.56e-08) in PPM1E which is independent of a previously reported association to the region covering BZRAP1-AS1, TEX14, RAD51C, PPM1E, TRIM37 genes. Moreover, we replicated two associations that have been previously reported by a pleiotropic study among AD and bipolar disease (Drange et al, 2019) and by a transethnic GWAS (Jun et al,2017); MTSS1L (rs2864755, pMTAG =2.88e-09) and HBEGF (rs7268, pMTAG =2.52e-11), respectively. The current study leveraged the power of GWAS and the phenomenon of pleiotropy to identify novel genetic loci and to construct biological hypotheses about the AD pathogenesis. Our knowledge of pleiotropy can aid to the diagnosis, intervention and personalised treatment via precision medicine.