Genetic and polygenic risk score analysis for Alzheimer's disease in the Chinese population

Abstract

Introduction Dozens of Alzheimer's disease (AD)‐associated loci have been identified in European‐descent populations, but their effects have not been thoroughly investigated in the Hong Kong Chinese population. Methods TaqMan array genotyping was performed for known AD‐associated variants in a Hong Kong Chinese cohort. Regression analysis was conducted to study the associations of variants with AD‐associated traits and biomarkers. Lasso regression was applied to establish a polygenic risk score (PRS) model for AD risk prediction. Results SORL1 is associated with AD in the Hong Kong Chinese population. Meta‐analysis corroborates the AD‐protective effect of the SORL1 rs11218343 C allele. The PRS is developed and associated with AD risk, cognitive status, and AD‐related endophenotypes. TREM2 H157Y might influence the amyloid beta 42/40 ratio and levels of immune‐associated proteins in plasma. Discussion SORL1 is associated with AD in the Hong Kong Chinese population. The PRS model can predict AD risk and cognitive status in this population.

Polygenic Score Models for Alzheimer’s Disease: From Research to Clinical Applications

“…Notably, a recent study also implies the involvement of the peripheral immune system in AD pathogenesis (Zhou et al, 2018), while other studies demonstrate associations between AD polygenic risk and plasma proteins (Morgan et al, 2017;Korologou-Linden et al, 2019b;Zhou et al, 2020) or metabolites (Papassotiropoulos et al, 2005;Korologou-Linden et al, 2019b). Specifically, by applying the proximity extension assay to plasma proteomic analysis, we investigated 280 proteins and revealed potential protein candidates (i.e., osteopontin and neurocan core protein) along with a protein network associated with AD polygenic risk-again implying global changes in plasma profiles that might be modulated by polygenic risk (Zhou et al, 2020).…”

“…Some other studies investigated the possible associations between AD polygenic score models and the risk of conversion to AD or mild cognitive impairment (MCI) (Filipovych et al, 2012;Rodríguez-Rodríguez et al, 2013;Verhaaren et al, 2013;Carrasquillo et al, 2015;Desikan et al, 2017;Tosto et al, 2017;Kauppi et al, 2018;Chaudhury et al, 2019;Elman et al, 2019;Ajnakina et al, 2020;Altmann et al, 2020;Andrews et al, 2020), cognitive function (Louwersheimer et al, 2016;Del-Aguila et al, 2018;Ge et al, 2018;Kauppi et al, 2018Kauppi et al, , 2020Porter et al, 2018a,c,b;Stephan et al, 2018;Tan et al, 2018Tan et al, , 2019Tasaki et al, 2018;Korologou-Linden et al, 2019a;Han et al, 2020;Zhou et al, 2020), and memory function (Barral et al, 2012;Verhaaren et al, 2013;Marden et al, 2014Marden et al, , 2016Adams et al, 2015;Carrasquillo et al, 2015;Mormino et al, 2016;Hayes et al, 2017;Marioni et al, 2017;Axelrud et al, 2018;Ge et al, 2018;Porter et al, 2018a,b,c;Tan et al, 2018Tan et al, , 2019Altmann et al, 2020). Notably, given ...…”

“…Notably, given that the brain's structure and functions are closely associated with cognitive ability, several studies have also investigated the use of polygenic score models to predict brain status including changes in brain structure (Sabuncu et al, 2012;Habes et al, 2016;Harrison et al, 2016;Nho et al, 2016;Desikan et al, 2017;Foley et al, 2017;Hayes et al, 2017Hayes et al, , 2020Xiao et al, 2017;Ge et al, 2018;Kauppi et al, 2018;Li et al, 2018;Tasaki et al, 2018;Chandler et al, 2019Chandler et al, , 2020Tan et al, 2019;Wang et al, 2019;Altmann et al, 2020;Matloff et al, 2020;Zhou et al, 2020) and function (Xiao et al, 2017;Axelrud et al, 2019;Chandler et al, 2019Chandler et al, , 2020. Moreover, some studies investigated biochemical changes indicative of brain status, such as AD pathological hallmarks including amyloid-beta (Aβ) load and tau tangles (Mormino et al, 2016;Darst et al, 2017;Desikan et al, 2017;Laiterä et al, 2017;Ge et al, 2018;Porter et al, 2018a,b,c;Tan et al, 2018Tan et al, , 2019Tasaki et al, 2018;Leonenko et al, 2019a;Yu et al, 2019;Altmann et al, 2020), enzyme activity in brain samples (Martiskainen et al, 2015;…”

“…For instance, linkage disequilibrium-based pruning, which removes variants in high linkage disequilibrium, or linkage disequilibrium-aware clumping, which simultaneously removes variants in high linkage disequilibrium while retaining variants with the smallest p -values, have been applied to select the most informative variants to construct a polygenic score model. In addition to p -value–based selection, other statistical learning methods such as lasso regression, which can select the most informative variants for AD classification by removing variants minimally associated with the disease, have been also incorporated into polygenic risk analysis for AD ( Romero-Rosales et al, 2020 ; Zhou et al, 2020 ).…”

“…Regarding model construction, PLINK and PRSice are the most widely used tools to select variants and construct polygenic score models. Other statistical analysis methods, such as linear support vector machine ( Filipovych et al, 2012 ), lasso regression ( Romero-Rosales et al, 2020 ; Zhou et al, 2020 ), multilocus genotype patterns analysis ( Barral et al, 2012 ), and decision tree ( Yokoyama et al, 2015 ; Porter et al, 2018c ), have also been adopted to construct polygenic score models for AD.…”