Comprehensive Analysis of Alzheimer's Disease Biologically Candidate Causal Genes Revealed by Function Association Study With GWAS

Although genome-wide association studies (GWAS) have successfully identified many risk loci associated with Alzheimer's disease (AD), the dispute about missing heritability and weak interpretability must be resolved to reveal the causal genes in the risk loci and explain the mechanism of AD. The aim of this study was to overcome the problems that involve moving from the risk loci to causal genes and to understand the genotype-to-phenotype relationship of AD. We integrated the prediction results from different methods (e.g., DAPPLE, DEPICT, Prix Fixe, etc.) based on GWAS data combining protein-protein interaction networks, gene functions, co-function networks or expression quantitative trait loci data. A total of 43 plausible causal genes of AD were identified, including eight high-confidence AD causal genes (BIN1, CR1, CLU, HMHA1, MS4A4A, MS4A6A, PICALM and PVR). Then the landscape of these 43 causal genes was generated. Gene Ontology analysis showed that these identified causal genes were enriched in lipid/lipoprotein-related complexes and processes, supporting that lipid/lipoprotein homeostasis has a critical role in AD. The distinct spatial-temporal expression patterns of these causal genes illustrated that they played diverse roles in different cell types and developmental stages. The top eight causal genes were dysregulated in AD cases compared with their expression in normal controls, indicating that these genes are important in the pathophysiology of AD. Results from our study could provide meaningful clues for understanding AD pathogenesis. Together, further functional validation of the causal genes of AD will help identify potential targets for AD therapy.