Analysis Of Pathogenesis And Intragenic SNPs Interaction On Alzheimer’s Disease Based On Brain Transcriptome And GWAS Data

Alzheimer’s disease (AD) is the most common neurodegenerative disease, affecting the health of millions of elderly people. There is still no effective strategy for early-diagnosis and therapy of AD, the major reasons of which may consist of elusive progression of AD and unavailable genetic factors to predict accurately the risk of AD. With the rapid development of high-throughput technologies, the transcriptome and genome study of AD have become increasingly popular. In the first part of this thesis, we focused on the central issue in the field of AD that is how to detect the causative trigger from the observed pathological characteristics of AD. Aiming to this, we collected many brain transcriptome datasets from aging and AD and divided them into four stages relevant to the progression of AD, including Aging, ND_H, AD_H and ADD. Based on this four stages, we mainly focused on the disturbance of cellular functional module during the progression of AD. It is observed that the biosynthesis and energy metabolism is down-regulated at the early stage of AD, the signal transduction is enhanced at the intermediate stage, and apoptosis is elevated at the late stage. In contrast to damage-themed hypothesis, we propose that the down-regulation of energy metabolism in AD is a protective response of the neurons to the reduced level of nutrient and oxygen supply in the microenvironment. In the second part of this thesis, we mainly focused on the genome analysis of AD. Genetic factor is an important risk factor for AD. Although the GWAS(Genome-Wide Association Study) study of AD have identified many risk SNPs, only the main effects of single SNPs(single-nucleotide polymorphisms) have been evaluated without considering the contribution of SNPs interaction to AD. Based on four publicly available GWAS dataset of AD, we leverage the non-parametric GMDR(Generalized Multifactor Dimensionality Reduction) method to perform an analysis of intragenic SNPs interaction. We have discovered ten potential risk genes consistent across four independent datasets, including PDE1A, RYR3, TEK, SLC25A21, LOC729852, KIRREL3, PTPN5, FSHR, PARK2, and NR3C2. Multiple level of functional evidences demonstrated that these genes are highly related to the pathologies of AD. In addition, we have also shown that intragenic high-order SNPs interaction by GMDR may be complementary to the conventional single SNP association analysis to detect more potential risk genes and the optimal order for SNPs combination is four.