Functional Connectivity And Genomic Association Analysis Of AD Brain Network

Alzheimer's disease(AD)is the most common form of dementia,accounting for 50-70% of all dementia cases.Currently,there are 35.6 million people with AD worldwide,and this number is expected to increase to 115.4 million by 2050.This is an incurable progressive and neurodegenerative disease that causes memory loss and other cognitive deficits.In recent years,twin studies have confirmed that AD has a very high heritability and is an important indicator for predicting the risk of the disease at the earliest.Brain imaging genetics is an emerging field of research aimed at studying the underlying genetic structure of brain structure and function by using different imaging modalities.However,not all changes in the brain are a direct result of genetic effects.In this thesis,using diffusion-weighted magnetic resonance imaging data to construct a brain network to obtain highly heritable brain network connections.Using data of twins got from the Human Connectome Project(HCP),the reliability of edge-level measures,namely fractional anisotropy,fiber length,and fiber number in the structural connectome as well as seven network-level measures,specifically assortativity coeffcient,local effciency,modularity,transitivity,cluster coeffcient,global effciency,and characteristic path length were evaluated using intraclass correlation coeffcients.In addition,estimates of the heritability of the reliable measures were also obtained.Using polygenic variance component model to evaluate the network heritability,obtain a genetic network with significant inheritance,and conduct reliability tests through intraclass correlation coeffcients(ICC)to extract high-reliability,high-heritable connections as the final heritable brain netwaok.Using the genotype and DTI data from the Alzheimer's Neuroimaging Initiative(ADNI),using DTI data to continue building the brain network,mapping the resulting brain network to a heritable brain network.This thesis is different from the previous research on node of brain network.We mainly focuses on the measurement of edge connection level,including fiber anisotropy,fiber length and fiber number.Furthermore,in order to avoid introducing potential biases in the imaging process,the imaging process flow is repeated,and the reliability of all connections is evaluated.For further imaging genetic association analysis,only highly reliable connectivity measurements will be considered.Using scan age and gender as covariates,a general linear regression model was used to study the association between each pair of candidate SNPs and connectivity measures.After implementing a strict Bonferroni correction,it was found that rs10498633 in SLC24A4 was significantly correlated with anisotropy,fiber length and number,including some of the brain hemispheres.With a lower level of significance at 5e-6,we observed significant genetic effect of SNPs in APOE,ABCA7,EPHA1 and CASS4 on various brain connectivity measures.