Genome-Wide Association And Interaction Study On Quantitaitve TraitsofAlzheimer's Disease

Alzheimer's disease(AD)isthe most common form of dementia and increase people's chances of developing diseaseswith the aging of people in the world.It's a neurodegenerative disease in the brain of old population.AD patients will assume 4-5 times expenses compared with non-neurodegenerative diseases.It will bring patients' families heavy emotional burden and huge economic pressure.At present,ADhas become a serious public health problem and a worldwide medical problem.AD is characterized by pathological results at autopsy of the accumulation of amyloid-?(A?)protein in senile plaques and hyper-phosphorylated tau in neurofibrillary tangles in brain.Twins experiments confirm the high heritability of AD diseases and drive genomics research based on AD phenotypes,but there is a long way from fully unraveling the genetic code for complex diseases such as AD.Current research shows that the combined main effects of identified AD risk genes can only account for about 35%of AD heritability,and there is a lot of "missing heritability" can not be explained.The paper conducted research on a "focus one disease then others" basis.To address the"missing heritability" issue and underlying biological and disease mechanism,we performed genome-wide association studies(GWAS)on AD related quantitative traits(Qts),aims to explore and discover the early AD risk gene and early AD sensitive quantitative traits,develops algorithms and software tools.First,the paperproposed quality control(QC)method,QC standards and QC processes,conducted QC and demographic analysis base on ADNI genotypic data,5 PET-AV45 data and 3 CSF level data,then performed GWAS on 8 level quantitative traits.The GWAS results replicated APOC1,APOE,TOMM40 and PVRL2 four AD risk related genes,confirmed the ADNI data,plink QC precess method,and GWAS method used in this paper are available.It makes subsequent researches more meaningful.Secondly,the paper posed a hypothesis of SNP-SNP interaction,then designed the SNP-SNP interaction detected method,detected strategy and detectedprocessbased on linear regression(LR)model,developed a genome-wide interaction study(GWIS)tools.8 GWIS of quantitative traits results confirmed that strong interaction effects can be detected between small main effects SNPs and identified 40 genes.Among the GWIS results,10 AD related risk genes were replicated,12 psychiatric disease related genes and 18 new findings were found.It indicates that GWIS method base on Qts is an effective method for detecting AD related genes,especially in the conditions of GWAS invalid.Thirdly,the paper proposeda concept of quantitative traitsensitivity,theoretical basis and experimental flow,performed GWAS and GWIS based on 12 FreeSurfer measures,discussed the sensitivityof 20Qts among AV45 amyloid burden,CSF level and FreeSurfer measures.The results presented thatA? and Tau related phenotypes were more sensitive to GWIS,and the FreeSurferrelated phenotypes were more sensitive to GWAS,indicatedthat the SNP-SNP interaction base on A? and Taurelated phenotypes presented more sensitive to early AD and MCI,validated the rationality and effectiveness of the quantitative trait sensitivity analysis,and foundthat GWIS is an effective method for detecting early AD risk genes.In order to explore a new sensitive quantitative trait for detecting changes of early AD pathological and genetic mutations,the paper proposed6-conditions for designing new Qt,and posed a hypothesis of T-tau/A?42 combined Qt based on these conditions,then performed GWAS and GWIS base new combined phenotype,identied 212 genes.Among the 212 findings,39 AD related risk genes included APOC1,APOE andTOMM40 were replicated,21 psychiatric disease related genes and 152 new findings were found.It indicates that GWIS method base on Qts is an effective method for detecting AD related genes,especially in the conditions of GWAS invalid.The post hoc analysisresults showed the main effects of 24 SNP identified by T-tau/A?42 GWAS accounted for 11.3%of variance,combined APOE,age,gender and diagnosis states can account for 41.6%variance of T-tau/A?42,all previous studies were not on this scale.The interaction effects of GWIS results post hoc analysisaccounted for 3.1%-5.1%of variance independently.The results of these studies indicated that T-tau/A?42 is the ultra-high sensitivity phenotype and showed high potential predictive power.Finally,the paper proposed a multifactor integrated gene-gene interaction detetct method,solved the problem of gene-gene interaction correlation coefficient matrixcalculation,interaction detection and multifactor integration,developed a gene-gene interaction(IGENEv2.0)tool,implemented 2 correlation coefficient matrix calculation methods and 6gene-gene interaction calculation methods,then analyzed the differences among the algorithms base on IGENE.The significance of IGENE results was consistent with the GWIS.The literatures and tissue-specific analysis showed great significance of the GWAS and GWIS studies on the sensitive phenotype T-tau/A?42,and the findings expressed high consistency on biological explanation levelwarrant further replication.In summary,the results of genome-wide association studies based on the quantitative traits of human brain showed that the quantitative traits are closer to genetic mechanisms than case-control,higher signal-to-noise ratio and more powerful for detectingearly AD related risk genes.In addition,SNP-SNP interactionstudy,genetic susceptibilityanalysis of quantitativetraits,and development of gene-gene interaction detectmethod and detect tool,provided a systematic approach for miningrisk gene of complex disease,worth of being generalized.